# The Titans · How judgment develops Follow Buffett, Bezos, Howard Marks, and Mark Leonard through the decisions, mistakes, pressures, and revisions in their original work. Reconstruct what they knew at the time, then test your judgment in historical cases, constructed 2026 businesses, and possible futures. ## Coverage This course follows development over time as well as business mechanisms. The Works inventory records the research actually completed, with separate statuses for full reading, partial reading, located works and unavailable material. A listed work is not necessarily a work read. Coverage is bounded by the dated inventories below; books, recordings and private material that remain inaccessible or unreviewed are gaps, not silently counted as complete. Warren Buffett A chronological development study built from a defined inventory of 90 serial primary documents: 29 located Buffett Partnership items dated 1957-1970, plus a separately disclosed unavailable 1956 first-letter gap, 60 Berkshire annual letters or reports covering 1965-2024, and Buffett's November 2025 Thanksgiving message. All 90 located serial documents were read in full from their openings through their signed closings and any Buffett-authored appendices. The reading also inventories major authored articles, the Columbia essay, the CNBC annual-meeting archive, the 2025 succession announcement, Greg Abel's first letter, and important books or collections. Unread books, partially reviewed standalone works, and access gaps remain explicit below. Jeff Bezos Defined development corpus of 35 works dated 1997 through 2021. All 24 Bezos-signed Amazon shareholder letters were read in full. Four additional original transcripts or testimony records and one independent investigation were read in full. One interview, one congressional report and one NLRB case record were read in part. The authored collection and two biographies were located but not read because full copies were unavailable. Andy Jassy's 2021 through 2024 letters were excluded after signature review. Four later Jassy letters are listed separately as partial authorship checks; they are not part of the 24-letter Bezos-authored corpus. Howard Marks The public inventory covers all 165 dated memos listed in Oaktree’s official archive through September 9, 2026, spanning October 12, 1990 through April 9, 2026, plus the official complete collection, three authored book editions and three material original programs. All 165 dated public memos and the founders transcript were read in full. The collection remains Read in part as a compilation because its front matter, contents pages, repeated legal disclosures and every decorative image were not reviewed page by page; all dated memo texts contained in it are tracked separately as Read in full, and meaningful figures identified during those readings were inspected. Full texts of the three books were not available locally or freely from their publishers, so book coverage is limited to publisher records and original later accounts of their formation. Oaktree refers to client-only writings outside the public archive; those remain unavailable and are not included as public memos. Mark Leonard All 19 letters in Constellation’s official president’s-letter archive and all 14 documents in its shareholder-Q&A archive were read in full for this development path. Reporting periods, actual letter dates and individual respondents were checked. The Q&A archive includes company responses, a notice and a CFO interview, not only Leonard’s own work. Original webcasts are inventoried separately as located but not reviewed. No complete authored-book bibliography or complete interview corpus has been established; the course does not claim all of Leonard’s work. Early-life and founding episodes rely on explicitly retrospective primary accounts where contemporary material is missing. Two later company announcements establish succession plans, not an independently demonstrated result. Historical stages are educational reconstructions of sourced decisions. They do not invent private conversations or promise that all information available to the original decision-maker is known. The 2026 adaptations and later stress tests are fictional scenarios with stated assumptions. They are not market forecasts or claims about what any of the four would do today. Quantities in a constructed case are teaching assumptions unless explicitly identified as historical source data. The teaching is original paraphrase and application. Direct quotations, if any, are explicitly marked and located. Letters and memos written by successors, collaborators, narrators, or commentators are not attributed to the featured author. Source material is linked for checking rather than reproduced in full. Written responses, revisions, recall ratings, and the capstone rubric support practice. They are not automatic professional certification, a validated investment test, or a measured expertise percentile. A good answer reconstructs a mechanism, considers alternatives, handles uncertainty, and updates responsibly. Agreement with a historical decision is not the scoring rule. ## Foundations · Study a decision before its outcome A business history gives you an advantage its participants did not have. You know which company survived and which decision became famous. That advantage can make the lesson less useful. Once you know the result, an uncertain investment begins to look inevitable, an ordinary warning becomes a decisive signal, and a rival's reasonable alternative begins to look foolish. This course asks you to reverse that process. Recover the decision as it looked before the answer was available. The four studies concern different jobs. Buffett allocates capital across businesses and securities. Bezos's letters concern building and operating Amazon. Marks writes about investment judgment, especially risk and credit. Leonard's letters concern acquiring and operating vertical-market software businesses. A practice can make sense in one of those jobs and misfire in another. Before borrowing it, identify who owns the assets, what they can change, when cash must be returned, and what failure would cost. The readings reconstruct selected arguments from the original letters and memos. They do not provide a biography of every period or a survey of every publication. Each source record states its author, reporting or publication period, and a usable passage locator. A letter is strong evidence for what its author said publicly. It can also report a transaction or operating result. It is weaker evidence for the claim that one favored practice caused all the organization's success. Other people, conditions, competing explanations, and unreported alternatives remain relevant. Use three distinct judgments. First reconstruct the author's position accurately enough that a sympathetic reader would recognize it. Next evaluate whether the evidence supports that position in the case. Finally decide whether the mechanism transfers to a different setting. You can succeed at the first task while rejecting the third. You can also make a defensible decision that differs from the historical actor's decision. The scenario laboratory separates history, a constructed 2026 setting, and a hypothetical later development. Write a choice and a rationale before opening the debrief. The historical reconstruction identifies its evidence limits; it does not invent a private management conversation. The modern and future stages introduce fictional facts and options. Their numbers are assumptions for an exercise, not company disclosures or forecasts. The labels stay visible because changing the date should not quietly change a teaching invention into a factual claim. Keep your initial rationale. After the debrief, identify the most important missing step and write a revision separately. A revision that changes only the final choice teaches little. A useful revision identifies the assumption that failed, the evidence that contradicted it, and the part of your decision procedure that should change. If your original reasoning was sound but an unfavorable possibility occurred, explain that too. The purpose of preserving the first answer is to distinguish learning from rewriting your history. Finish a case by carrying one question into an unfamiliar decision. Do not promise yourself that you will always be patient, always move quickly, or always trust local managers. Write the conditions under which patience, speed, or delegation makes sense. Those conditions are what make a principle usable. ### Work the question Choose a business success you know. Write one fact known before the decision, one fact learned afterward, and one causal explanation that the outcome alone cannot establish. Then state what a fair historical exercise must withhold. ### Compare with the guide’s reasoning Keep contemporary information and later interpretation separate. Withhold the result and later evidence that would make the choice easy. Do not withhold a contemporary fact merely because it undermines the celebrated explanation. A strong answer names an alternative cause and identifies evidence that could distinguish it. These are the guide’s assessment criteria, not a historical quotation. ### Sources - Warren E. Buffett, Berkshire Hathaway 2012 shareholder letter (2012 reporting year; published 2013), Dividends, including reinvestment, acquisitions, repurchases, and the conditions behind the payout argument.. https://www.berkshirehathaway.com/letters/2012ltr.pdf Evidence note: Written by the allocator. His assessments of acquisition value and alternative outcomes are his judgments, not independent causal measurements. - Jeffrey P. Bezos, Amazon 2016 shareholder letter (2016 reporting year; published 2017), Resist Proxies and High-Velocity Decision Making.. https://www.aboutamazon.com/news/company-news/2016-letter-to-shareholders Evidence note: Use the signed Bezos letter, not the hosting page’s staff byline. Distinguish his reported practices from proof that they caused Amazon’s results. - Howard Marks, Fewer Losers, or More Winners? (September 12, 2023), Putting It in Brief; Not Risk Avoidance; The Role of Risk Bearing; What About in Practice?. https://www.oaktreecapital.com/insights/memo/fewer-losers-or-more-winners Evidence note: A conceptual and retrospective memo. It distinguishes bond upside, strategies needing winners, and risk control. It does not establish that a reader can identify investment bargains. - Mark Leonard, Constellation Software 2014 president’s letter (2014 reporting year; published 2015), Pages 1–5, particularly intangible-asset condition, reliable capital, and business-unit decentralization.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: The URL reflects later file hosting, not a 2026 letter. Business figures in the original refer to their stated historical periods. Author explanations remain author explanations. ## Foundations · Make the decision reviewable A principle becomes operational when another person can see what it asks you to do, why, and under which conditions you would stop. The decision record used here is an original teaching device. It organizes the reasoning so a later review can find a mistaken assumption instead of merely congratulating or blaming the decision-maker. Begin with the decision itself. Name the action, the owner, the deadline, and the alternatives. Include doing less, proceeding in stages, negotiating different terms, and declining altogether when those are real options. An option that you cannot finance or implement is not yet an alternative. A list of aspirations is not a decision. Improving customer experience, growing profit, and reducing risk can all be desirable without identifying which commitment should be made now. Next identify the mechanism. Describe how the action changes a customer's behavior, an operating capability, a cash flow, or a source of uncertainty. The causal chain should have intermediate observations. If you predict that faster support increases retention, specify which customers are affected, what counts as a resolved problem, when renewals occur, and what other changes might explain the renewal result. A rise in support messages answered does not establish that problems were solved. A renewal does not by itself establish that faster support caused it. Separate evidence from assumptions. Evidence includes dated records, observations, and calculations whose inputs can be checked. An assumption fills a gap. It can be necessary and reasonable, but it should remain visible. Give the most consequential assumptions a plausible adverse alternative. Confidence is not evidence. It is a judgment about evidence, model uncertainty, and missing information. State what your confidence depends on rather than substituting an emphatic adjective for a reason. Then specify the constraint that cannot be violated. This might be a contractual payment, service obligation, access to customer data, safety requirement, or minimum cash reserve. A reversible interface change can sit inside an irreversible contractual commitment. A short trial can leave long-lived reputational or operational consequences. Do not classify a decision as reversible merely because its software can be rolled back. Finally set a review trigger before the result arrives. It should identify an observation, a responsible person, and the action to reconsider. Include both a failure signal and a reason to expand. Otherwise an experiment can linger indefinitely, or a successful test can remain too small to matter. A trigger is not a substitute for judgment. An unexpected event can make the original metric inadequate. Record why it did, so flexibility does not become permission to move every disappointing target. Two source distinctions sharpen this procedure. In Amazon's 2016 letter, Bezos distinguishes lighter treatment of reversible decisions from escalation when objectives are misaligned. In Berkshire's 2012 letter, Buffett describes the danger of deciding on the desired answer and then constructing its rationale. The guide's record combines neither into an automatic rule. It asks whether the process fits the stakes and whether the reasoning would remain persuasive if another person proposed the same action. After a decision, review the process before reading the outcome into it. What information was available but ignored? What could not reasonably have been known? Did the team follow its stated conditions? Did a favorable external event rescue an otherwise weak plan? A useful organization can learn from lucky successes as well as painful failures. ### Work the question Write a decision record for one live or fictional choice. Include two feasible alternatives, the causal chain, the weakest assumption, a binding constraint, an owner, and the evidence that would cause you to revise. Then write the strongest objection to your preferred option. ### Compare with the guide’s reasoning The objection should threaten the mechanism, economics, feasibility, or acceptable downside. Merely stating that execution must be good is insufficient. The alternative must be possible under the same resource constraints. The review condition should be observable and capable of changing the choice, rather than a vague promise to monitor results. ### Sources - Jeffrey P. Bezos, Amazon 2016 shareholder letter (2016 reporting year; published 2017), Resist Proxies and High-Velocity Decision Making.. https://www.aboutamazon.com/news/company-news/2016-letter-to-shareholders Evidence note: Use the signed Bezos letter, not the hosting page’s staff byline. Distinguish his reported practices from proof that they caused Amazon’s results. - Warren E. Buffett, Berkshire Hathaway 2012 shareholder letter (2012 reporting year; published 2013), Dividends, including reinvestment, acquisitions, repurchases, and the conditions behind the payout argument.. https://www.berkshirehathaway.com/letters/2012ltr.pdf Evidence note: Written by the allocator. His assessments of acquisition value and alternative outcomes are his judgments, not independent causal measurements. ## Foundations · Compare money on the same basis Before comparing allocation philosophies, make sure the quantities being compared describe the same thing. Revenue is what a business earns from sales under its accounting policies. Profit deducts the relevant expenses. Cash flow records cash movements. A cash balance is a stock measured at a date, whereas annual cash flow is measured over a period. A profitable company can be short of cash if customers pay late, inventory grows, or debt comes due. A cash-rich company can still be consuming the capability that generates future receipts. Capital allocation asks where resources should go next. It therefore concerns incremental future consequences. A historical return on an old purchase does not establish the return on a new purchase at today's price. A return measured against depreciated assets can look exceptional while the next expansion is ordinary. A share repurchase can increase earnings per share and still transfer too much value to departing owners if the purchase price is excessive. The accounting ratio and the economic proposition need separate examination. The following arithmetic is entirely constructed for a 2026 teaching exercise. All amounts are US dollars. A project costs $120,000 at the start of 2026. Assume it produces $50,000 of net cash at each of the next three year-ends, has no residual value, and requires no further investment. Those cash receipts total $150,000. The $30,000 difference is a cumulative cash surplus over three years. Calling it a 25 percent annual return would be wrong; the 25 percent is the undiscounted surplus divided by the initial outlay over the whole three-year exercise. Now use a fictional teaching discount rate of 10 percent per year. This is an assumption, not a recommendation or an estimate of any actual company's cost of capital. The present value of the three receipts is 50,000/1.10 + 50,000/1.10² + 50,000/1.10³, approximately $124,343. Subtract the initial $120,000 and the net present value is approximately $4,343. The favorable margin in this model is small enough that a change in cash generation can matter more than precision in the displayed result. Suppose the same fictional project instead generates $40,000 at each year-end. The present value at the same teaching rate is approximately $99,474, giving a net present value of approximately negative $20,526. This calculation does not attach a probability to either path. It also omits tax changes, financing frictions, and any strategic option not already included in the cash receipts. A spreadsheet can correctly calculate the implications of weak assumptions. Correct arithmetic cannot make the assumptions correct. State the claim before selecting a metric. To compare projects, use consistent cash-flow timing, currency, financing treatment, horizon, and residual assumptions. To examine an acquisition, include the cash needed to keep the acquired business functioning. To examine a distribution, include the organization's remaining obligations. To examine a customer investment, identify which benefit is expected to become cash, when, and at what additional cost. Do not count the same expected gain once in operating cash and again in a separate strategic premium. Buffett's 2012 discussion makes purchase price and the uses available for retained earnings central to his payout argument. Leonard's 2014 letter warns that cash remaining after expenses is not automatically a sound measure of economic return if the underlying intangible assets are deteriorating. These are complementary cautions. The guide's arithmetic is deliberately simple so that the difficult part remains visible: what future cash is plausible, what preserves it, and what you give up by committing now. ### Work the question For the fictional project, name two assumptions that could reverse the small positive net present value. Explain why the $30,000 undiscounted surplus is neither annual profit nor a 25 percent annual return. Then identify one cash need that an acquisition headline price could omit. ### Compare with the guide’s reasoning Lower cash receipts, later receipts, additional sustaining investment, a higher appropriate discount rate, or an overstated residual value can reverse the result. Cash and accounting profit differ, and a three-year cumulative ratio is not an annual rate. Acquisition funding may also require working capital, integration expenditure, deferred consideration, or refinancing. Check which are already included before adding them. ### Sources - Warren E. Buffett, Berkshire Hathaway 2012 shareholder letter (2012 reporting year; published 2013), Dividends, including reinvestment, acquisitions, repurchases, and the conditions behind the payout argument.. https://www.berkshirehathaway.com/letters/2012ltr.pdf Evidence note: Written by the allocator. His assessments of acquisition value and alternative outcomes are his judgments, not independent causal measurements. - Mark Leonard, Constellation Software 2014 president’s letter (2014 reporting year; published 2015), Pages 1–5, particularly intangible-asset condition, reliable capital, and business-unit decentralization.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: The URL reflects later file hosting, not a 2026 letter. Business figures in the original refer to their stated historical periods. Author explanations remain author explanations. ## Warren Buffett · Development · 1951-1958 | Analysis before the legend The earliest Buffett artifact located for this course is his 1951 GEICO article, but the exact original scan was not fully extracted. The factual account here therefore rests on Buffett's 1995 shareholder letter, a later retrospective that describes his January 1951 visit to GEICO, Lorimer Davidson's explanation of the direct-sales cost advantage, four stock purchases during 1951 and Buffett's sale in 1952. That later account is evidence of Buffett's own reconstruction, not a substitute for the unread original article. It shows a twenty-year-old student combining Graham's demand for a discount with direct investigation of a company's operating advantage. Berkshire returned to GEICO during the insurer's 1976 crisis and did not acquire the whole company until 1996. The durable pattern was not uninterrupted ownership. It was the ability to recognize the same economic advantage again after circumstances and price changed. The 1957 and 1958 partnership letters show the operating system Buffett built around that analysis. He told partners that he did not forecast the market, but he did change the portfolio mix as opportunities changed. In the 1957 reporting year, a falling market increased the supply of undervalued general issues, so the balance shifted away from announced corporate workouts. The same letter warned that the partnership would probably lag in a strongly rising market. That warning became a live test in the 1958 reporting year when the Dow, including dividends, gained 38.5 percent. In the 1958 reporting year, Buffett's five partnerships gained between 36.7 percent and 46.2 percent. He did not use the favorable result to erase the forecast. He explained why a strategy with dormant, illiquid positions could look slow during a boom. Commonwealth Trust makes the early craft concrete. By the 1958 reporting year, Buffett had bought an illiquid bank around $51 a share while estimating a much higher private-owner value. The position took patience because only a few shares traded. A possible merger offered a route to value, but no fixed timetable. He sold around $80 late in the 1958 reporting year when another opportunity looked better, even though he believed the buyer could still do well. This is not the later slogan of never selling a wonderful business. It is capital moving from one discount to another. The pressure was scarcity, liquidity and the need to buy without moving the price. The response was secrecy, patience and a willingness to leave some upside for the next owner. The development to notice is not a switch from ignorance to wisdom. Buffett began with a coherent discipline and then met the practical limits of small pools of capital. A tiny partnership could buy neglected securities that a future Berkshire could not. The early letters also reveal a habit that persisted. He specified in advance when the approach should look weak, then judged results against that expectation. The question for a learner is whether a principle survives an unfriendly comparison, not whether a successful person can tell a persuasive story after the outcome is known. ### Work the question Write the 1958 partner note before seeing the year's result. State where this strategy should lag, what evidence would count as failure, and whether Commonwealth should be sold below your estimate of full value when a better opportunity appears. ### Compare with the guide’s reasoning A strong answer separates market direction from strategy fit. It expects relative weakness in a sharp bull market, uses a multi-year comparison rather than one year's gain, and treats the sale as an opportunity-cost decision rather than proof that the original estimate was wrong. ### Sources - Warren E. Buffett, Second Annual Letter to Limited Partners (1957 reporting year; circulated 1958), Reproduced original, printed pages 1-3. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. The reproduced original identifies itself as the second annual letter and includes a fragment of the unavailable prior letter. - Warren E. Buffett, 1958 partnership letter (1958 reporting year; circulated 1959), Reproduced original, printed pages 4-6. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Includes the complete Commonwealth Trust account and the stated expectation for performance in a rising market. - Warren E. Buffett, The Security I Like Best (December 6, 1951), Commercial and Financial Chronicle archive; exact article scan located but not completely extracted in this pass. https://archive.org/details/pub_commercial-and-financial-chronicle Evidence note: The original periodical archive was located, but the exact article was not verified by a complete direct read. The course therefore uses no quotation from it and labels the detailed early GEICO account as Buffett's 1995 retrospective. - Warren E. Buffett, 1995 Chairman's Letter (1995 reporting year; published 1996), GEICO section, including Buffett's retrospective account of 1951-1952, the 1976 return and the 1995 agreement to acquire the remaining shares. https://www.berkshirehathaway.com/letters/1995.html Evidence note: Official original. This is later self-reporting about the 1951 episode, not contemporaneous evidence of what the unread 1951 article said in full. - Warren E. Buffett, Buffett Partnership letters compilation (1957-1970), Complete 152-page compilation, especially the 1957 and 1958 letters, January 18, 1963 Ground Rules and Dempster discussion, January 25 and October 9, 1967 letters, January 22 and May 29, 1969 letters, and February 25, 1970 letter. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: A third-party-hosted compilation of reproduced original letters. It is not an official Berkshire archive. Datelines and signatures were checked in the documents. All 29 located documents dated 1957 through 1970 were read in full; the earlier 1956 original referenced in the 1957 letter was not present. ## Warren Buffett · Development · 1961-1963 | Control introduced a different kind of risk The partnership's move into control investments changed the work. A cheap minority holding could be sold when price approached value. Control made Buffett responsible for turning assets into value, choosing managers and accepting consequences for employees and owners. Dempster Mill exposed that difference. The partnership obtained control in August 1961 after buying most shares around $28. Buffett valued the assets below their book figures because the operating business produced little profit. The spreadsheet looked protected. The organization did not. Buffett first worked with the existing managers to reduce inventories, overhead and tied-up capital. His January 1963 account says those efforts went nowhere for about six months. The later course correction was not another valuation calculation. Buffett met Harry Bottle on April 17, 1962, offered compensation tied to the required outcomes, and placed him in the president's role on April 23. Bottle cut the break-even point, sold slow inventory, changed marketing and disposed of facilities. By the 1962 year-end valuation, Dempster had no notes payable, much lower inventory and a portfolio of marketable securities. Buffett's asset discount had created room for a merely adequate sale, but the investment still needed an operator who could make the assets move. This episode reveals a harder edge than the later Berkshire promise to leave successful managers alone. Buffett changed management quickly after concluding that stated agreement was not producing action. He also praised Bottle for taking difficult steps first. The pressure came from a business whose apparent balance-sheet value could be consumed by continued operating losses. Delay was not neutral. Each month could convert inventory and plant from protection into waste. The revision was to treat personnel and incentives as part of valuation rather than as matters that could be solved after the investment case. The January 1963 Ground Rules, read beside Dempster, show Buffett trying to protect the partnership from another kind of human risk. He told partners that monthly payments could be returns of capital, that three years was the minimum useful performance test, and that his family wealth would remain invested alongside theirs. The rules did not make outcomes safe. They aligned expectations before volatility arrived. That is a different form of control. Dempster required authority over a company; the partnership required clarity about what Buffett could and could not promise. A tidy history might say Dempster proved the superiority of hard action. The record is narrower. We have Buffett's own account of the former managers, not their version or an independent operating history. We can document the sequence, balance-sheet change and sale. We cannot infer motives. The teaching point is therefore bounded. Asset value bought time, poor execution consumed it, and Buffett revised both management and incentives when observable results failed to change. ### Work the question You control Dempster in February 1962. What evidence would justify replacing management, and which facts would remain unknown even after the balance sheet improves? ### Compare with the guide’s reasoning Use operating evidence such as inventory conversion, overhead, cash generation and completed actions. Improved assets do not prove employee effects, customer durability, managerial motives or repeatable earning power. ### Sources - Warren E. Buffett, Ground Rules and 1962 partnership review (January 18, 1963), Reproduced original, printed pages 32-41, especially Dempster on pages 37-39. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Contemporary account of Dempster before and after the April 1962 management change. - Warren E. Buffett, 1963 partnership review (January 18, 1964), Reproduced original, printed pages 51-62, especially the Dempster outcome on pages 59-61. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full in the local original compilation, printed pages 51-62. The source records the expanding capital base, performance comparisons and Dempster follow-through. - Warren E. Buffett, Buffett Partnership letters compilation (1957-1970), Complete 152-page compilation, especially the 1957 and 1958 letters, January 18, 1963 Ground Rules and Dempster discussion, January 25 and October 9, 1967 letters, January 22 and May 29, 1969 letters, and February 25, 1970 letter. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: A third-party-hosted compilation of reproduced original letters. It is not an official Berkshire archive. Datelines and signatures were checked in the documents. All 29 located documents dated 1957 through 1970 were read in full; the earlier 1956 original referenced in the 1957 letter was not present. ## Warren Buffett · Development · 1964-1969 | Success made the original method harder to use The pressure was visible before the partnership's tenth anniversary. In the January 20, 1966 review of the 1965 reporting year, Buffett said capital had reached $43.6 million and that greater size was now more likely to hurt results than help them. He stopped admitting ordinary new partners. He also made a fourth investment category explicit, expanded the role of relatively undervalued securities, and formalized a rule permitting as much as 40 percent of net assets in one holding only when both the expected advantage and protection against permanent loss were exceptional. He disclosed that the same year contained his worst single investment experience to date. Method revision and error disclosure arrived inside a record 47.2 percent partnership year, rather than after a collapse. By the partnership's tenth anniversary, the problem was no longer whether Buffett's method worked. It was whether it could keep working at a much larger scale and in a changed market. The January 25, 1967 letter reports that Buffett Associates had begun in May 1956 with $105,100 and that partnership capital stood at about $54.1 million at the start of 1967. Buffett said the first decade's results had no chance of being duplicated. The number of understandable, sufficiently large bargains had fallen, competition had risen, and a few ideas now dominated results. Success had altered the opportunity set that produced it. The October 9, 1967 letter is an unusually clear revision made before failure forced it. Buffett reduced the goal from a ten-percentage-point annual advantage over the Dow to the lesser of a 9 percent annual return or a five-percentage-point advantage. Four causes were separated. Quantitative bargains had become scarce. Professional investors were chasing shorter performance windows. The larger capital base made small opportunities less useful. Buffett also said his own desired level of effort had changed. The admission matters because it mixes market conditions, organizational scale and personal preference without pretending they are the same thing. He was also between methods. Buffett still described himself as primarily quantitative, yet admitted that his most exceptional results had come from rare qualitative judgments. He would not chase fashionable securities he did not understand, but the statistical bargains that had been the partnership's bread and butter were disappearing. That is development under pressure. An old method had not become false. Its supply of usable opportunities had shrunk. The emerging alternative required judgments that were harder to repeat and harder to prove from numbers alone. The 1968 reporting year made persistence harder to interpret. The partnership gained 58.8 percent while the Dow gained 7.7 percent, including dividends. Buffett called the result a freak and repeated that idea quality was at an all-time low. If performance alone had set the decision, the partnership looked stronger than ever. He instead treated the exceptional gain as weak evidence about future capacity. On May 29, 1969, he told partners he intended to wind the partnership down. The decision preserved a promise made earlier. If he could no longer operate within the stated goals and methods, he would say so. The closure was not withdrawal from business. Buffett already controlled Berkshire and other companies, and the liquidation letters devoted substantial work to distributions, tax consequences and possible managers for partners' money. Persistence changed form. He stopped accepting the old mandate while retaining businesses and investments that fit a different mandate. That distinction resists the heroic story of relentless continuation. Sometimes persistence means keeping the standard and ending the vehicle that can no longer meet it. The decision is more demanding than simply lowering the yardstick after a great year and carrying on. ### Work the question Should partners reject the closure because the 1968 result was exceptional? Build the case using capacity, opportunity supply, incentives and the promise made before the result. ### Compare with the guide’s reasoning The 1968 gain does not answer whether the method remained scalable or repeatable. The prior warning, reduced opportunity supply and changed mandate support closure even when trailing performance was excellent. ### Sources - Warren E. Buffett, 1965 partnership review (January 20, 1966), Reproduced original, printed pages 85-94. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Buffett limits new partners, describes scale as an emerging constraint, formalizes concentration limits and records his worst investment experience to date. - Warren E. Buffett, The First Decade partnership review (January 25, 1967), Reproduced original, printed pages 100-107. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Buffett states that the first decade could not be duplicated, identifies the shrinking idea flow, and distinguishes method from market fashion. - Warren E. Buffett, Partnership objective-revision letter (October 9, 1967), Reproduced original, printed pages 111-114. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Buffett explains why scale, scarce bargains and a changed market led him to reduce the partnership objective. - Warren E. Buffett, 1967 partnership review (January 24, 1968), Reproduced original, printed pages 115-119. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. The letter measures the first year after the objective revision and separates headline outperformance from a poor workout result and the lower-return control portfolio. - Warren E. Buffett, 1968 partnership review (January 22, 1969), Reproduced original, printed pages 123-128. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Buffett calls the 58.8 percent partnership result for the 1968 reporting year a freak and repeats that idea quality and quantity were at an all-time low. - Warren E. Buffett, Partnership retirement announcement (May 29, 1969), Reproduced original, printed pages 129-131. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Buffett accepts responsibility for the 1969 workout losses and announces liquidation rather than continue without first-class ideas. ## Warren Buffett · Development · 1965-1976 | Berkshire was not one clean pivot Buffett's late letters call the purchase of Berkshire a major mistake. The 1965-1969 documents are signed by Malcolm G. Chace Jr. and/or Kenneth V. Chace, although the local originals attribute their letter text to Buffett. The contemporary reports show why that judgment took time to become obvious in practice. The 1965 report described profitable operations, lower inventories, repaid bank loans and additional machinery. The 1966 report again discussed operating improvement and declared Berkshire's only cash dividend. In the 1967 reporting period, textile sales fell from $49.4 million in the prior fiscal year to $39 million. Management curtailed production, lost skilled workers and expected retraining costs when demand returned. These are the actions of people trying to improve a difficult business, not yet behaving as though extinction were settled. At the same time, capital was beginning to escape the textile trap. The 1966 report said marketable securities offered a temporary use for surplus funds and a way to participate in earnings outside textiles. Berkshire acquired National Indemnity in 1967. The insurance subsidiaries earned more that year than the textile business despite using less capital. The report called the purchase a first major step toward a diversified earnings base, but it also expected the relative profitability of textiles and insurance to reverse in some years. The belief at the time was diversification, not immediate abandonment. Pressure then arrived from both businesses. Insurance underwriting deteriorated sharply during the 1974 reporting year. Buffett wrote that Berkshire's own management lacked the information and pricing knowledge required for a particular area of auto insurance. Liquidity was raised and volume restrained. The 1975 reporting year then produced the lowest return on opening equity since 1967. Textile employment had fallen as much as 53 percent during the first half of 1975, yet the letter forecast better textile earnings in 1976. The 1976 report called textiles a significant disappointment and still mentioned an acquisition under consideration. These details matter because they show persistence after the broad economic diagnosis was already visible. The later story says Charlie Munger immediately recognized the error and supplied a blueprint for buying wonderful businesses. That is Buffett's retrospective account, and it captures the eventual direction. It does not fully describe the sequence. Capital moved toward insurance, banking, publishing and See's while Berkshire continued trying to repair, modernize and even expand textiles. The course correction was gradual, partly financed by cash released from the original operation and partly constrained by employees, managers and Buffett's reluctance to close a business that could cover its cash needs. This before-and-after evidence changes the lesson. Recognizing a bad industry is not the same as exiting it. A leader can correctly see low returns and still delay because current profits, human obligations and hope make each next investment look defensible. Berkshire's development came from redirecting marginal capital before the old operation disappeared, maintaining liquidity through insurance mistakes, and eventually admitting that incremental textile improvements could not fix the industry's structure. The late diagnosis is useful, but the contemporary record shows the cost and duration of learning it. ### Work the question At the end of 1976, decide whether another textile acquisition is justified. Use only facts available then, and state which later facts you are forbidden to import. ### Compare with the guide’s reasoning A defensible answer weighs current cash needs, industry returns, required equipment, employment and alternative uses of capital. It cannot import the 1985 shutdown, later See's cash generation or Buffett's sixty-year retrospective as though they were known in 1976. ### Sources - Warren E. Buffett; signed by Berkshire officers as noted, Early Berkshire Hathaway annual reports (1965-1976 reporting years), Independent archive of hosted original-report facsimiles. Exact report locators used here: 1965 report, printed pages 1-2, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1965-Berkshire-AR.pdf; 1966 report, printed pages 1, 3-6, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1966-Berkshire-AR.pdf; 1967 report, printed pages 1, 3-4, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1967-Berkshire-AR.pdf; 1975 report, printed pages 1 and 7-8, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1975-Berkshire-AR.pdf; 1976 report, printed pages 1-2, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1976-Berkshire-AR.pdf. Berkshire’s official individual-letter archive begins at 1977.. https://theoraclesclassroom.com/archives/ Evidence note: These are independently hosted reproductions of the original reports, checked against the local original PDFs. The 1965-1969 reports are signed by Malcolm G. Chace Jr. and/or Kenneth V. Chace, while the local originals state that Buffett wrote the letter. The 1970-1976 reports are signed by Buffett. Berkshire’s official page supplies individual letters only from 1977 and a broader 1965-2024 compilation. - Warren E. Buffett, Berkshire Hathaway 1974 annual letter (1974 reporting year; signed March 31, 1975), Insurance underwriting, Home and Auto, liquidity and Blue Chip Stamps. https://www.berkshirehathaway.com/letters/letters.html Evidence note: Contemporary report during an insurance downturn. The full text was available locally; the public URL is Berkshire's archive page because it does not expose an individual 1974 file. - Warren E. Buffett, 2022 Chairman's Letter (2022 reporting year; signed February 25, 2023), A dozen good decisions, luck, Coca-Cola and American Express, and Berkshire's 1967 insurance pivot. https://www.berkshirehathaway.com/letters/2022ltr.pdf Evidence note: A late retrospective that deliberately compresses a long record. The course tests it against contemporary documents rather than accepting the compression as the whole story. - Warren E. Buffett, 2024 Chairman's Letter (2024 reporting year; signed February 22, 2025), Mistakes, delayed correction, GEICO, cash, Berkshire's sixty-year record and succession. https://www.berkshirehathaway.com/letters/2024ltr.pdf Evidence note: Buffett's final Berkshire annual letter. It was reviewed from the complete local original. - Warren E. Buffett, Controlled-company distribution letter (December 5, 1969), Reproduced original, printed pages 137-140. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Buffett reports the weak textile economics, the redeployment into insurance and banking, and the planned distribution of Berkshire and Diversified Retailing shares. - Warren E. Buffett, Partnership liquidation questions and answers (December 26, 1969), Reproduced original, printed pages 141-144. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Buffett explains why Berkshire textiles remained open despite poor returns and details the constraints on distributing illiquid control shares. ## Warren Buffett · Development · 1972-1985 | Quality changed the opportunity cost of patience See's Candies did more than add a famous example to Buffett's vocabulary. It changed the comparison applied to every dollar retained in a weak business. Berkshire's 1975 report already called the 1972 acquisition a winner. The 1977 report said See's pretax earnings had risen from $4.2 million in 1972 to $12.6 million in 1977 with little additional capital. The operation did not need large sums merely to stand still. Cash could leave the candy business and be used elsewhere. That experience made the textile operation's demands harder to excuse. The surviving December 1972 letter to See's manager Charles Huggins also qualifies a simple hands-off story. In the readable portion, Buffett used observations from a Brandeis store visit to propose controlled tests of display, product presentation and territorial availability, while explicitly deferring to Huggins's deeper product knowledge. The record supports selective operating input near the acquisition, not a claim that Buffett personally managed the business or that the proposals were implemented. Buffett's 1983 goodwill appendix supplied the revised explanation. Accounting book value recorded capital put into a business. Economic goodwill represented earning power that could grow without equivalent additions of tangible capital. Buffett admitted that his Graham-trained preference for tangible assets had caused important errors of omission. The revision was not that price stopped mattering. It was that a high return on a small tangible base could justify paying above book value, while a business sold below book could remain expensive if it repeatedly consumed capital. The 1985 shutdown account shows why insight did not produce immediate action. Buffett listed employees, candid management and cooperative labor as reasons for continuing textiles while the business could avoid large cash drains. He also acknowledged that he had expected conditions to improve and had been wrong. Managers changed product lines, machinery and distribution. Berkshire bought Waumbec Mills in 1975. Proposals for equipment often appeared attractive in isolation because each machine lowered a particular cost. Collectively, however, competitors made similar investments, prices fell, and the owners received poor returns on a larger capital base. The pressure therefore came from local rationality. One machine could pass a return calculation. One more year could protect jobs. One acquisition could add scale. Each decision looked bounded, but together they prolonged exposure to a commodity industry with excess capacity. Buffett closed the operation in July 1985 only after concluding that continued support could become an endless claim on capital. The later letter does not blame the operating managers. It places the allocation error with Berkshire. That allocation frame preserved a distinction between effort and economics. Capable people could execute well inside a structure that still denied owners an adequate return. The tension should not be resolved too neatly. See's did not prove that every branded consumer company was wonderful or that every capital-intensive company was bad. Buffett later invested heavily in utilities and a railroad. The deeper update concerned the price of required reinvestment, the durability of customer preference and where incremental capital could earn its best risk-adjusted return. Persistence remained valuable when it compounded an advantage. It became destructive when it financed survival without improving industry position. The hard part was that the difference appeared over years, not in one decisive spreadsheet. ### Work the question A proposed textile machine meets its standalone return hurdle in 1982. What additional evidence is required before Berkshire should fund it? ### Compare with the guide’s reasoning Test whether competitors can copy the saving, whether prices will absorb the benefit, how much later reinvestment is required, and what the same capital could earn in a business with stronger customer preference. ### Sources - Warren E. Buffett, 1983 Chairman's Letter (1983 reporting period; published 1984), Goodwill and its Amortization, especially the See's comparison. https://www.berkshirehathaway.com/letters/1983.html Evidence note: Contemporary discussion of accounting and economic goodwill using historical See's figures. The acquisition rationale is partly retrospective. - Warren E. Buffett, 1985 Chairman's Letter (March 4, 1986), Shutdown of Textile Business. https://www.berkshirehathaway.com/letters/1985.html Evidence note: Contemporary account written after the July 1985 closure decision. It reports Buffett's own diagnosis and hindsight about earlier delay. - Warren E. Buffett, 2007 Chairman's Letter (2007 reporting period; published 2008), Businesses, The Great, the Good and the Gruesome. https://www.berkshirehathaway.com/letters/2007ltr.pdf Evidence note: Retrospective comparison of See's, FlightSafety, capital-intensive growth, and Dexter. The Dexter share value is measured at the letter's February 2008 publication, not in the 2007 reporting year or in 1993. - Warren E. Buffett, 2014 Chairman's Letter (2014 reporting period; published 2015), Berkshire, Past, Present and Future, especially the textile, See's and Dexter passages. https://www.berkshirehathaway.com/letters/2014ltr.pdf Evidence note: Fifty-year retrospective. It supplies later cumulative figures and Buffett's revised interpretation, not a contemporaneous record of the earlier decisions. The Dexter share value is measured at the letter's February 2015 publication. - Warren E. Buffett, Letter to Charles N. Huggins at See’s Candy Shops (December 13, 1972), Independent hosted scan, numbered points 1-4 and continuation headed Page #2. Partial OCR read through the Buffett signature; the full five-page scan was not visually accessible.. https://theoraclesclassroom.com/wp-content/uploads/2019/10/1972-Buffett-Letter-to-Sees-Candies.pdf Evidence note: Read in part from OCR with visible errors. No quotation or numerical claim is drawn from it. The readable portion documents Buffett proposing bounded merchandising tests while crediting Huggins’s greater product knowledge. ## Warren Buffett · Development · 1974-1990 | Risk had to be rebuilt from consequences Insurance forced Buffett to replace simple labels with consequence-based risk. In the 1974 reporting year, Berkshire's insurance result deteriorated far more than expected. Buffett did not attribute the loss only to a bad industry cycle. In one auto-insurance operation, he said management lacked adequate underwriting information and pricing knowledge. The response was to restrict volume, build liquidity and wait for rates that paid for the risk. Growth stopped being evidence of success when the cost of the product could not be known until claims arrived. A partially readable July 1976 letter about GEICO shows the same discipline under more personal pressure. Buffett separately identified his sentimental connection, GEICO's enduring low-cost position and his own judgment that survival was plausible, while leaving underwriting decisions to National Indemnity's manager. He also favored contracting the book over pursuing growth. The surviving OCR is incomplete and corrupts figures, so it supports this decision structure rather than quotations, quantities or proof that the commitment was attractive. The inflation problem added another layer. In his May 1977 Fortune article, Buffett argued that aggregate corporate returns on equity behaved like a sticky coupon. Inflation required more dollars in receivables, inventory and plant merely to maintain the same physical output. Retained earnings could therefore produce nominal growth without real growth. Berkshire's own textile experience supplied an operating example. A large tangible base needed replacement and modernization while competition constrained prices. The article's argument and the See's experience pushed in the same direction. The attractive asset was not simply one protected from inflation in name. It was a business able to raise prices or grow without continually rebuilding its capital base. Buffett's 1984 Columbia essay defended the Graham and Dodd tradition against the claim that sustained outperformance was random. Its strongest evidence was not that value investors shared identical portfolios. They differed. Buffett argued that they shared a decision rule centered on price versus business value and had been identified before the records were assembled. The essay also defined risk as permanent economic loss rather than price fluctuation. That fit the partnership's earlier emphasis, but Berkshire had added operational evidence. An insurance book could look calm until delayed claims revealed underpricing. A bond could pay exactly as promised and still lose purchasing power. A cheap textile asset could consume more capital. By the 1989 and 1990 letters, Buffett was cataloguing failures caused by the institutional imperative and leverage. Managers copied peers, expanded because others expanded, and used debt that made ordinary business trouble fatal. Berkshire's own insurance mistakes remained visible for years because reserve errors developed slowly. Buffett wrote in the 1990 report that old errors kept sending bills. This is a more severe account of persistence. Some decisions cannot be exited when the mistake is recognized because the contracts already exist. The revision across these years was from classifying an asset as safe to tracing how it could fail. The relevant questions became concrete. What future cash must be paid? Who can demand it, and when? Can inflation change the real result? Does leverage remove the time needed for value to emerge? Can growth hide underpricing? That approach did not eliminate error. It made error legible enough to change underwriting, liquidity and capital allocation before the next cycle. ### Work the question Compare a long-term bond, a rapidly growing insurer and See's in an inflationary period. Rank them by permanent-loss risk without using price volatility as the deciding metric. ### Compare with the guide’s reasoning Trace purchasing power, unknown claim costs, pricing power, capital needs and liquidity. The ranking can change with price and contract terms; the asset-class label is insufficient. ### Sources - Warren E. Buffett, Berkshire Hathaway 1974 annual letter (1974 reporting year; signed March 31, 1975), Insurance underwriting, Home and Auto, liquidity and Blue Chip Stamps. https://www.berkshirehathaway.com/letters/letters.html Evidence note: Contemporary report during an insurance downturn. The full text was available locally; the public URL is Berkshire's archive page because it does not expose an individual 1974 file. - Warren E. Buffett, How Inflation Swindles the Equity Investor (May 1, 1977), The sticky equity coupon, retained earnings and capital required merely to maintain physical output. https://fortune.com/article/buffett-how-inflation-swindles-the-equity-investor-fortune-classics-1977/ Evidence note: Original publisher page. The article develops an argument that was contemporaneous with Berkshire's capital-intensive textile experience. - Warren E. Buffett, The Superinvestors of Graham-and-Doddsville (May 17, 1984), Accessible article text and performance tables. https://business.columbia.edu/insights/chazen-global-insights/superinvestors-graham-and-doddsville Evidence note: Columbia Business School's publication of Buffett's argument against a pure-chance explanation for the records of value investors he had identified in advance. - Warren E. Buffett, 1989 Chairman's Letter (1989 reporting year; signed March 2, 1990), Mistakes of the First Twenty-five Years and the institutional imperative. https://www.berkshirehathaway.com/letters/1989.html Evidence note: Retrospective classification of commission, omission and institutional errors after twenty-five years of Berkshire control. - Warren E. Buffett, Letter to George D. Young about GEICO (July 22, 1976), Independent hosted scan, numbered reasons on pages 1-3 and allocation of survival and underwriting responsibility at the end of page 3. Partial OCR ends before the fourth scan page.. https://theoraclesclassroom.com/wp-content/uploads/2020/10/Buffett-Letter-about-GEICO-1976.pdf Evidence note: Read in part from visibly imperfect OCR. No quotation or numerical claim is drawn from it. The readable portion separates sentiment, GEICO’s cost position, survival judgment and underwriting responsibility. ## Warren Buffett · Development · 1991 | Salomon put reputation under a clock Salomon Brothers converted Buffett from an investor commenting on institutional behavior into the interim chairman responsible for a firm under investigation. Berkshire had invested in Salomon preferred stock in 1987. In 1991, revelations about unauthorized bids in United States Treasury auctions threatened the firm's access to government-securities markets and therefore its survival. Buffett took the interim chair on August 18, 1991. The decision concentrated legal, regulatory, liquidity and reputational pressure in a business whose daily operations depended on trust. His congressional testimony did not present a clever financial rescue. It offered cooperation, accountability and a behavioral boundary. Buffett apologized for the firm's conduct, committed to supplying facts to authorities and said employees should worry about whether an action would survive public scrutiny, not merely whether a lawyer could defend it. That standard was broader than compliance because Salomon's problem had included failures of escalation and culture. The practical revision was to make reputation an operating constraint that could stop profitable activity before a rulebook did. The 1991 Berkshire letter shows the cost of this role. Buffett said his job at Salomon was intense and personal, while Berkshire's decentralized businesses continued because their managers did not require his daily direction. That experience tested a structure Berkshire had been building for years. Delegated operating authority created capacity for the parent to absorb an emergency elsewhere. The same structure also depended on selecting managers whose conduct would not create another emergency. Autonomy and trust were therefore complements, not substitutes for oversight. Salomon also resists a pure hero narrative. Berkshire had chosen the exposure and had praised the instrument. Buffett did not discover the misconduct before it became public. His later intervention was consequential, but it followed a governance failure inside a company in which Berkshire was a major investor. The episode therefore teaches both response and prior blindness. A strong reputation can buy a hearing from regulators, but it does not prove that the organization saw the problem early. The record also does not establish that every disputed fact was resolved by Buffett's testimony. His statements document the commitments he made under oath and the standard he imposed after taking charge. Buffett carried the lesson into later acquisition criteria and managerial letters. The desired manager needed ability, energy and integrity because the first two without the third could magnify damage. Yet the 2025 Thanksgiving message admits that Buffett and Munger later failed several times to act when a previously excellent chief executive suffered cognitive decline. The Salomon standard was clear under public crisis. Applying an uncomfortable personnel standard early, privately and to a trusted colleague remained harder. Development did not end when the principle was articulated. That contrast turns reputation from a crisis statement into a recurring governance problem. ### Work the question Separate the Salomon evidence into what Buffett repaired, what Berkshire failed to detect, and what the episode cannot prove about decentralized management. ### Compare with the guide’s reasoning He restored cooperation and set a public conduct standard after the scandal. Berkshire did not prevent or discover the violations. One successful emergency response does not prove that trust-based decentralization reliably detects misconduct early. ### Sources - Warren E. Buffett, 1991 Chairman's Letter (1991 reporting year; signed February 28, 1992), Salomon, omission errors, USAir and fixed-income securities. https://www.berkshirehathaway.com/letters/1991.html Evidence note: Buffett's own report after taking the interim chair at Salomon. It does not replace the hearing record or independent reporting on the scandal. - Warren E. Buffett, Salomon Brothers securities-trading investigation testimony (September 4-5, 1991), Opening apology, cooperation commitments, compliance standard and response to lawmakers. https://www.youtube.com/watch?v=MtaeGt3KwuA Evidence note: Public recording of Buffett's testimony before a House subcommittee. The hearing was reviewed in part; the course does not claim a complete transcript read. - Warren E. Buffett, Thanksgiving Message to Fellow Shareholders (November 10, 2025), Complete message, including succession, philanthropy, delayed personnel action and final reflections. https://berkshirehathaway.com/news/nov1025.pdf Evidence note: Latest Buffett-authored material located on Berkshire's official site as of September 9, 2026. Read in full. ## Warren Buffett · Development · 1993-2002 | Price and payment were separate bets Dexter Shoe and General Re show that an acquisition contains more than one forecast. Berkshire must judge the business, the managers, the price, the financing instrument and its own ability to respond if the thesis fails. In the 1993 letter, Buffett described Dexter favorably. It had substantial production, retail outlets, supplier recognition and a strong position in golf shoes. Berkshire paid with its own shares. The case at the time assumed that Dexter could continue resisting lower-cost foreign competition. By the 2001 reporting year, Dexter was losing money. Buffett separated three mistakes. Berkshire bought the company, paid with stock and delayed operational changes after the need became apparent. The distinction matters. A cash purchase of a failing business would have lost the purchase price. A stock purchase transferred part of every future Berkshire winner to the sellers. As Berkshire grew, the opportunity cost of that currency kept expanding. The delayed response then added an operating cost after the strategic error was becoming visible. General Re began with the same optimism at far greater scale. The 1998 letter called the company first-class and explained that issuing Berkshire shares increased book value without necessarily increasing intrinsic value. Buffett believed the exchange was roughly fair. After the September 11, 2001 attacks, he acknowledged that General Re had accumulated terrorism and aggregation exposure without adequate safeguards. He said he had recognized the general danger but failed to turn the thought into action. The 2002 letter then recorded another $1.31 billion of reserve corrections for earlier years and said weaknesses should have been detected before the merger. The acquisition eventually produced a strong operation, which makes hindsight harder. A recovered business can still have been purchased on bad terms. In the 2016 letter, Buffett said issuing 272,200 Berkshire shares for General Re increased the share count by 21.8 percent and caused owners to give more than they received. He contrasted that with BNSF, where some stock issuance made sense. The update was not a permanent ban on stock. It was a stricter test of the value surrendered and a preference for internally generated cash when the buyer's own shares were undervalued. These cases also revise the popular account that Buffett simply learned to buy better businesses after See's. Dexter was presented as a quality manufacturer in 1993. General Re had decades of respected underwriting. The failures came from industry change, hidden or underestimated liabilities, overconfidence in normalized economics, and payment terms. Quality was a hypothesis that required monitoring. The later letters are valuable because Buffett preserved the favorable contemporary case and then named distinct errors instead of rewriting the original decision as obviously foolish. The learner's task is to do the same before the outcome makes every signal look clear. ### Work the question Build a five-part acquisition decision for General Re in 1998. Which parts failed later, and which could still be judged successful? ### Compare with the guide’s reasoning Separate business quality, management and controls, price, Berkshire-share currency, and post-close correction. General Re later became valuable, while reserving, aggregation control and the amount of Berkshire equity surrendered were still serious errors. ### Sources - Warren E. Buffett, 1993 Chairman's Letter (1993 reporting period; published 1994), Dexter Shoe. https://www.berkshirehathaway.com/letters/1993.html Evidence note: Contemporary, favorable account after the November 7, 1993 merger. It is useful evidence of the case made then, not proof that all deliberations were disclosed. - Warren E. Buffett, 2001 Chairman's Letter (2001 reporting period; published 2002), Non-insurance operating highlights, shoe operations and Dexter. https://www.berkshirehathaway.com/2001ar/2001letter.html Evidence note: Retrospective self-assessment identifying the purchase, use of stock, and delayed operational response as three separate errors. - Warren E. Buffett, 1998 Chairman's Letter (1998 reporting year; signed March 1, 1999), General Re merger, share issuance and intrinsic value versus book value. https://www.berkshirehathaway.com/letters/1998pdf.pdf Evidence note: Contemporary favorable account of the General Re acquisition. Later letters identify underwriting, reserving and acquisition-currency errors that were not visible here. - Warren E. Buffett, 2002 Chairman's Letter (2002 reporting year; signed February 21, 2003), General Re reserving, aggregation risk and derivatives. https://www.berkshirehathaway.com/letters/2002pdf.pdf Evidence note: Follow-up after the 2001 letter. It documents another year of reserve corrections and Buffett's account of what he failed to detect before the General Re merger. - Warren E. Buffett, 2016 Chairman's Letter (2016 reporting year; signed February 25, 2017), Dexter and General Re acquisition currency, BNSF share issuance and future returns. https://www.berkshirehathaway.com/letters/2016ltr.pdf Evidence note: Later comparison of three stock-funded acquisitions, showing why the instrument used to pay can be a separate decision from the asset purchased. ## Warren Buffett · Development · 1999-2003 | The circle of competence carried an omission cost During the technology boom, Buffett's restraint looked outdated. Berkshire's per-share market value fell 19.9 percent in the 1999 reporting year while the S&P 500, including dividends, gained 21.0 percent. Buffett's Fortune article that year did not predict the exact date of a crash. It compared market value, interest rates and corporate profits across two seventeen-year periods and argued that late-1990s expectations had outrun plausible economic growth. The position was analytical, but living through it required accepting a visible relative loss and public doubt. The 2000 letter described speculation as a transfer among participants rather than value created by a business. Buffett also said Berkshire was not equipped to select the few technology companies that would become lasting winners. This boundary prevented many losses after the bubble broke. It also created real omission costs. Berkshire did not own the extraordinary early growth of companies such as Microsoft, Google or Amazon. A circle of competence protects only if the manager admits what the boundary excludes. It cannot be praised by counting avoided failures while ignoring winners that were also declined. The General Re derivatives operation then complicated Buffett's critique of financial complexity. Berkshire inherited a large book of contracts in the 1998 merger. In the 2002 letter, Buffett described faulty earlier accounting, collateral spirals and counterparty chains. Exiting took years because a derivative contract cannot always be closed at a clean quoted price. Berkshire had entered the exposure through an acquisition it believed it understood. The lesson was not merely to avoid exotic instruments. It was to examine the obligations embedded inside an apparently familiar company and the time required to unwind them. These years show two kinds of persistence. Buffett held the competence boundary through a boom even when comparison made him look obsolete. He also persisted through the slow cleanup of contracts and reserves that Berkshire already owned. The first protected choice. The second was an imposed obligation. Conflating them turns patience into a universal virtue. Chosen patience is valuable when the thesis remains intact and financing permits time. Forced patience can be the consequence of inadequate diligence or contracts that make exit expensive. Buffett later invested heavily in Apple. That did not require declaring the earlier boundary foolish. Apple by then could be evaluated through a large installed customer base, recurring behavior and capital returns in a way that differed from forecasting a young technology market. The update was in what evidence made a business understandable. The category label changed less than the operating facts. A useful circle of competence therefore has a permeable edge. It excludes decisions that cannot be supported now while allowing new evidence, people and business models to change what can be supported later. That edge must be explained with evidence each time it moves. ### Work the question Defend the 1999 technology omission without pretending it was costless. Then state what evidence would justify expanding the competence boundary later. ### Compare with the guide’s reasoning The defense rests on unsupported long-range economics and extreme price expectations, while the cost includes missed winners. Expansion requires evidence about customer behavior, durable economics, management and price that can be evaluated rather than a change in fashion. ### Sources - Warren E. Buffett; edited by Carol Loomis, Mr. Buffett on the Stock Market (November 22, 1999), Interest rates, corporate profits and the two seventeen-year market periods. https://fortune.com/article/warren-buffett-on-stock-market/ Evidence note: Original publisher page. It records Buffett's late-1999 valuation argument and should not be rewritten as a precise forecast of the market's turning date. - Warren E. Buffett, 2000 Chairman's Letter (2000 reporting year; signed February 28, 2001), Speculation, technology businesses and limits of competence. https://www.berkshirehathaway.com/2000ar/2000letter.html Evidence note: Contemporary explanation of why Berkshire avoided most technology investments. It also admits earlier failures to understand several supposedly familiar industries. - Warren E. Buffett, 2002 Chairman's Letter (2002 reporting year; signed February 21, 2003), General Re reserving, aggregation risk and derivatives. https://www.berkshirehathaway.com/letters/2002pdf.pdf Evidence note: Follow-up after the 2001 letter. It documents another year of reserve corrections and Buffett's account of what he failed to detect before the General Re merger. - Warren E. Buffett, 2020 Chairman's Letter (2020 reporting year; signed February 27, 2021), Precision Castparts $11 billion write-down and Buffett responsibility statement, printed pages 3-4. https://www.berkshirehathaway.com/letters/2020ltr.pdf Evidence note: Buffett attributes the Precision Castparts overpayment to his own optimism rather than deception by the seller. ## Warren Buffett · Development · 2008-2010 | Liquidity created the right to act The financial crisis tested whether decades of warnings about leverage and liquidity would produce action under real fear. Berkshire entered the 2008 reporting year with substantial cash and businesses that generated more. During the crisis it invested in Goldman Sachs and General Electric on terms unavailable in normal markets, completed the Wrigley financing commitment and bought other securities. Buffett's October 2008 opinion article disclosed that he was moving personal money from government bonds into United States equities. He did not say business conditions were about to improve. He said fear had reduced prices while long-term productive capacity remained. The 2008 annual letter was written after Berkshire's per-share book value fell 9.6 percent during the year, its worst result to that point. Buffett supported forceful government intervention because a financial-system collapse would damage productive businesses regardless of who caused the crisis. At the same time, he admitted investment errors, including buying ConocoPhillips near the peak in oil and gas prices. The contrast matters. General confidence in the country's long future did not make each security or entry price correct. A sound macro posture still contained poor individual judgments. By the end of the 2009 reporting year, Berkshire's cash equivalents had fallen to $30.6 billion from $44.3 billion at the start of 2008, even after retaining $17 billion of operating earnings during the two-year period. Eight billion dollars of the remaining cash was earmarked for BNSF. The company had used the panic rather than merely commenting on it. Yet the BNSF acquisition also required Berkshire stock. The November 2009 agreement committed Berkshire to stock consideration, and the acquisition closed in February 2010. At closing, Berkshire issued about 95,000 shares, equal to 6.1 percent of the shares previously outstanding. Buffett believed Berkshire's market price undervalued those shares, so the cost to continuing owners was higher than the quoted amount delivered to sellers. This was a deliberate exception to his strong preference against issuing stock. BNSF provided an essential service, could absorb very large amounts of capital and had long-duration demand. It also required investment far above depreciation for decades. The purchase therefore revised another oversimplified Buffett rule. Capital-intensive businesses could be attractive when they had durable demand, sensible regulation and an ability to earn on the incremental capital. The textile comparison was not capital-light versus capital-heavy. It was protected reinvestment versus reinvestment that competitors could neutralize. The crisis lesson is usually summarized as being greedy when others are fearful. The record supports a more operational reading. Berkshire could act because it had accepted lower returns on cash before the crisis, limited parent debt, controlled insurance aggregation and maintained a stream of operating earnings. Fear supplied prices. Prior balance-sheet choices supplied agency. The cost was visible in foregone yield before 2008 and in dilution for BNSF. Persistence under pressure was therefore designed years earlier. Without that design, conviction in October 2008 would have been rhetoric rather than deployable capital. ### Work the question Explain why confidence was insufficient in October 2008. Identify the balance-sheet and organizational conditions that turned it into action, then price the BNSF dilution as a separate cost. ### Compare with the guide’s reasoning Cash, operating earnings, limited parent obligations and controlled insurance exposure created capacity. BNSF still required judging the value of Berkshire shares surrendered, even if the railroad itself was attractive. ### Sources - Warren E. Buffett, 2008 Chairman's Letter (2008 reporting year; signed February 27, 2009), Financial panic, government action, liquidity and investments made during disarray. https://www.berkshirehathaway.com/letters/2008ltr.pdf Evidence note: Contemporary report after Berkshire's worst annual decline in per-share book value to that date. - Warren E. Buffett, Buy American. I Am. (October 17, 2008), Personal equity-allocation disclosure and the distinction between forecasting business conditions and buying productive assets. https://www.nytimes.com/2008/10/17/opinion/17buffett.html Evidence note: Original publisher page. Access restrictions prevented a fresh complete page extraction, so the work is not marked read in full. - Warren E. Buffett, 2009 Chairman's Letter (2009 reporting year; signed February 26, 2010), BNSF acquisition, stock consideration and liquidity, printed pages 15-17; signed February 26, 2010. https://www.berkshirehathaway.com/letters/2009ltr.pdf Evidence note: Contemporary account of the cash Berkshire deployed in 2008-2009 and the tradeoff created by using Berkshire shares for part of BNSF. ## Warren Buffett · Development · 2014-2020 | Scale did not remove delayed exits or overpayment The 2014 Tesco exit is an immediate warning against treating marketable securities as automatically self-correcting. At the end of 2012, Berkshire held 415 million Tesco shares at a cost of $2.3 billion. Buffett sold 114 million shares during 2013 after souring on management, but later said the pace was too slow as market share, margins and accounting problems worsened during 2014. Berkshire finished the exit with a $444 million after-tax loss for 2014. The position was easier to sell than a controlled business, yet Buffett still delayed after his judgment changed. Liquidity created an exit option; it did not force timely action. By 2015, Berkshire's opportunity problem had reversed from the early partnership. It needed investments large enough to matter to hundreds of billions of dollars of capital. Precision Castparts appeared to fit. The 2015 letter called it the world's premier aerospace-components supplier and said the acquisition for more than $32 billion in cash fit Berkshire's model. Mark Donegan was praised, demand appeared durable and the deal could absorb a meaningful amount of capital without issuing Berkshire stock. The 2020 letter recorded an $11 billion write-down, almost entirely tied to the price Buffett had paid in 2016. He said the seller had not misled him. He had been too optimistic about normalized profit. He still believed the company could earn good returns on tangible assets, but the expected level of earnings had been wrong, so the price was wrong. This distinction prevents an impairment from becoming a claim that the entire business was worthless. It also makes the failure more useful. A strong company can be a poor acquisition at an optimistic price. The period also exposed pressure in a different form. The pandemic damaged aerospace demand and made the overestimate impossible to ignore, but it did not create the original valuation error. The error occurred when normalized earnings were estimated in 2015-2016. The later shock revealed how little margin existed around that estimate. Buffett's retrospective places responsibility on his forecast rather than on an unforeseeable event. That is a better model of accountability, but it does not tell us which assumptions in the acquisition model were challenged before signing. The unavailable deal file remains an evidence gap. Meanwhile, Berkshire's partial ownership of Apple demonstrated the opposite path. Berkshire could benefit from Apple's retained earnings and repurchases without controlling operations. The 2020 letter showed that Apple's share count reduction increased Berkshire's percentage ownership even after Berkshire sold some shares. Berkshire also repurchased its own shares when Buffett judged the price below intrinsic value. The old preference for complete ownership had softened. A minority position in an exceptional business could be more valuable and easier to resize than complete ownership of a marginal operation. Scale therefore produced two revisions at once. It pushed Berkshire toward whole companies large enough to absorb cash, raising the risk of paying for an optimistic normalization. It also made large marketable securities useful again because public companies such as Apple could absorb billions without requiring Berkshire to operate them. The lesson is not that Buffett finally learned technology. It is that organizational size changed the available set, while price discipline remained necessary in every form. Persistence after Precision Castparts meant retaining a still-useful business, admitting the overpayment and avoiding a false claim that future recovery would erase the acquisition mistake. ### Work the question You are reviewing Precision Castparts in 2015. Design a downside case that separates business quality from purchase-price risk and specifies what a cyclical shock would reveal. ### Compare with the guide’s reasoning Stress normalized margins, aerospace volumes, customer concentration and recovery time. A good tangible return at a lower earnings base can coexist with a poor purchase price, so the acquisition needs a margin around normalization. ### Sources - Warren E. Buffett, 2014 Chairman's Letter (2014 reporting year; signed February 28, 2015), Tesco exit and delayed response, printed pages 17-18; Buffett retrospective, printed pages 24-37. https://www.berkshirehathaway.com/letters/2014ltr.pdf Evidence note: The Tesco discussion is a contemporary admission about the 2012-2014 holding and exit. The fifty-year review later in the same document is retrospective. - Warren E. Buffett, 2015 Chairman's Letter (2015 reporting year; signed February 27, 2016), Precision Castparts acquisition, printed pages 4-5; BHE regulatory assumptions, printed pages 12-13. https://www.berkshirehathaway.com/letters/2015ltr.pdf Evidence note: Contemporary enthusiastic account of Precision Castparts before the later impairment. - Warren E. Buffett, 2020 Chairman's Letter (2020 reporting year; signed February 27, 2021), Precision Castparts $11 billion write-down and Buffett responsibility statement, printed pages 3-4. https://www.berkshirehathaway.com/letters/2020ltr.pdf Evidence note: Buffett attributes the Precision Castparts overpayment to his own optimism rather than deception by the seller. ## Warren Buffett · Development · 2021-2025 | Late candor, Munger's absence and succession Buffett's last letters increasingly compress the record into a few decisive choices. The 2022 letter says Berkshire's satisfactory result came from about a dozen truly good decisions over Buffett's career, roughly one every five years, combined with luck and long holding periods. That is a useful antidote to constant activity. It is also a retrospective selection. The earlier letters contain many more attempted repairs, insurance errors, abandoned positions and businesses that demanded work. A dozen winners explain much of the outcome, but they do not describe the full process that produced and financed them. Charlie Munger's death in November 2023 changed the voice of the 2023 letter. Buffett credited Munger with identifying the Berkshire textile purchase as foolish and with supplying the design for a better company. The praise is supported by decades of letters that show Munger influencing the move toward businesses with stronger economics. Yet the contemporary 1965-1976 reports show that Buffett continued investing managerial attention and some capital in textiles. The clean architect story is Buffett's later interpretation of Munger's influence, not a contemporaneous record of an immediate conversion. Holding both records makes the partnership more instructive. Munger supplied a direction; implementation and conviction accumulated slowly. The 2023 letter also admitted a recent error. In the 2015 letter, Buffett had argued that Berkshire Hathaway Energy's diverse earnings streams protected it from serious harm by any one regulator, while acknowledging that its investment program required trust in future regulation. By 2023, adverse regulation and wildfire liabilities had made earnings and asset values hard to project in some states. Buffett said he had failed even to consider those developments and called that omission costly. The reversal matters because utilities had been presented as capital-intensive businesses whose regulated returns made reinvestment attractive. A condition supporting the thesis changed or had been underweighted. Development here meant revising a category that had once solved the textile problem. No business type remained exempt from renewed analysis. The 2024 letter, Buffett's final annual letter, made delayed correction the cardinal sin. It distinguished errors in business economics from errors about managerial ability or fidelity and noted that controlled companies were harder to sell than marketable securities. The same letter acknowledged that Berkshire's size could make both entry and exit take a year or more. Candor had become part of the succession system. Greg Abel was expected to report failures because concealment would corrupt internal judgment as well as shareholder communication. At the May 3, 2025 annual meeting, Buffett announced that Abel should become chief executive at year-end. In the November 10 Thanksgiving message, Buffett said he would stop writing the annual report and would continue with annual Thanksgiving messages. He also admitted that he and Munger had failed several times to act when a once-excellent chief executive developed cognitive impairment. Greg Abel's first annual letter, published February 28, 2026 for the 2025 reporting year, belongs to Abel. Treating it as another Buffett letter would erase the transition the archive itself now marks. The latest Buffett-authored material therefore ends without a claim of finished mastery. He revised the pace of his philanthropy because his children were aging, retained some voting shares while owners developed confidence in Abel, and described past personnel delays that violated his own stated standard. Persistence became institutional. The work was to transfer principles, authority and reporting duties while accepting that successors would face facts he could not forecast. ### Work the question Test Buffett's late claim that a dozen decisions explain the outcome. What does the claim illuminate, and what development work does it hide? ### Compare with the guide’s reasoning It shows the asymmetric value of a few long-held winners. It hides the financing, failed repairs, risk controls, manager changes, omissions and succession work that allowed those winners to matter and prevented errors from ending the record. ### Sources - Warren E. Buffett, 2015 Chairman's Letter (2015 reporting year; signed February 27, 2016), Precision Castparts acquisition, printed pages 4-5; BHE regulatory assumptions, printed pages 12-13. https://www.berkshirehathaway.com/letters/2015ltr.pdf Evidence note: Contemporary enthusiastic account of Precision Castparts before the later impairment. - Warren E. Buffett, 2022 Chairman's Letter (2022 reporting year; signed February 25, 2023), A dozen good decisions, luck, Coca-Cola and American Express, and Berkshire's 1967 insurance pivot. https://www.berkshirehathaway.com/letters/2022ltr.pdf Evidence note: A late retrospective that deliberately compresses a long record. The course tests it against contemporary documents rather than accepting the compression as the whole story. - Warren E. Buffett, 2023 Chairman's Letter (2023 reporting year; signed February 24, 2024), BHE regulatory and wildfire reassessment, printed pages 13-14; Charlie Munger tribute, printed pages 5-6. https://www.berkshirehathaway.com/letters/2023ltr.pdf Evidence note: First annual letter after Munger's death. Buffett's account of their relationship is retrospective and affectionate; the contemporaneous letters show the revisions unfolding more gradually. - Warren E. Buffett, 2024 Chairman's Letter (2024 reporting year; signed February 22, 2025), Mistakes, delayed correction, GEICO, cash, Berkshire's sixty-year record and succession. https://www.berkshirehathaway.com/letters/2024ltr.pdf Evidence note: Buffett's final Berkshire annual letter. It was reviewed from the complete local original. - Warren E. Buffett, Greg Abel and Ajit Jain, 2025 Berkshire Hathaway annual meeting (May 3, 2025), Buffett's closing succession announcement and meeting record. https://buffett.cnbc.com/2025-berkshire-hathaway-annual-meeting/ Evidence note: CNBC's authorized archive page. The meeting was reviewed in part, with emphasis on the succession announcement. - Warren E. Buffett, Thanksgiving Message to Fellow Shareholders (November 10, 2025), Complete message, including succession, philanthropy, delayed personnel action and final reflections. https://berkshirehathaway.com/news/nov1025.pdf Evidence note: Latest Buffett-authored material located on Berkshire's official site as of September 9, 2026. Read in full. - Greg Abel, 2025 Berkshire Hathaway annual letter (2025 reporting year; published February 28, 2026), Opening, authorship, culture and stewardship. https://www.berkshirehathaway.com/letters/2025ltr.pdf Evidence note: This is Greg Abel's first annual letter as Berkshire CEO. It is successor evidence, not Buffett-authored material. ## Jeff Bezos · Development · 1994-1997: A clean origin story rests on shared risk and fast correction Bezos later organized Amazon's beginning around a single choice. Internet use was growing quickly, he left a strong position at D. E. Shaw, and he acted before future regret could harden. That version is useful because it identifies a decision under uncertainty. It is incomplete as history because the regret frame comes from later interviews and a 2010 speech, after Amazon's outcome was known. The 2001 Academy of Achievement interview supplies a rougher sequence. Bezos says he first built a 30-page plan while knowing it would not survive contact with reality. His boss, David Shaw, advised him that the idea might make more sense for someone without a good job. MacKenzie Bezos supported the move. His parents put a few hundred thousand dollars into a venture Bezos told them had a 70-percent chance of losing their money. Roughly 60 investor meetings produced about 20 checks, commonly near $50,000. This was persistence, but it was persistence financed and enabled by other people. The operating beginning also resists the image of a lone visionary executing a finished idea. Bezos credited Shel Kaphan as the early system architect and one of Amazon's most important founders in practice. Within about 30 days of launch, the company had taken orders from all 50 states and 45 countries, but the team was badly prepared to pack them. Ten people worked at night on a concrete floor. Bezos proposed knee pads. A coworker proposed packing tables, and productivity roughly doubled. The point is developmental. The founder was not the person with the best immediate answer. The useful behavior was to accept the obvious correction and change the physical system. The 1997 shareholder letter shows what had become more explicit after those improvised years. Bezos declared that Amazon would favour long-term market leadership, customer value and cash generation over short-term accounting appearance. He also promised measurement, stopping programs with unacceptable returns, scaling the ones that worked, disciplined spending and careful attention to dilution. The letter therefore did more than celebrate boldness. It paired a long horizon with review and abandonment. Its evidence was specific to the 1997 reporting period. Sales rose from $15.7 million in 1996 to $147.8 million in 1997. Customer accounts rose from 180,000 to 1.51 million. Distribution space expanded to 285,000 square feet. Amazon ended 1997 with $125 million in cash and investments after financing. Those numbers showed demand, capacity strain and a temporary capital cushion. They did not prove that the declared philosophy caused the growth. Two developments matter most. First, Bezos had already revised his education and career path. In later accounts he describes leaving theoretical physics after encountering classmates whose mathematical command exceeded his own and moving toward computer science and finance. This was not persistence with one identity. It was persistence in finding a field where his skills and opportunity met. Second, Amazon's first operational lesson came from a coworker, not from the strategic plan. A serious student should separate the later narrative from the event. The later narrative gives Bezos's meaning. The event record shows distributed contribution, borrowed risk and rapid correction. The transferable practice is to write down the choice, tell capital providers the real downside, and keep the operating method open to people closer to the work. ### Work the question Reconstruct Amazon's founding without using the regret-minimization story as proof. What evidence belongs under decision, risk bearer, operating correction and later interpretation? ### Compare with the guide’s reasoning The decision was Bezos's move from D. E. Shaw after identifying rapid web growth. MacKenzie Bezos, his parents, early investors and employees bore or enabled material risk. Packing-table adoption is an operating correction that came from a coworker after demand exposed a bad process. The regret frame is Bezos's later interpretation. It can explain how he wanted the choice understood, but it cannot establish his full state of mind in 1994. ### Sources - Jeffrey P. Bezos, Jeff Bezos interview (May 4, 2001), Full Academy of Achievement interview transcript. https://achievement.org/achiever/jeffrey-p-bezos/ Evidence note: Early retrospective account with specific credit to MacKenzie Bezos, Shel Kaphan, Bezos's parents, investors and early coworkers. Motivational claims remain his account. - Jeffrey P. Bezos, 2010 Baccalaureate Remarks (May 30, 2010), Complete Princeton transcript. https://www.princeton.edu/news/2010/05/30/2010-baccalaureate-remarks Evidence note: Later moral retelling of childhood and the decision to start Amazon. Use as retrospective framing rather than contemporaneous evidence of 1994 motives. - Jeffrey P. Bezos, 1997 Letter to Shareholders (1997 reporting period; published 1998), It's All About the Long Term, Obsess Over Customers, Infrastructure, and Goals for 1998. https://www.aboutamazon.com/news/company-news/amazons-original-1997-letter-to-shareholders Evidence note: Contemporary statement of management policy and 1997 results. Management's claims about market leadership and future economics are its judgments at the time. ## Jeff Bezos · Development · 1998-1999: Growth stops being a website problem The 1998 and 1999 letters show a founder learning that demand is only the beginning of retail. During the 1998 reporting period, sales reached $610 million and cumulative customer accounts reached 6.2 million. Repeat customers represented more than 64 percent of fourth-quarter orders. Amazon was no longer validating whether people would buy books online. It was trying to build systems, distribution capacity, teams and processes quickly enough to carry demand across more products and countries. Bezos wrote that customers could switch with a click if another offer became better. That fact turned customer loyalty from an asset Amazon owned into an operating performance the company had to earn repeatedly. Management's 1999 plan was costly by design. It called for distribution and systems expansion, brand investment, new categories and stronger management processes. The following letter reports how violent that expansion became. During the 1999 reporting period, sales reached $1.64 billion and customer accounts reached 16.9 million. Amazon added auctions, zShops, toys, electronics, home improvement, software, video games, payments and a wireless initiative. It expanded distribution capacity from roughly 300,000 square feet to more than 5 million square feet in less than 12 months. A student looking only at sales would miss the developmental pressure. Amazon had complicated the business faster than it had accumulated operating knowledge. Bezos's task moved from choosing an attractive market to coordinating warehouses, software, acquisitions and managers without letting service collapse. The letters also reveal a belief that had not yet been fully tested. Amazon expected an online platform with many stores and third-party offers to gain power from shared customers, infrastructure and brand. Some elements endured. Marketplace later became central. Others named confidently in 1999 did not. Auctions and zShops attracted little demand and were later absorbed into the Marketplace learning story. This before-and-after comparison matters because a successful platform strategy can make every early launch look like a deliberate stepping stone. The contemporary letter does not know which seed will survive. It lists a portfolio of efforts, records high growth and argues that the infrastructure will serve a much larger company. Costs were already distributed. Employees had to build and keep a round-the-clock store operating while remaking its architecture. Suppliers and logistics partners had to accommodate a fast-changing retailer. Investors financed capacity before the eventual economics were visible. Competing retailers faced a company willing to reinvest and price for long-run share. The 1998 letter warns candidates that work at Amazon demands long, hard and smart effort. That admission establishes an intended work standard, not its fairness or sustainability. Later reporting supplies evidence that the same standard could impose personal and health costs. The developmental mechanism was escalation with partial controls. Bezos did not respond to complexity by narrowing Amazon immediately. He added categories and capacity while asking managers to set goals, review businesses and preserve cash. That approach produced a larger option set, but it also magnified what had to be corrected in the crash that followed. For application, track three inventories during hypergrowth. Record customer demand, capability debt and commitments that cannot be unwound quickly. Growth can improve the first while silently worsening the other two. The point at which a website becomes an operating system for many businesses is also the point at which ambition needs sharper sequencing. ### Work the question What evidence in 1998 and 1999 would have warned a board that Amazon's growth was creating capability debt, even while customer measures improved? ### Compare with the guide’s reasoning The distribution footprint expanded by more than an order of magnitude in less than 12 months, the category and geography list multiplied, and management explicitly identified systems, bench strength and processes as constraints. Those facts show coordination load and fixed commitments rising faster than a simple customer count reveals. They do not prove failure, but they justify liquidity tests, capacity milestones and business-level stop rules. ### Sources - Jeffrey P. Bezos, 1998 Letter to Shareholders (1998 reporting period; published 1999), A Recap of 1998, Our Customers, and Goals for 1999. https://s2.q4cdn.com/299287126/files/doc_financials/annual/Shareholderletter98.pdf Evidence note: Contemporary follow-up reporting customer, revenue, repeat-order, cash-flow, inventory and infrastructure figures. The outcome does not isolate the effect of the 1997 policy. - Jeffrey P. Bezos, 1999 Letter to Shareholders (1999 reporting period; published 2000), A Recap of 1999, Goals for 2000, distribution capacity and platform expansion. https://s2.q4cdn.com/299287126/files/doc_financials/annual/Shareholderletter99.pdf Evidence note: Contemporary management account of hypergrowth, category expansion and infrastructure burden. Later outcomes show that several named initiatives failed. ## Jeff Bezos · Development · 2000-2002: The land rush breaks, and Bezos changes the sequence The 2000 letter is the sharpest contemporaneous revision in the Bezos corpus. Amazon's share price had fallen by more than 80 percent during the 2000 reporting period. Bezos opened with an admission of pain for owners, then separated the stock movement from what management believed about customers and the business. That distinction could preserve rational action under market panic. It could also protect management from valid criticism, so the operating evidence matters more than the slogan. Amazon held about $1.1 billion in cash and marketable securities after raising convertible debt early in 2000. The cushion meant it could respond to the collapse from a financed position. The letter then changes an earlier belief. Amazon had treated online commerce as a land rush and invested in single-category companies such as Pets.com and living.com. Bezos says management had underestimated how much time and difficulty those companies would face in reaching scale and profitability. Amazon had lost significant money in those investments and would not keep funding them indefinitely. The update was not a rejection of the internet, customer growth or platform economics. It was a revision in capital sequencing. A market can be real while a vehicle, timing or funding structure is wrong. Persistence belonged to the larger thesis. Abandonment belonged to ventures that could not survive the new constraint. The 2001 letter shows the operational consequence. Bezos describes four years of concentrated growth followed by almost two years spent chiefly reducing costs. Only then did Amazon believe it could balance growth with cost improvement. The fourth quarter of the 2001 reporting period produced $59 million of pro forma operating profit and $35 million of pro forma net profit under Amazon's stated measures. Sales for the year rose 13 percent to $3.12 billion, slower than the hypergrowth years. Inventory turns increased from 12 in 2000 to 16 in 2001. Marketplace reached 15 percent of United States orders in the fourth quarter. The sequence changed from build broadly, then seek scale to cut cost, lower selected prices, grow volume and spread fixed expense. The 2002 letter adds evidence and limits. Amazon reported $135 million of free cash flow under its definition, compared with negative $170 million in 2001. It described lower defects, improved customer contacts and a 100-book price comparison. Those observations supported management's price and experience loop for that period. They did not establish that every category had attractive economics or that all costs had disappeared. The crash narrative can become too neat if 2000 is remembered only as proof of resolve. The 2000 letter also records capital loss and failed companies. Its account of the cost program does not establish the full employee consequences, so those should not be inferred from the financial improvement alone. The development is best understood as a change in what persistence meant. In 1999, persistence looked like expanding the opportunity set before competitors. By 2001, it looked like preserving liquidity, lowering structural cost and concentrating on mechanisms that could fund themselves. Later interviews retell the stock decline as a period when internal measures improved, and that is partly supported by the letter sequence. Yet the ability to persist depended on the capital already raised and on imposing costs across employees, investees and owners. A useful response to collapse therefore has four columns. Keep the enduring customer fact, name the broken assumption, stop the exposed commitments and state the financial bridge to the next proof point. ### Work the question Which part of Bezos's thesis survived the dot-com collapse, which assumption failed, and what changed in Amazon's operating sequence? ### Compare with the guide’s reasoning The belief that customers valued online selection, convenience and lower prices survived. The land-rush assumption that focused e-commerce companies could quickly reach scale and profitability failed in the ventures Bezos named. Amazon stopped open-ended funding, spent almost two years reducing costs, and then paired growth projects with a price-cost-volume loop and cash discipline. The bridge depended on prior financing, so survival was not resolve alone. ### Sources - Jeffrey P. Bezos, 1999 Letter to Shareholders (1999 reporting period; published 2000), A Recap of 1999, Goals for 2000, distribution capacity and platform expansion. https://s2.q4cdn.com/299287126/files/doc_financials/annual/Shareholderletter99.pdf Evidence note: Contemporary management account of hypergrowth, category expansion and infrastructure burden. Later outcomes show that several named initiatives failed. - Jeffrey P. Bezos, 2000 Letter to Shareholders (2000 reporting period; published 2001), Opening share-price discussion, platform investments and Goal for 2001. https://s2.q4cdn.com/299287126/files/doc_financials/annual/00ar_letter.pdf Evidence note: Contemporary admission that the land-rush metaphor had understated the time and difficulty required for single-category companies to reach scale. - Jeffrey P. Bezos, 2001 Letter to Shareholders (2001 reporting period; published 2002), Opening discussion of price, cost reduction, growth, fixed costs and cash flow. https://ir.aboutamazon.com/files/doc_financials/annual/2001_shareholderLetter.pdf Evidence note: Management's contemporary explanation of a reinforcing operating cycle after reporting its first profitable quarter on the letter's pro forma measures. - Jeffrey P. Bezos, 2002 Letter to Shareholders (2002 reporting period; published 2003), Opening discussion of customer experience, price, satisfaction, the 100-book comparison, and free cash flow. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2002_shareholderLetter.pdf Evidence note: Contemporary company evidence. The 100-book basket, satisfaction score and management's free-cash-flow definition have specific scopes and do not prove a universal price rule. - Jeffrey P. Bezos, Jeff Bezos interview (May 4, 2001), Full Academy of Achievement interview transcript. https://achievement.org/achiever/jeffrey-p-bezos/ Evidence note: Early retrospective account with specific credit to MacKenzie Bezos, Shel Kaphan, Bezos's parents, investors and early coworkers. Motivational claims remain his account. ## Jeff Bezos · Development · 2003-2006: Survival lessons become a decision system After the crash, Bezos's letters become more explicit about how decisions should connect customer behaviour to owner returns. The 2003 letter uses the difference between an owner and a tenant. Amazon would accept some lost sales by displaying critical reviews and warning a customer about a duplicate purchase because management expected trust to matter more over time. This was written after Amazon reported its first full-year profit for the 2003 reporting period. The financial context matters. Long-term language was no longer floating above a company with only a future profitability target. Management now had to decide which apparent sacrifices actually built an enduring asset. The 2004 letter supplies the financial control. Its transportation-machine example shows earnings rising while free cash flow collapses under repeated capital spending. Amazon's stated objective becomes free cash flow per share, with attention to capital intensity, working capital and dilution. In the 2004 reporting period, Amazon reported $477 million in free cash flow under its definition, $480 million in year-end inventory on nearly $7 billion in sales and more than $600 million of convertible debt repaid. The figures do not prove that every Amazon investment met the standard. They show what management now said an investment had to become. Growth needed to release cash without consuming ownership faster than it created value. The 2005 letter separates repeated operating decisions from choices that cannot be settled by near-term data. Fulfilment-centre location and inventory placement could use historical demand and logistics data. Repeated price cuts could not be judged only by weekly elasticity because management believed their effect on customer trust would unfold across years. Marketplace supplies a later result. Third-party units rose from 6 percent of total units in 2000 to 28 percent in 2005 while Amazon's own retail revenue roughly tripled. That outcome is consistent with the choice to let third-party sellers compete on the detail page. It does not reveal what sales or cash would have been under another design. Bezos calls the decision judgment-based because the missing long-run counterfactual could not be calculated cleanly at the start. The 2006 letter then turns the system toward new businesses. A proposed seed had to offer attractive returns at scale, address an underserved need and use capabilities that could create real differentiation. Kindle, Fulfilment by Amazon and Amazon Web Services were emerging from this period. So were categories such as apparel and grocery. Bezos explicitly rejected opening physical stores because he could not then see a differentiated model and expected high capital requirements. That statement is important precisely because it did not remain permanent. Whole Foods and Amazon Go later changed the evidence. A criterion earned more loyalty than a conclusion. This is development through codification. The company moved from surviving a financing and cost crisis to writing a stack of filters. Customer trust could justify a near-term sacrifice. Cash per share disciplined growth. Models handled repeated operations. Judgment handled uncertain long-term effects. New ventures needed return, scale and differentiation. The risk is that a good framework can become a retrospective explanation attached only to winners. Auctions, zShops and failed investments also passed through Amazon's decision-making environment. A board using these letters should require a decision journal before the result, including who pays during the learning period. Sellers may fund selection, employees may fund speed through workload, suppliers may fund working capital through payment terms, and shareholders may fund options through dilution. The decision system is credible only when those transfers stay visible. ### Work the question How did Amazon's decision architecture change from the 2003 owner frame through the 2006 seed test, and where can hindsight still enter? ### Compare with the guide’s reasoning The sequence added an owner horizon, free cash flow per share, a distinction between model-based and judgment-based choices, and a return-scale-differentiation test for new businesses. Hindsight enters when later winners are used to validate the framework without counting rejected ideas, failed launches or costs shifted to other participants. A contemporaneous decision journal and full attempt denominator reduce that bias. ### Sources - Jeffrey P. Bezos, 2003 Letter to Shareholders (2003 reporting period; published 2004), Long-term owner discussion, negative reviews and Instant Order Update. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2003_-Shareholder_-Letter041304.pdf Evidence note: Contemporary owner-versus-tenant frame following Amazon's first full-year profit. The examples describe deliberate short-term sales sacrifices. - Jeffrey P. Bezos, 2004 Letter to Shareholders (April 2005), Our Most Important Financial Measure, Free Cash Flow Per Share, including the transportation-machine example. https://ir.aboutamazon.com/files/doc_financials/annual/2004_Annual_report.pdf Evidence note: Uses a deliberately simplified fictional machine to separate earnings growth from cash generation, followed by Amazon's 2004 company figures. - Jeffrey P. Bezos, 2005 Letter to Shareholders (2005 reporting period; published 2006), Opening discussion of math-based and judgment-based decisions, price reductions, and the single detail page. https://ir.aboutamazon.com/files/doc_financials/annual/AMZN2005AnnualReport.pdf Evidence note: Contemporary account of Amazon's decision policy with retrospective results for the Marketplace choice made in 2000. Long-run price effects were explicitly not measurable in advance. - Jeffrey P. Bezos, 2006 Letter to Shareholders (2006 reporting period; published 2007), Planting Seeds section and discussion of physical stores, FBA, AWS and new categories. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2006.PDF Evidence note: Contemporary new-business test. The rejection of physical stores records a position that later changed when management saw differentiation. ## Jeff Bezos · Development · 2007-2010: Product pressure becomes reusable capability Kindle's launch is useful because the 2007 letter describes both preparation and surprise. Amazon worked for more than three years on the device and service. The team was trying to improve the book rather than merely place a screen around the same transaction. When Kindle went on sale, launch inventory sold out in 5.5 hours. Management had to scramble for supply. Long development did not eliminate forecast error. Demand exceeded launch inventory, making production, availability and demand forecasting immediate constraints. The letter does not establish how much of the sellout reflected product appeal versus the quantity initially stocked. Bezos also says Amazon had no guaranteed right to lead the transition. If execution failed, someone else could do it. The 2008 recession put a second kind of pressure on that commitment. Bezos argued that Amazon should work backward from durable customer needs instead of forward from skills the company already possessed. Customers would continue to value lower prices, greater selection and faster delivery. During the 12 months before the letter, Amazon estimated that free shipping saved customers more than $800 million. Prime still cost $79 per year in the United States during the 2008 reporting period, and Fulfilment by Amazon shipped more than 3 million third-party units in the fourth quarter. These measures showed adoption of a combined retail and fulfilment system. They did not measure the total investment required or the labour burden inside the network. By 2010, the letters describe technology less as a separate project and more as an operating substrate. Amazon's scale had forced teams to break the company into hundreds of services with defined interfaces. That architecture let internal capabilities become external products through AWS. The sequence matters. Bezos did not begin with a generic plan to sell cloud computing to the world. Later accounts describe application and infrastructure teams repeatedly confronting the same internal bottlenecks, then recognizing that the tools could serve outside developers. The 2010 letter links technical work directly to free cash flow, but this is a management claim about contribution rather than a measured attribution. The 2018 Economic Club interview adds team and luck to the mature story. Bezos credits the internal work that exposed the need for standard infrastructure and calls the long period without a comparable cloud competitor exceptional luck. That admission qualifies a simple persistence narrative. AWS required conviction and continued investment, but its extraordinary result also depended on market timing and a seven-year competitive opening that management could not command. Kindle likewise depended on publishers, authors, device engineers and customers accepting a new reading system. The developmental mechanism is capability retention. A difficult product or internal tool can create people, architecture and operating knowledge that outlast its first use. That does not make every troubled project valuable. The relevant test is whether the capability is identifiable, transferable and cheaper to retain than to rebuild. Kindle turned book retail into a device, content and wireless-service problem. AWS turned internal service discipline into a customer business. In each case, the useful persistence was not repeating the first plan. Teams kept the durable customer need and revised the implementation under new constraints. A practical review should ask what demand signal changed, what capability the work created, what bottleneck appeared next and whether luck or market structure, rather than management skill, explains part of the result. ### Work the question Why is AWS weak evidence for the claim that persistence alone produces large businesses, and what stronger lesson does the sequence support? ### Compare with the guide’s reasoning AWS grew from internal operating constraints, reusable service architecture, sustained investment and an outside customer need. Bezos later acknowledged an unusually long competitive opening. The stronger lesson is to identify transferable capability produced by hard work, test whether outsiders share the need and keep luck visible. Persistence without a capability or market update is repetition, not development. ### Sources - Jeffrey P. Bezos, 2007 Letter to Shareholders (2007 reporting period; published 2008), Kindle development, launch demand and supply response. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2007letter.pdf Evidence note: Contemporary account of a product developed for more than three years and then constrained by demand that exhausted launch inventory in 5.5 hours. - Jeffrey P. Bezos, 2008 Letter to Shareholders (2008 reporting period; published 2009), Working backward, durable customer needs and recession-era investment. https://s2.q4cdn.com/299287126/files/doc_financials/annual/Amazon_SH_Letter_2008.pdf Evidence note: Management's recession-era account of choosing opportunities from customer needs rather than from existing skills. - Jeffrey P. Bezos, 2009 Letter to Shareholders (April 2010), Opening results and the 452 goals for 2010. https://s2.q4cdn.com/299287126/files/doc_financials/annual/AMZN_Shareholder-Letter-2009-(final).pdf Evidence note: Reports the content of one annual planning set. It shows Amazon's chosen management emphasis, not proof that financial outputs can be ignored. - Jeffrey P. Bezos, 2010 Letter to Shareholders (2010 reporting period; published 2011), Service architecture, technology examples and connection to free cash flow. https://s2.q4cdn.com/299287126/files/doc_financials/annual/117006_ltr_ltr2.pdf Evidence note: Management account of technical capabilities developed under operating scale. It links engineering work to cash flow without isolating causal contribution. - Jeffrey P. Bezos and David M. Rubenstein, Interview at the Economic Club of Washington (September 13, 2018), Complete edited transcript. https://www.economicclub.org/sites/default/files/transcripts/Jeff_Bezos_Edited_Transcript.pdf Evidence note: Later retrospective account that distributes credit among mentors, colleagues and family, and identifies exceptional luck in AWS's competitive lead. ## Jeff Bezos · Development · 2011-2014: The platform that removes gatekeepers becomes one The 2011 letter presents a coherent development in Amazon's identity. Marketplace, Fulfilment by Amazon, Kindle Direct Publishing and AWS let sellers, authors and developers reach customers without persuading a traditional retailer, publisher or technology buyer. Bezos describes these as self-service platforms. The argument is not only that Amazon grew. It is that outside participants gained tools, distribution and demand. The contemporary measures are large. Amazon reported 900 billion objects stored in S3 by the end of the 2011 reporting period, and the letter gave selected examples of authors and sellers building businesses on Amazon's infrastructure. Those examples demonstrate possibility, not a representative outcome. The 2012 and 2013 letters show the organizational burden behind that platform expansion. AWS added 159 features and services in the 2012 reporting period and had cut prices 27 times since launch, according to management. Bezos admits that proactive invention produces blind alleys and self-inflicted mistakes. Amazon also reports a growing Weblab program, from 546 tests in 2011 to 1,092 in 2012 and 1,976 in 2013. Amazon Fresh had remained in limited trial for about five years before broader expansion. Fulfilment-centre associates participated in thousands of improvement efforts, while Career Choice offered training for jobs that might take employees outside Amazon. The record shows repeated revision and long incubation, but it still gives selected initiatives rather than a complete attempt denominator. The 2014 letter retells the era after outcomes were visible. Marketplace, Prime and AWS now appear as three bold ideas that met attractive business tests. Bezos does acknowledge that Auctions and zShops attracted almost no customers. Prime required Amazon to surrender many millions of dollars in shipping revenue before management knew whether membership behaviour would repay it. AWS involved high switching costs and long investment. The account is valuable because it includes failure and delay. It remains a winner-centred retrospective. The businesses that became large receive detailed origin stories, while abandoned work receives less economic accounting. Independent evidence complicates the empowerment narrative. The 2020 House Judiciary majority staff report describes Amazon as both marketplace operator and competing seller. It cites seller dependence, rising fees, forced arbitration, asymmetric access to seller data and internal documents about competitive actions. Those findings do not erase the real distribution and fulfilment services third parties received. They show that removing one gatekeeper can create another when participants cannot reach the same customers elsewhere. The report is a congressional staff assessment, not a judicial finding, and Amazon disputed central conclusions in its testimony. That disagreement belongs inside the teaching rather than below it. This before-and-after test changes the developmental question. Early Amazon had to persuade sellers that joining the platform would increase opportunity. Mature Amazon had to govern conflicts produced by its own success. The same flywheel that expanded selection and seller sales could increase dependence, allowing Amazon to set fees, rank offers, adjudicate disputes and introduce competing products. Costs fell on sellers who had invested around rules they did not control. Customers could receive lower prices and more selection while suppliers carried bargaining risk. Platform persistence therefore needs a constitutional update. As a platform gains power, it should separate rule-setting from self-preference, define data boundaries, publish appeal standards and measure participant concentration. A founder story about empowerment is strongest when it survives evidence from people who became dependent on the system. ### Work the question How can Marketplace be both an empowering invention and a source of dependence, and what evidence would distinguish those effects? ### Compare with the guide’s reasoning Marketplace can lower entry barriers by supplying traffic, fulfilment and trust while also concentrating access, data and dispute control in Amazon. Measure seller revenue and survival alongside fee burden, multi-homing, suspension appeals, data use and the share of sellers dependent on Amazon. The House report supplies contrary evidence and allegations, while the letters supply Amazon's claimed benefits. Neither alone settles the full effect. ### Sources - Jeffrey P. Bezos, 2011 Letter to Shareholders (2011 reporting period; published 2012), Self-service platforms, AWS, Marketplace, KDP and FBA. https://s2.q4cdn.com/299287126/files/doc_financials/annual/letter.PDF Evidence note: Management's case that Amazon platforms let outside participants bypass older gatekeepers. Later institutional evidence tests Amazon's own gatekeeper role. - Jeffrey P. Bezos, 2012 Letter to Shareholders (2012 reporting period; published 2013), Proactive customer invention, blind alleys, device economics and AWS iteration. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2012-Shareholder-Letter.pdf Evidence note: Management acknowledges self-inflicted mistakes and blind alleys while describing continuing investment in devices and cloud services. - Jeffrey P. Bezos, 2013 Letter to Shareholders (2013 reporting period; published 2014), Initiative catalogue, experiments, Amazon Fresh, fulfilment-centre improvement and employee programs. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2013-Letter-to-Shareholders.pdf Evidence note: A wide management inventory of initiatives, experimentation and operating practices. It gives selected results, not a complete denominator of attempts or costs. - Jeffrey P. Bezos, 2014 Letter to Shareholders (2014 reporting period; published 2015), Opening business-quality test and the Marketplace, Prime, FBA and AWS sections. https://ir.aboutamazon.com/files/doc_financials/annual/AMAZON-2014-Shareholder-Letter.pdf Evidence note: Retrospective account of several initiatives after they had shown traction. Management's business-quality assessments and causal explanations remain judgments. - U.S. House Judiciary Subcommittee on Antitrust, Commercial and Administrative Law, Investigation of Competition in Digital Markets: Majority Staff Report and Recommendations (October 6, 2020), Amazon overview and selected marketplace, seller-data, fee, private-label, acquisition and logistics passages. https://www.govinfo.gov/content/pkg/GOVPUB-Y4_J89_1-PURL-gpo145949/pdf/GOVPUB-Y4_J89_1-PURL-gpo145949.pdf Evidence note: Read in part. The majority staff report draws on documents and market participants but is not a judicial finding and was not read outside the Amazon-relevant portions. ## Jeff Bezos · Development · 2015-2016: Bezos tries to preserve speed, while employees absorb the contest By 2015, Amazon had passed $100 billion in annual sales and AWS was approaching a $10 billion annual run rate. The founder's problem was no longer proving that small teams could invent. It was preventing a large organization from routing reversible choices through slow, uniform process. Bezos divided decisions by reversibility. Hard-to-reverse commitments deserved heavy analysis. Reversible choices could move with less information and correction. Prime Now, built in 111 days, illustrated the speed he wanted. The 2016 letter expands the response into Day 1 and Day 2. Day 2 meant stasis, process as a substitute for outcomes and declining decision velocity. Bezos advised acting with about 70 percent of the information one might eventually obtain, using disagreement without requiring consensus and escalating misalignment quickly. This method was a revision to scale, not a complete founding doctrine. Earlier letters concentrated on capital, price, selection and cash. The 2015 and 2016 letters spend more attention on how hierarchy changes decisions. They also try to turn failure into an economic portfolio. A large win can repay many bounded attempts, but the illustrative 10-percent chance of a 100-times payoff is not a measured Amazon base rate. Fire Phone had already shown that a highly visible bet could destroy capital. The useful management question is whether teams can close a failed project, retain relevant capability and prevent the winner narrative from hiding the full loss denominator. The same period provides a contrary test of what speed and standards cost. Kantor and Streitfeld's 2015 New York Times investigation drew on more than 100 current and former employees. It describes late-night demands, combative meetings, continuous metrics, confidential peer feedback, forced ranking and employees who said family or health crises damaged their standing. Some interviewees also described exceptional colleagues, responsibility and invention. Amazon disputed that the harshest accounts represented policy or normal practice. The article therefore does not prove one uniform culture. It does establish that the mechanism Bezos praised had human consequences and serious internal disagreement. The praised decision-speed system coexisted with reports of heavy workload, uncertainty and interpersonal contest. The reporting does not isolate whether those burdens improved speed or instead impaired it. That cost cannot be dismissed as comfort sacrificed for ambition. If a system burns out capable people, hides information from managers or rewards political attacks, it can also weaken the very decision quality it is meant to protect. The New York Times reported that employees prepared defensive records for ranking sessions and sometimes hoarded candidates or support. Those behaviours are plausible balancing loops against the ideal of candid challenge. The 2016 Code Conference liveblog shows Bezos publicly discussing large opportunities and technological humility, but the complete audio was not reviewed. It cannot repair or rebut the reporting. Development here is incomplete. Bezos correctly identifies that a mature company needs different process for different commitments. He does not, in the shareholder letters, offer an equally developed distinction between productive pressure and destructive pressure. A better operating rule would classify both decision and burden. State how reversible the commercial choice is, who must work outside normal limits to deliver it, whether that burden is voluntary, and what evidence shows learning rather than exhaustion. Speed can be an advantage. Speed purchased through hidden personal cost is also a liability that should appear in the decision record. ### Work the question Apply Bezos's reversible-decision idea to the work system itself. Which employee costs are reversible, and which can become hard-to-reverse commitments? ### Compare with the guide’s reasoning A short, consented surge with recovery may be reversible. Chronic overwork, health harm, family disruption, reputation loss, talent flight and a culture of defensive ranking can compound and become difficult to repair. Decision reviews should record burden, duration, consent, recovery and retention effects alongside launch speed. A fast product decision does not make the labour used to execute it reversible. ### Sources - Jeffrey P. Bezos, 2015 Letter to Shareholders (2015 reporting period; published 2016), Failure, long-tailed business returns, and Type 1 versus Type 2 decisions. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2015-Letter-to-Shareholders.PDF Evidence note: Retrospective management account. The ten-percent and one-hundred-times illustration explains an asymmetric payoff shape, not a measured base rate for Amazon experiments. - Jeffrey P. Bezos, Amazon 2016 shareholder letter (2016 reporting year; published 2017), Resist Proxies and High-Velocity Decision Making.. https://www.aboutamazon.com/news/company-news/2016-letter-to-shareholders Evidence note: Use the signed Bezos letter, not the hosting page’s staff byline. Distinguish his reported practices from proof that they caused Amazon’s results. - Jeffrey P. Bezos and Walter Mossberg, Jeff Bezos live from Code 2016 (May 31, 2016), Recode liveblog of the full stage interview; complete 81-minute audio separately located. https://live.recode.net/jeff-bezos-2016-code/ Evidence note: Read in part. The liveblog is not a verbatim transcript and the complete audio was not directly reviewed, so no omitted answer is treated as evidence. - Jodi Kantor and David Streitfeld, Inside Amazon: Wrestling Big Ideas in a Bruising Workplace (August 15, 2015), Complete reported article, including Amazon's responses. https://www.nytimes.com/2015/08/16/technology/inside-amazon-wrestling-big-ideas-in-a-bruising-workplace.html Evidence note: Reporting based on more than 100 current and former employees. Accounts vary, but the article documents costs and contest around the culture Bezos praised. ## Jeff Bezos · Development · 2017-2018: Bezos admits what colleagues had to teach him The 2017 letter contains a rare direct admission about founder development. Bezos says high standards are teachable and domain-specific, then explains that his own early standards for operating processes were low. Colleagues had to teach him what strong operations looked like. This matters because Amazon's later mythology can make operational excellence appear to have arrived fully formed from customer obsession. The earlier letters show the opposite. Demand outran packing methods, capacity expanded violently, the crash forced cost work, and mature operating systems emerged through many people. Bezos's admission gives those participants explicit causal weight. The same letter makes the time cost of quality visible. A strong six-page narrative may require a week or more of work even when the final document reads easily. A team can fail by underestimating the scope rather than lacking skill. Bezos also reports that Amazon has produced billions of dollars in failures. That is more candid than celebrating only small, reversible experiments. It still does not provide a project-level ledger, the burden on the people whose work ended, or a denominator that would let an outsider estimate Amazon's hit rate. Whole Foods supplies a concrete change of mind. In the 2006 letter, Bezos rejected physical stores because he could not see differentiation and expected high capital needs. During the 2017 reporting period, Amazon completed the Whole Foods acquisition and expanded Amazon Go. The contradiction is productive. The original criterion survived while the conclusion changed. Physical presence became acceptable when management believed technology, brand, logistics or the acquired asset could differentiate the experience. A rigid founder would have defended the earlier answer. A learning founder applies the earlier test to new facts. The 2018 letter adds another revision. Third-party gross merchandise sales had risen from 3 percent of Amazon's total in the 1999 reporting period to 58 percent in 2018. Bezos says third-party sellers had grown faster than Amazon's first-party business and attributes part of that difference to seller tools, FBA and Prime. He also uses wandering to describe discovery that cannot be planned cleanly. Early machine-learning tools failed to attract many outside developers before later work became SageMaker. Fire Phone failed as a product, while people and knowledge from the effort moved to Echo and Alexa. Retaining capability after failure is evidence of revision only because the original product was actually stopped. The 2018 Economic Club interview further distributes invention. Bezos credits David Shaw for lessons about hiring, Bing Gordon for emphasizing recurring loyalty and a junior engineer for proposing the unlimited-shipping idea that became Prime. He says the first financial model for Prime looked alarming because heavy users arrived first. Trend lines, not initial averages, supported continued investment. He also identifies AWS's long competitive lead as luck. These credits make the mature story more accurate. Bezos still chose and defended large commitments, but ideas, operating standards, capital and correction came from a network. The developmental lesson is to preserve criteria while allowing conclusions to expire. Document why a rejected format failed the test in 2006, then show which fact changed by 2017. Close a product that failed, then name the specific capability worth transferring. Credit the person who supplied the decisive idea or correction. This form of persistence is demanding because it denies the founder exclusive authorship and denies the old answer permanent status. It makes the revision teachable rather than merely successful. ### Work the question What is the strongest evidence that Bezos changed rather than simply applied one fixed doctrine throughout Amazon's history? ### Compare with the guide’s reasoning He explicitly says colleagues raised his weak operating standards, reverses his 2006 physical-store conclusion when differentiation changes, and retells Prime and AWS with credit to employees, mentors and luck. The underlying criteria remain recognizable, but the answers, capabilities and source of knowledge change. That is stronger development evidence than a later claim that Day 1 explained everything from the start. ### Sources - Jeffrey P. Bezos, 2006 Letter to Shareholders (2006 reporting period; published 2007), Planting Seeds section and discussion of physical stores, FBA, AWS and new categories. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2006.PDF Evidence note: Contemporary new-business test. The rejection of physical stores records a position that later changed when management saw differentiation. - Jeffrey P. Bezos, 2017 Letter to Shareholders (2017 reporting period; published 2018), High standards, operating standards admission, six-page memos and Whole Foods. https://s2.q4cdn.com/299287126/files/doc_financials/annual/Amazon_Shareholder_Letter.pdf Evidence note: Bezos states that colleagues taught him operating standards he initially lacked. The Whole Foods acquisition also revises his 2006 physical-store position. - Jeffrey P. Bezos, 2018 Letter to Shareholders (2018 reporting period; published 2019), Third-party sales, wandering, Fire Phone transfer, wage challenge and employee programs. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2018-Letter-to-Shareholders.pdf Evidence note: Retrospective comparison of first-party and third-party growth, plus management's account of retaining people and capability after the Fire Phone failure. - Jeffrey P. Bezos and David M. Rubenstein, Interview at the Economic Club of Washington (September 13, 2018), Complete edited transcript. https://www.economicclub.org/sites/default/files/transcripts/Jeff_Bezos_Edited_Transcript.pdf Evidence note: Later retrospective account that distributes credit among mentors, colleagues and family, and identifies exceptional luck in AWS's competitive lead. ## Jeff Bezos · Development · 2018-2019: The labour question moves into the shareholder letter Amazon's labour model was present from the beginning, but it was usually described as culture, ownership and high standards. The 2018 letter changes the emphasis. Amazon raised its minimum wage to $15 per hour for more than 250,000 employees and more than 100,000 seasonal workers in the United States during the 2018 reporting period, then challenged competitors to match the move. Bezos also highlighted Career Choice and other employee programs. The policy delivered a concrete benefit to a large workforce. It also arrived after years of criticism about warehouse work, corporate pressure and the distribution of gains from Amazon's scale. The development should not be reduced to either benevolence or capitulation. A wage floor can improve pay, recruitment and retention while still leaving pace, surveillance, injury, scheduling and voice unresolved. The 2015 New York Times investigation concentrated on corporate employees and included some warehouse context. It described a system in which data, peer contest and high workload helped Amazon extract more effort. Amazon's response was that the harshest accounts did not reflect policy or the experience of many employees. The coexistence of those accounts is the evidence. Amazon had built a system some people found unusually generative and others found personally damaging. Higher pay did not answer every claim about that system. The 2019 shareholder letter was published in April 2020, during the first acute stage of COVID-19. It reports more than 150 operating changes, masks, temperature checks, temporary wage premiums, a testing effort, a $25 million relief fund and 175,000 announced jobs. Amazon estimated more than $500 million of pandemic-related spending through April 2020. These quantities describe management's response during a defined early period. They should not be read as proof that protection was sufficient in every facility. The same rapid network that created customer value moved risk onto workers required to enter warehouses and deliver packages while many customers stayed home. The founder narrative also begins to widen from customer and owner to other stakeholders. Climate commitments, jobs, wages and pandemic service occupy growing space. This is partly an expansion of responsibility. It is also a response to political, labour and reputational pressure. The letters do not let us assign a single motive, and the course should not invent one. We can observe the sequence. Criticism intensified, Amazon made a wage commitment, a public-health emergency exposed worker dependence, and Bezos devoted more owner communication to employee claims. A serious update requires metrics that do not collapse different burdens into one compensation number. Track hourly pay by period, injury rate, time away from work, schedule stability, turnover, worker complaints and access to challenge automated or managerial decisions. Separate a pandemic surge from a normal operating target. Record who could remain remote and who could not. The Bezos method often asks which controllable inputs lead to customer outcomes. The labour revision requires the same precision for employee outcomes. If the company claims high standards are teachable, it should make safe pace, fair review and recovery standards as operational as package speed. ### Work the question Why is the 2018 wage increase evidence of development but not a complete answer to Amazon's labour costs? ### Compare with the guide’s reasoning It changed pay for a defined group and moved employee benefit into the shareholder narrative. Compensation does not measure injury, surveillance, workload, scheduling, turnover or voice. The earlier reporting and pandemic-era exposure show those dimensions remained material. A fair assessment keeps the wage gain and unresolved operating burdens in the same record. ### Sources - Jeffrey P. Bezos, 2018 Letter to Shareholders (2018 reporting period; published 2019), Third-party sales, wandering, Fire Phone transfer, wage challenge and employee programs. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2018-Letter-to-Shareholders.pdf Evidence note: Retrospective comparison of first-party and third-party growth, plus management's account of retaining people and capability after the Fire Phone failure. - Jeffrey P. Bezos, 2019 Letter to Shareholders (2019 reporting period; published April 2020), COVID-19 response, testing work, temporary pay, jobs and climate commitments. https://s2.q4cdn.com/299287126/files/doc_financials/2020/ar/2019-Shareholder-Letter.pdf Evidence note: Management's early-pandemic account, written while operating conditions were changing quickly. It does not independently assess worker experience or adequacy of protection. - Jodi Kantor and David Streitfeld, Inside Amazon: Wrestling Big Ideas in a Bruising Workplace (August 15, 2015), Complete reported article, including Amazon's responses. https://www.nytimes.com/2015/08/16/technology/inside-amazon-wrestling-big-ideas-in-a-bruising-workplace.html Evidence note: Reporting based on more than 100 current and former employees. Accounts vary, but the article documents costs and contest around the culture Bezos praised. ## Jeff Bezos · Development · 2020: Scale turns operating doctrine into a public-power question The 2020 House testimony repeats many elements of Bezos's origin story, but the setting changes their meaning. He was no longer persuading investors to finance an uncertain retailer. He was defending one of the world's largest companies before a congressional antitrust inquiry. The testimony states that Amazon had accumulated nearly $3 billion in losses by the end of the 2001 reporting period and recalls the stock falling from $116 to $6. It emphasizes customers, small businesses, jobs and continued invention. It also reports 1.7 million small and medium-sized businesses selling through Amazon stores and more than 200,000 sellers with more than $100,000 in sales during 2019. These are company claims made under scrutiny and need a denominator, distribution and dependency measure before they can establish seller welfare. The House Judiciary majority staff report provides the contrary record. Drawing on internal documents, interviews and submissions, the report concludes that Amazon's role as both marketplace operator and seller creates conflicts. It describes sellers who rely heavily on Amazon, fee increases, account suspensions, forced arbitration, Amazon's access to non-public seller data and acquisitions that may reduce competitive alternatives. The report also argues that low pricing can entrench power when a company can subsidize one business from another. Amazon disputed allegations and said it prohibited the use of individual seller data for private-label decisions. The report is a majority staff assessment rather than a court judgment. Its value here is that it identifies costs and power relationships missing from Bezos's empowerment account. The comparison changes how we read earlier principles. Customer obsession can improve price and convenience while weakening suppliers who have no equivalent route to demand. A long horizon can support Kindle or AWS, and it can also permit years of pricing that rivals cannot finance. Working backward from customers does not automatically include competitors, sellers or communities in the objective. Free cash flow per share says little about how bargaining surplus is divided. None of this proves that the principles were insincere. It shows they are incomplete as governance for a platform whose decisions shape other firms' survival. Pressure also changed Bezos's role. In 1997, he warned shareholders about competition, execution and financing. By 2020, the public question was whether Amazon itself had become infrastructure with power to set the conditions of competition. Persistence at this scale could no longer be judged only by surviving skeptics or delivering another service. It needed rules for conflicts of interest and evidence from people subject to Amazon's systems. The developmental test is whether a founder can revise doctrine when the firm changes category. A startup needs permission to pursue uncertain advantage. A dominant platform needs constraints on how it uses control, data and cross-subsidy. The same action can change character as alternatives disappear. A practical board exercise should map every major principle to a non-customer constituency. For Marketplace, record seller concentration and appeal outcomes. For private labels, audit data access. For acquisitions, preserve counterfactual competition. For logistics, record contractor and worker outcomes. Scale does not invalidate customer focus. It removes the excuse that customer benefit is the only public effect worth measuring. ### Work the question Which Bezos principles become incomplete when Amazon changes from retailer to infrastructure, and what governance additions follow? ### Compare with the guide’s reasoning Customer obsession omits sellers and rivals, free cash flow omits surplus distribution, and long-term investment can hide cross-subsidy or foreclosure. Add conflict rules for Amazon as operator and seller, auditable seller-data boundaries, transparent appeals, fee and dependency measures, and acquisition review that considers lost future competition. The House report supplies allegations and evidence to test, not a final judicial verdict. ### Sources - Jeffrey P. Bezos, Written Testimony before the House Judiciary Subcommittee (July 29, 2020), Complete written testimony. https://docs.house.gov/meetings/JU/JU05/20200729/110883/HHRG-116-JU05-Wstate-BezosJ-20200729.pdf Evidence note: Founder narrative delivered under antitrust scrutiny. It supplies specific claims but is advocacy, not an independent assessment of Amazon's market power. - U.S. House Judiciary Subcommittee on Antitrust, Commercial and Administrative Law, Investigation of Competition in Digital Markets: Majority Staff Report and Recommendations (October 6, 2020), Amazon overview and selected marketplace, seller-data, fee, private-label, acquisition and logistics passages. https://www.govinfo.gov/content/pkg/GOVPUB-Y4_J89_1-PURL-gpo145949/pdf/GOVPUB-Y4_J89_1-PURL-gpo145949.pdf Evidence note: Read in part. The majority staff report draws on documents and market participants but is not a judicial finding and was not read outside the Amazon-relevant portions. - Jeffrey P. Bezos, 2019 Letter to Shareholders (2019 reporting period; published April 2020), COVID-19 response, testing work, temporary pay, jobs and climate commitments. https://s2.q4cdn.com/299287126/files/doc_financials/2020/ar/2019-Shareholder-Letter.pdf Evidence note: Management's early-pandemic account, written while operating conditions were changing quickly. It does not independently assess worker experience or adequacy of protection. ## Jeff Bezos · Development · 2021: The final CEO letter adds an employee promise after the model is challenged Bezos's final CEO shareholder letter is the most explicit attempt to widen Amazon's value-creation story. For the 2020 reporting period, he estimates value created for shareholders, employees, third-party sellers and customers. The employee figure of $91 billion is largely compensation and benefits, so it is an economic cost to Amazon and income to workers rather than a clean measure of surplus. The customer estimate depends on survey assumptions that Bezos says have wide error. The exercise is still a development from letters that concentrated primarily on customers and owners. It tries to name multiple constituencies and confront the question of who benefited from Amazon's scale. The immediate pressure is visible. Bezos responds to the Bessemer, Alabama union election and rejects the idea that Amazon treats employees as desperate people forced into punishing work. He says 94 percent of fulfilment-centre employees surveyed would recommend Amazon to a friend as a workplace. He also says Amazon terminated fewer than 2.6 percent of employees for inability to perform their jobs and that the rate was lower during the 2020 reporting period because of COVID-19 operations. Those measures have defined populations and periods, and neither resolves the broader claims. Referral willingness can coexist with dissatisfaction. A termination percentage does not measure voluntary exit, injury or whether production expectations are fair. The letter nevertheless makes a new commitment. Amazon should become Earth's best employer and safest place to work, alongside its customer mission. Bezos reports that musculoskeletal disorders represented about 40 percent of work-related injuries and proposes algorithmic job rotation to vary physical movements. Amazon planned more than $300 million in 2021 safety spending and had about 6,200 safety professionals. These are inputs and intentions at the moment of transition to Andy Jassy. They are not outcomes. The course should not award the revision before injury, retention and worker-voice evidence changes. The NLRB record complicates the final narrative. In the first 2021 Bessemer tally, 738 votes supported the union and 1,798 opposed it, with challenged ballots insufficient to change the result. A later regional order directed a new election after finding objectionable conduct in the first process. The complete decision was not read for this course, so the chapter uses the official case and tally record only as a bounded institutional challenge. It does not infer what every Bessemer employee wanted and does not generalize from one facility. It shows that Amazon's claim of direct, positive employee relations was contested through a formal labour process at the moment Bezos elevated the employer mission. This ending should remain open. Bezos's letter records recognition, diagnosis and a promised response. It does not demonstrate completed development. The relevant before-and-after evidence belongs after his CEO tenure and under Jassy's authorship, which is outside this Bezos corpus. That authorship boundary matters. The letters for the 2021 through 2024 reporting periods are signed by Andy Jassy and append a separately identified reprint of Bezos's 1997 letter. Attributing their results to Bezos would turn succession into a founder story. The final lesson is that persistence can require expanding the objective rather than defending the old scorecard. Amazon's customer and cash systems produced extraordinary capability, and they also produced worker and platform pressures that the early doctrine did not fully govern. A serious leader should state the new promise, define outcome measures, accept an independent challenge and leave future evaluators free to say the promise failed. Recognition is development. Proof requires later results. ### Work the question What would count as evidence that Bezos's final employee commitment became an operating revision rather than a farewell aspiration? ### Compare with the guide’s reasoning Later injury rates, musculoskeletal-disorder rates, voluntary turnover, schedule and pace measures, appeal outcomes, worker surveys with disclosed methods and independent labour findings would need to improve. Spending, staffing and job-rotation plans are inputs. Because Jassy authored the later letters, the result should be evaluated as succession evidence rather than attributed automatically to Bezos. ### Sources - Jeffrey P. Bezos, 2020 Letter to Shareholders (2020 reporting period; published April 2021), Stakeholder value estimates, Bessemer discussion, safety commitments and final CEO reflection. https://s2.q4cdn.com/299287126/files/doc_financials/2021/ar/Amazon-2020-Shareholder-Letter-and-1997-Shareholder-Letter.pdf Evidence note: Bezos's final CEO letter. It responds to worker criticism and makes new employee and safety commitments while acknowledging uncertainty in some stakeholder-value estimates. - National Labor Relations Board, Amazon.com Services LLC, Case 10-RC-269250 (2021), Official case page, initial tally and election-result records. https://www.nlrb.gov/case/10-RC-269250 Evidence note: Read in part. The official case and tally pages were reviewed, but the complete docket and later regional decision were not read in full. - Jodi Kantor and David Streitfeld, Inside Amazon: Wrestling Big Ideas in a Bruising Workplace (August 15, 2015), Complete reported article, including Amazon's responses. https://www.nytimes.com/2015/08/16/technology/inside-amazon-wrestling-big-ideas-in-a-bruising-workplace.html Evidence note: Reporting based on more than 100 current and former employees. Accounts vary, but the article documents costs and contest around the culture Bezos praised. - Andy Jassy, 2021 Letter to Shareholders (2021 reporting period; published 2022), Author byline, closing signature and appended 1997-letter label. https://www.aboutamazon.com/news/company-news/2021-letter-to-shareholders Evidence note: The current-period letter is authored by Andy Jassy. The appended 1997 letter is separately authored by Jeffrey P. Bezos. Only authorship and the reprint boundary were checked for this source; no full-reading claim is made. - Andy Jassy, 2022 Letter to Shareholders (2022 reporting period; published 2023), Author byline, closing signature and appended 1997-letter label. https://www.aboutamazon.com/news/company-news/amazon-ceo-andy-jassy-2022-letter-to-shareholders Evidence note: The current-period letter is authored by Andy Jassy. The appended 1997 letter is separately authored by Jeffrey P. Bezos. Only authorship and the reprint boundary were checked for this source; no full-reading claim is made. - Andy Jassy, 2023 Letter to Shareholders (2023 reporting period; published 2024), Author byline, closing signature and appended 1997-letter label. https://www.aboutamazon.com/news/company-news/amazon-ceo-andy-jassy-2023-letter-to-shareholders Evidence note: The current-period letter is authored by Andy Jassy. The appended 1997 letter is separately authored by Jeffrey P. Bezos. Only authorship and the reprint boundary were checked for this source; no full-reading claim is made. - Andy Jassy, 2024 Letter to Shareholders (2024 reporting period; published 2025), Author byline, closing signature and appended 1997-letter label. https://www.aboutamazon.com/news/company-news/amazon-ceo-andy-jassy-2024-letter-to-shareholders Evidence note: The current-period letter is authored by Andy Jassy. The appended 1997 letter is separately authored by Jeffrey P. Bezos. Only authorship and the reprint boundary were checked for this source; no full-reading claim is made. ## Howard Marks · Development · 1968-1988. Two schools, a failed equity assignment, and the unloved bond desk Marks’s mature voice can sound as though defensive credit investing was a philosophy he selected from a menu. His own chronology is less tidy. He entered First National City Bank’s investment research department in the summer of 1968 and began full-time work in 1969, just as the bank and its peers concentrated on the Nifty Fifty. The companies were admired for good reasons, but their stock prices incorporated a belief that quality removed the need to ask what one paid. The 1973-74 collapse did more than supply an anecdote about bubbles. Marks says the disastrous performance cost him his position as director of equity research. The career redirection came before the theory that later made it look inevitable. His education had already given him a productive conflict. Wharton taught practical, qualitative finance. Chicago, where he studied from 1967 to 1969, taught new quantitative theories, including the efficient market hypothesis. He did not discard either school. He later described his work as an attempt to reconcile what Chicago said markets should do with what emotional people actually did. That tension became more useful than loyalty to either camp. Efficiency explained why easy superiority was unlikely. The Nifty Fifty showed that intelligent institutions could still turn reasonable facts into unreasonable prices. In 1978 a new chief investment officer moved him to convertibles and high yield bonds. Marks calls this his great lucky break, not an act of foresight. Low-rated bonds fit his temperament and the period. They offered a contractual upside and obvious ways to lose, so progress depended on excluding credits that would not pay. The market was also small, stigmatized and under-researched. Marks could use Chicago’s lesson in reverse. An unpopular field could remain inefficient because capable capital had not yet crowded in. He learned that risk did not make an asset uninvestable. Price, yield, structure and diversification determined whether bearing it made sense. The decisive extension came from another person. Bruce Karsh, then at SunAmerica, approached Marks with the idea for a distressed debt fund. Marks’s later explanation is wonderfully concrete. A high yield investor might buy a bond at $100, watch it collapse to $10 in bankruptcy, work through the restructuring, and receive $30. Why not organize a fund devoted to the $10-to-$30 part? The first Special Credits fund began in 1988. This was not Marks discovering every adjacent opportunity himself. It was a partnership in which Karsh supplied an idea and a more opportunistic investing engine, while Marks recognized how it fit what his group already knew. The developmental lesson is therefore not “be defensive.” Marks accumulated a set of mismatched experiences. An elite equity process failed. A forced transfer opened a stigmatized market. Two schools supplied conflicting models. A future partner saw a strategy Marks had not proposed. His eventual philosophy arose from making those accidents cohere. The cost was real. He left the prestigious equity track and spent years in securities many fiduciaries would not touch. The advantage was also period-specific. A neglected $2 billion high yield market in 1978 offered structural inefficiency that a mature trillion-dollar market cannot reproduce. ### Work the question Which parts of Marks’s early philosophy were chosen, and which were adaptations to failure, reassignment and other people’s ideas? ### Compare with the guide’s reasoning The synthesis was his, but the raw material was contingent. The Nifty Fifty loss displaced him, the 1978 reassignment put him in low-rated debt, Chicago and Wharton created a theoretical tension, and Karsh proposed distressed debt. Marks’s skill was recognizing what those events taught and building a repeatable practice around them. ### Sources - Howard Marks, Getting Lucky (2014-01-16), Complete memo, the Wharton, Chicago, 1978 high-yield reassignment, and decision-quality passages. https://www.oaktreecapital.com/docs/default-source/memos/2014-01-16-getting-lucky.pdf?sfvrsn=c4b70f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, What's It All About, Alpha? (2001-07-11), Complete memo, opening Wharton-Chicago synthesis and the alpha, beta, and efficient-market discussions. https://www.oaktreecapital.com/docs/default-source/memos/2001-07-11-whats-it-all-about-alpha.pdf?sfvrsn=13bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Bruce Karsh, and Sheldon Stone, 30 Years of Oaktree (2025-04-30), Official eight-page transcript, pp. 1-3 on formation and early funds, pp. 4-6 on partnership and operating choices, and pp. 7-8 on succession. https://www.oaktreecapital.com/insights/insight-podcast/education/30-years-of-oaktree-with-howard-marks-bruce-karsh-and-sheldon-stone Evidence note: Read in full. The founders reconstruct their partnership, early funds, culture, mistakes avoided and succession aims. ## Howard Marks · Development · 1990-1993. The memo habit turns observations into a defensive identity The first memo did not announce a finished system. In October 1990 Marks wrote that his views on managing portfolios had “come increasingly into focus” after two recent observations. A pension fund had never produced a spectacular annual rank but, by staying within a solid band for fourteen years, reached the fourth percentile over the full period. At the same time, a prominent value manager excused a terrible year by arguing that top-five-percent aspirations required willingness to finish in the bottom five percent. Marks rejected that bargain. His clients did not need a heroic year enough to accept a disastrous one. The Route to Performance converted those observations into an operating preference. He would seek modest consistent advantage and especially strong relative results in bad markets. The principle matched high yield credit, where one default can erase the income from several successful positions. It also matched his temperament. Later he would call himself more worrier than dreamer. The important point is that the memo records the idea arriving through comparison, rather than appearing fully formed in 1978. First Quarter Performance, written in April 1991, added a different mechanism. The late-1990 market had reached a negative extreme, followed by strong gains in early 1991. Marks used that reversal to develop the pendulum image. He did not infer that investors could forecast each swing. He argued for alertness to extremes, modest adjustment and resistance to the crowd. That distinction matters. Cycle awareness was initially a way to change posture at extremes, not a timing system. By 1993 The Value of Predictions pushed the anti-forecasting case further. A forecast shared by the consensus is already embedded in price. A profitable non-consensus forecast must differ from the market, prove right, and matter on an investable horizon. Forecasters also tend to publicize successes without maintaining scorecards of failures. Marks’s answer was not passivity. He proposed concentrating on inefficient markets and on company or security facts that could be known more reliably than the macro future. These memos show writing doing intellectual work. Each starts from a live irritant. An inconsistent manager, a market reversal, or the prestige of prediction gives Marks something to test. The published principle is the result of the test. The sequence also exposes a lasting tension. He wants investors to resist consensus and adjust at extremes, yet he distrusts forecasts about when consensus will reverse. That tension never disappears. It becomes the reason for small posture changes most of the time and aggressive action only in rare dislocations. Persistence belongs in the story. Marks later recalled that he wrote for roughly a decade without receiving a single response. The memos did not begin as a celebrated public franchise. They were a private discipline with little evidence that an audience cared. Continuing anyway allowed observations from one cycle to remain available when the next cycle forced revision. ### Work the question How did the first three years of memos change Marks’s decision process? ### Compare with the guide’s reasoning They connected defensive consistency, pendulum extremes and distrust of forecasts. Together they produced a rule for acting without pretending to know the future. Stay invested, study knowable securities, alter risk modestly when psychology is extreme, and preserve the written record for later comparison. ### Sources - Howard Marks, The Route to Performance (1990-10-12), Printed pp. 1-2, the fourteen-year pension-plan record and the concluding avoidance-of-losers passage. https://www.oaktreecapital.com/docs/default-source/memos/1990-10-12-the-route-to-performance.pdf?sfvrsn=33bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, First Quarter Performance (1991-04-11), Printed pp. 1-2, opening pendulum discussion and the closing three-step response to extremes. https://www.oaktreecapital.com/docs/default-source/memos/1991-04-11-first-quarter-client-performance.pdf?sfvrsn=d7bd0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, The Value of Predictions, or Where'd All This Rain Come From? (1993-02-15), Complete memo, the seven-step forecasting chain and the closing tests for whether a forecast is actionable. https://www.oaktreecapital.com/docs/default-source/memos/1993-02-15-the-value-of-predictions-or-where-39-d-all-this-rain-come-from.pdf?sfvrsn=6fbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, What Lies Ahead? (2001-10-04), Printed pp. 6-7, the “A Bear’s Eye View” response and closing discussion of caution, lost upside, and investor agility. https://www.oaktreecapital.com/docs/default-source/memos/2001-10-04-what-lies-ahead.pdf?sfvrsn=bbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, It’s not Easy (2015-09-09), Printed pp. 1-3, section “Second-Level Thinking” and the account of drafting the 2009 sample chapter. https://www.oaktreecapital.com/docs/default-source/memos/2015-09-09-its-not-easy.pdf?sfvrsn=47bb0f65_7 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Easy Money (2024-01-09), Printed p. 1, ten years without reader response; p. 14, 3.0-to-3.5-percent guess over the next five to ten years. https://www.oaktreecapital.com/insights/memo/easy-money Evidence note: Read directly in the official Oaktree complete collection or official memo page. ## Howard Marks · Development · 1994-2005. Founding Oaktree without a growth plan, then learning to limit success Oaktree’s origin was informal before it became principled. At a dinner in June 1994, the Marks and Karsh families discussed how appealing it would be to start a firm in Montecito. Karsh asked Marks to return with proposed economics on Monday. They agreed within half an hour. Marks, Karsh, Sheldon Stone, Larry Keele and Richard Masson had already worked together for an average of about nine years, so the new firm did not need to invent a philosophy from scratch. They wrote down six investment tenets and business principles that described what had worked for them. The continuity was real, but it can obscure how uncertain the undertaking was. How the Game Should Be Played, written weeks after Oaktree’s 1995 formation, supplied the public investment logic. Marks contrasted baseball’s home-run pursuit with a steadier game. Inefficient markets offered a place to work, but recognizing one’s limitations mattered as much as spotting opportunity. Avoiding strikeouts could compound into a superior record. The firm’s commercial plan was much thinner. In the founders’ 2025 retelling, Marks and Karsh said they had no profit plan, budget or forecast. They assumed that continuing to serve clients well would create a successful firm. Oaktree began with $10 million of capital and was cash-flow positive in its first year. That absence of a growth plan did not mean growth was easy or automatic. The founders had to build accounting and operations that their former employer had supplied. They deliberately kept investment teams in Los Angeles and New York rather than disrupting people’s lives. They treated shared values and complementary skills as infrastructure. A balanced life and refusal to hire people they did not want to work with constrained the usual route of maximizing near-term output. By Oaktree at Ten in April 2005, those choices had produced a much larger institution. Marks reported roughly $28 billion under management and about 300 employees. More revealing than the scale was the capacity discipline. Oaktree said it had turned away $14 billion of high yield mandates since November 1998 because additional capital could weaken results. Marks also noted that every one of the firm’s 24 closed-end funds had been profitable, with reported net internal rates of return ranging from 4 percent to 49 percent over their respective periods. Those are firm supplied figures, not an independent audit, but they show the feedback the founders were receiving. Success created a new problem. A neglected niche becomes less attractive when money recognizes it. An organization built to exploit scarcity can impair its own edge by gathering too much capital. Oaktree’s answer was to ration growth, decentralize investment authority and treat reputation as an asset. That choice had a cost. Competitors could accumulate more assets in buoyant markets, and clients could prefer managers willing to deploy faster. The 2025 founders’ discussion also acknowledges that the old prohibition on “mission creep” no longer fits a world in which the firm must evolve across markets. The durable lesson is not that the 1995 principles never changed. The six investment tenets remained word-for-word stable, while the business built around them had to expand, professionalize and plan for succession. Marks’s development lies in separating a stable decision ethic from the institutional forms used to express it. ### Work the question What did Oaktree preserve as it grew, and what did it have to revise? ### Compare with the guide’s reasoning It preserved risk control, bottom-up work, limited reliance on forecasts and a partnership culture. It revised the scale, operational apparatus, product range and leadership structure. Capacity limits show that preserving an investment method sometimes requires refusing otherwise attractive business growth. ### Sources - Howard Marks, How the Game Should Be Played (1995-05-26), Complete memo, the baseball analogy and the closing discussion of Oaktree’s game plan. https://www.oaktreecapital.com/docs/default-source/memos/1995-05-26-how-the-game-should-be-played.pdf?sfvrsn=d3bd0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Are You An Investor or a Speculator? (1997-09-03), Complete memo, passages defining investment versus speculation and applying the distinction to technology stocks. https://www.oaktreecapital.com/docs/default-source/memos/are-you-an-investor-or-a-speculator.pdf?sfvrsn=1e37cf65_2 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Oaktree at Ten (2005-04-11), Printed pp. 1-5, sections “Priorities,” “Clients,” “Performance,” “Growth,” and “Plans for the Future”. https://www.oaktreecapital.com/docs/default-source/memos/2005-04-11-oaktree-at-ten.pdf?sfvrsn=5fbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Bruce Karsh, and Sheldon Stone, 30 Years of Oaktree (2025-04-30), Official eight-page transcript, pp. 1-3 on formation and early funds, pp. 4-6 on partnership and operating choices, and pp. 7-8 on succession. https://www.oaktreecapital.com/insights/insight-podcast/education/30-years-of-oaktree-with-howard-marks-bruce-karsh-and-sheldon-stone Evidence note: Read in full. The founders reconstruct their partnership, early funds, culture, mistakes avoided and succession aims. ## Howard Marks · Development · 1997-2006. Being early, being wrong for a while, and turning the memos into a book Marks’s defensive temperament generated correct warnings and uncomfortable timing. In September 1997, Are You an Investor or a Speculator? argued that the stock market was not undervalued. He revisited the Nifty Fifty and contrasted investment based on value with speculation based on what someone else might pay. Yet he also stressed that no bell announces an extreme. The technology boom continued for more than two years. An investor who translated his caution into a large short or an exit from equities could have suffered badly before being vindicated. The January 2000 bubble.com memo is a useful admission. Marks said his recent caution had been wrong. He did not withdraw the valuation concern. He separated three propositions that euphoric markets tempt people to merge. A technology can transform the world. Its companies can become important. Their securities can still be priced too high. Reading Edward Chancellor’s history of speculation helped him recognize the rhyme between the South Sea episode and the technology, media and telecommunications boom. He published the memo while acknowledging that calling prices high did not reveal the next move. The timing happened to be excellent. The technology bubble broke soon afterward, and readers finally responded. Marks later said he had written for ten years without a single reply. The attention prompted by bubble.com encouraged the long memo sequence that followed. The persistence preceded the reputation. It also creates a selection problem. A timely warning is remembered more vividly than years of cautious language that did not lead quickly to a reversal. The Most Important Thing began as a July 2003 memo that gathered recurrent phrases Marks noticed himself using with clients. The eventual 2011 book was not simply that memo bound between covers. In It’s Not Easy, Marks explains that he expected to write a book after retirement, then accelerated when Warren Buffett offered in 2009 to supply a blurb. Columbia Business School Publishing asked for a sample chapter. Sitting down to produce it, Marks named and developed “second-level thinking,” a concept he says had not appeared explicitly in the memos. The episode shows codification changing the philosophy. A publishing demand created a label and an organizing device that later readers may mistake for an idea that guided every earlier decision. This period therefore contains two kinds of pressure. Markets made caution look foolish before the reversal. Publishing required tacit habits to become teachable concepts. Marks handled the first by refusing to equate overvaluation with an imminent decline. He handled the second by revisiting his archive and accepting that clear rules could be useful without becoming mechanical formulas. The cost of being early is central. A manager must retain clients, confidence and liquidity during the interval between a reasonable diagnosis and a market response. Marks could tolerate that interval partly because Oaktree was not making large directional equity bets. His warnings often governed capacity, credit selection and risk posture rather than a single all-or-nothing trade. The institutional design made his intellectual temperament survivable. ### Work the question Why is bubble.com weaker evidence of forecasting skill than it first appears, yet still valuable? ### Compare with the guide’s reasoning Its timing was unusually good, but Marks had been cautious earlier and explicitly said high prices did not predict the next move. Its value lies in distinguishing technological truth from investment price and in preserving a disciplined warning without turning it into a timed market bet. ### Sources - Howard Marks, Are You An Investor or a Speculator? (1997-09-03), Complete memo, passages defining investment versus speculation and applying the distinction to technology stocks. https://www.oaktreecapital.com/docs/default-source/memos/are-you-an-investor-or-a-speculator.pdf?sfvrsn=1e37cf65_2 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, bubble.com (2000-01-02), Complete memo, the opening bubble diagnosis and the numbered discussion of what happened, why, and what follows. https://www.oaktreecapital.com/docs/default-source/memos/2000-01-02-bubble.pdf?sfvrsn=37bc0f65_5 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Irrational Exuberance (2000-05-01), Complete memo, the Robertson, Soros-Druckenmiller, and Brinson cases and the closing discussion of pressure while early. https://www.oaktreecapital.com/docs/default-source/memos/2000-05-01-irrational-exuberance.pdf?sfvrsn=cfbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, We're Not In 1999 Anymore, Toto (2000-12-31), Complete memo, opening correction retrospective and closing valuation and prospective-return discussion. https://www.oaktreecapital.com/docs/default-source/memos/2000-12-31-were-not-in-1999-anymore.pdf?sfvrsn=3bbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, What's It All About, Alpha? (2001-07-11), Complete memo, opening Wharton-Chicago synthesis and the alpha, beta, and efficient-market discussions. https://www.oaktreecapital.com/docs/default-source/memos/2001-07-11-whats-it-all-about-alpha.pdf?sfvrsn=13bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, You Can't Predict. You Can Prepare. (2001-11-20), Complete memo, discussions of staying power, cash, early caution, and preparing for unknowable cycles. https://www.oaktreecapital.com/docs/default-source/memos/2001-11-20-you-cant-predict-you-can-prepare.pdf?sfvrsn=bc00f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, It’s not Easy (2015-09-09), Printed pp. 1-3, section “Second-Level Thinking” and the account of drafting the 2009 sample chapter. https://www.oaktreecapital.com/docs/default-source/memos/2015-09-09-its-not-easy.pdf?sfvrsn=47bb0f65_7 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ## Howard Marks · Development · 2007. Four memos record the difference between diagnosis and timing The 2007 sequence is more revealing than a single famous warning. In February’s The Race to the Bottom, Marks described lenders competing away their own protections. Mortgage underwriting weakened, leverage rose, covenants disappeared and financial engineering made risk look transferable. He could see the credit system becoming fragile. He could not identify when that fragility would matter. Oaktree’s response was to hold higher standards and accept looking old-fashioned while transactions continued. It’s All Good on July 16 described the broadest and strongest upswing Marks had seen. Investors reached for return because safe yields were low. Financing became abundant, risk premiums shrank and leverage made weak propositions saleable. He warned that a long benign period could not continue forever, but also wrote that one could not know whether a small disturbance marked the turn. This is a diagnosis of vulnerability, not a dated crash call. Only two weeks later, It’s All Good . . . Really? revised the picture. Credit markets had abruptly weakened. Marks candidly said that when he wrote the prior memo, he did not expect the cycle to be curtailed before July ended. The change taught a specific lesson. The catalyst that ends an excessive market is usually not forecastable. He still resisted declaring that the great turn had arrived. His formulation was closer to “this is more like it” than “this is it.” By September, Now It’s All Bad? recorded the pendulum moving toward fear. The prior five years of easy money had not disappeared from the causal story, but price declines and tighter credit created a different decision problem. If investors extrapolated the new pessimism as mechanically as they had extrapolated optimism, they could repeat the error in reverse. Marks expected economic weakness but refused to make the portfolio depend on certainty about it. This sequence demonstrates revision under pressure. The stable principle was that lenders must be paid for risk and protected against adverse outcomes. The situational judgment changed three times. February emphasized declining standards. July 16 emphasized a dangerous but continuing boom. July 30 recognized that the break could be underway earlier than expected. September asked whether fear had already overshot. A polished retrospective can compress all four into “we saw the crisis coming.” The originals show something more useful. Marks saw fragility, missed the timing, updated promptly and kept the magnitude of his claims smaller than the magnitude of the possible event. There was an institutional consequence. In his 2023 retrospective, Marks says Oaktree raised an $11 billion reserve fund for distressed opportunities between January 2007 and March 2008. That amount and period are not claims drawn from the 2007 originals. The earlier memos document the caution behind the preparation, while the retrospective also reports that Oaktree sold assets, liquidated mature funds and kept some strategies small. The preparation did not require knowing the exact catalyst. It required protecting the ability to act if capital markets closed. This is the practical distinction behind the 2001 phrase “you can’t predict, you can prepare.” Preparation buys time for revision. A forecast demands that reality arrive on schedule. The sacrifice was visible in the late boom. Refusing weak deals meant lost fees and relative underperformance while risk taking still paid. That is why disciplined credit is not merely an analytical stance. The firm has to withstand clients asking why competitors are growing faster. ### Work the question What does the full 2007 memo sequence show that a single crisis warning cannot? ### Compare with the guide’s reasoning It separates structural diagnosis from timing. Marks identified weakening standards, did not expect the turn when it arrived, revised within two weeks, and then questioned whether pessimism was becoming excessive. Preparation and institutional patience made those updates possible without a forced binary bet. ### Sources - Howard Marks, You Can't Predict. You Can Prepare. (2001-11-20), Complete memo, discussions of staying power, cash, early caution, and preparing for unknowable cycles. https://www.oaktreecapital.com/docs/default-source/memos/2001-11-20-you-cant-predict-you-can-prepare.pdf?sfvrsn=bc00f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, The Race to the Bottom (2007-02-14), Complete memo, opening covenant and leverage evidence and closing “race to the bottom” diagnosis. https://www.oaktreecapital.com/docs/default-source/memos/2007-02-14-the-race-to-the-bottom.pdf?sfvrsn=9bbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, It’s All Good (2007-07-16), Complete memo, passages on the flattened risk-return line, thin risk premiums, and reduced reserve capacity. https://www.oaktreecapital.com/docs/default-source/memos/2007-07-16-its-all-good.pdf?sfvrsn=8fbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, It’s All Good . . . Really? (2007-07-30), Complete memo, opening admission that the credit turn arrived sooner than expected and the revised assessment that follows. https://www.oaktreecapital.com/docs/default-source/memos/2007-07-30-its-all-good-really.pdf?sfvrsn=97bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Now It’s All Bad? (2007-09-10), Complete memo, opening reassessment of the credit reversal and closing discussion of whether pessimism had gone too far. https://www.oaktreecapital.com/docs/default-source/memos/2007-09-10-now-its-all-bad.pdf?sfvrsn=9fbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Taking the Temperature (2023-07-10), Printed p. 1, five-memo selection rule; pp. 3-4, reserve-fund deployment and $6 billion/$7.5 billion figures; pp. 8-11, method and limits. https://www.oaktreecapital.com/insights/memo/taking-the-temperature Evidence note: Read directly in the official Oaktree complete collection or official memo page. ## Howard Marks · Development · 2008-2009. Fear, forced buying decisions, and an admission that the crisis was different The Global Financial Crisis forced Marks to move from warning to capital deployment. In March 2008, The Tide Goes Out linked leverage, mark-to-market accounting, short-term funding and forced selling into what he called a new kind of crisis, different in degree and kind from prior credit downturns. He admitted that the integrated view was new to him. He also refused apocalyptic speculation. There was no reliable roadmap, and imagining every possible failure would not tell Oaktree what to buy. Lehman’s September bankruptcy made the uncertainty operational. In Nobody Knows, Marks reduced the decision to a harsh asymmetry. If the financial world ended, portfolio decisions would scarcely matter. If it survived and Oaktree did not buy cheap claims on viable businesses, the firm would have failed at its purpose. That was not a prediction that rescue was certain. It was a choice of the only assumption under which action remained meaningful. The emotional evidence is unusually specific. In later accounts Marks recalls Karsh alternating between concern that the team was moving too slowly and concern that it was moving too fast. The oscillation did not stop purchases. Oaktree invested an average of roughly $400 million to $450 million per week during the final fifteen weeks of 2008, depending on the later source and scope used. The different figures should not be silently reconciled. The 2023 retrospective says Karsh’s group invested $6 billion from September 18 through year-end and Oaktree overall invested $7.5 billion. Those are firm reported period figures. The Limits to Negativism on October 15 records the intellectual pivot. Marks had long treated skepticism as protection from optimistic claims. A meeting with a client who could imagine an even worse outcome after every stress case made him see that skepticism must also challenge excessive pessimism. Oaktree’s purchase list had become long and its sales nearly nonexistent. The emotional admission matters because it avoids the fantasy of calm omniscience. He called those weeks the greatest panic he had witnessed and described the insight as an epiphany. He did not claim certainty that prices had bottomed. January 2009’s The Long View supplies the revision that triumphant crisis stories often omit. Marks said he had correctly observed the shorter credit cycle but failed to appreciate the larger picture. He had treated 2003-07 as a familiar cycle, only more extreme, when the build-up of leverage and financial-sector fragility made it different in kind. The crisis expanded his model from recurring short cycles toward longer secular movements in leverage, rates and institutional behavior. This was not a painless demonstration of principles. In Volatility + Leverage = Dynamite in December 2008, Marks disclosed that three Oaktree evergreen funds had borrowed to buy bank loans and another had borrowed to buy low-priced Japanese small-cap stocks. Their underlying companies were generally performing, but collapsing loan and equity prices hurt the funds. Marks apologized if he sounded holier than thou and said Oaktree was not perfect. The admission is stronger than a generic warning about other people’s leverage. Even a firm organized around risk control had combined borrowed money with assets whose observed price history understated the possible drawdown. Prices also continued to fall after purchases began. The team had to average down while counterparties and clients questioned the premise. Preparation supplied capital, but it did not remove doubt or protect every existing strategy. It made survival and continued buying possible while the firm absorbed losses elsewhere. The developmental result was a more demanding version of humility. Knowing that forecasts fail is not enough. Investors must still choose a pace, protect liquidity and update their model when the current crisis defeats historical analogy. Marks’s strongest crisis work joins doubt to action rather than using doubt as permission to wait. ### Work the question What did Marks learn in 2008 that was not already contained in “be fearful when others are greedy”? ### Compare with the guide’s reasoning He learned that skepticism must operate in both directions, that action can rest on a survivability assumption rather than a forecast, and that a familiar short-cycle model can miss a longer structural change. He also learned that doubt persists during good decisions and must be managed through pace and liquidity. ### Sources - Howard Marks, The Tide Goes Out (2008-03-18), Leveraging and Inflating; Mark-to-Market Accounting; Should Does Not Equal Will; Leverage and Risk. https://www.oaktreecapital.com/insights/memo/the-tide-goes-out Evidence note: Contemporaneous analysis written before the worst phase of the financial crisis. Examples and conclusions express the author's assessment and are not an independent performance study. - Howard Marks, Nobody Knows (2008-09-19), Does the Market Know?; The End of the Financial System; final bear-market-stage discussion. https://www.oaktreecapital.com/insights/memo/nobody-knows Evidence note: Contemporaneous memo after Lehman's bankruptcy. It records a decision premise and a search for bargains, but no security-level trade list. - Howard Marks, The Limits to Negativism (2008-10-15), Final discussion of third-stage pessimism and Oaktree purchases. https://www.oaktreecapital.com/insights/memo/the-limits-to-negativism Evidence note: Contemporaneous practitioner report. It describes purchase direction and averaging down, but not positions, allocations, or later investment results. - Howard Marks, Volatility + Leverage = Dynamite (2008-12-17), Printed pp. 10-12, “Are You Tall Enough to Use Leverage?” and Oaktree’s disclosure of four affected evergreen funds. https://www.oaktreecapital.com/docs/default-source/memos/2008-12-17-volatility-leverage-dynamite.pdf?sfvrsn=c7bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, The Long View (2009-01-09), Complete memo, “The Short View” and “The Long View” passages distinguishing the foreseen credit cycle from the missed systemic mechanism. https://www.oaktreecapital.com/docs/default-source/memos/2009-01-09-the-long-view.pdf?sfvrsn=c3bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Taking the Temperature (2023-07-10), Printed p. 1, five-memo selection rule; pp. 3-4, reserve-fund deployment and $6 billion/$7.5 billion figures; pp. 8-11, method and limits. https://www.oaktreecapital.com/insights/memo/taking-the-temperature Evidence note: Read directly in the official Oaktree complete collection or official memo page. ## Howard Marks · Development · 2009-2018. Books codify the memos, while collaborators expose what the labels hide The title The Most Important Thing predates the book. In July 2003 Marks used it for a thirteen-page memo that gathered eighteen precepts, from price and intrinsic value through capacity, leverage, candor, personnel and partnership. The memo was already an effort to turn recurring judgments into an institutional creed. It also preserved conflicts rather than hiding them. A manager should state a clear mandate yet retain tactical flexibility, hold convictions strongly yet stop short of hubris, and restrict capital even when performance attracts more of it. The Most Important Thing appeared as a book in May 2011 after the crisis gave Marks’s archive a wider audience. The publisher describes it as a concentration of the memos, but Marks’s later account shows that the act of concentration created new structure. “Second-level thinking” emerged when Columbia requested a sample chapter. The book’s twenty chapters turned recurring judgments about price, risk, cycles, contrarianism and luck into a sequence a reader could retain. That clarity is useful, but it can make development look more linear than the source memos were. The 2013 Illuminated edition made revision visible on the page. Christopher Davis, Joel Greenblatt, Paul Johnson and Seth Klarman added comments and counterpoints. Marks added annotations and a chapter on reasonable expectations. The publisher page confirms that it was more than a reprint. It placed the philosophy in dialogue with investors who could qualify it. The full edition was not available locally for this study, so its exact arguments remain a gap. The honest conclusion is limited to the publisher’s description, contents and Marks’s later references to particular commentary. Mastering the Market Cycle in 2018 developed another idea that had been present but incomplete in early memos. The 1991 pendulum described recurrent movement around extremes. While writing the cycle book, Marks says he reached a deeper question about two-thirds of the way through. Why do cycles exist at all? His answer became “excesses and corrections.” Events do not merely follow in an up-down pattern. Easy money changes behavior, behavior builds fragility, losses change psychology, and the correction can overshoot in the other direction. Causality became more important than periodicity. An original 2022 interview at Panmure House also reveals two limits hidden by the final title. Nick Train objected that a pendulum is mechanical, regular and predictable, while market movement is none of those things. Marks defended the word by treating it as a mood swing rather than a physical clock. He also disclosed that his preferred working title had been Listening to the Cycle. The publisher believed Mastering the Market Cycle would sell more copies. The published title therefore promises more control than Marks’s own formulation. His method is to listen for conditions and alter odds, not to command the cycle. Marks’s son Andrew again served as a corrective. During work on the book in 2017, Marks remarked that his market calls had been about right. Andrew replied that this was because he had made only five in fifty years. Marks accepted the criticism. Rare calls at extreme prices are compatible with distrust of routine market timing. A long list of calls from ordinary conditions would not be. The books should therefore be read as revisions and teaching instruments, not as timeless transcripts of an unchanged mind. The original book was pulled forward by Buffett’s offer, shaped by a publisher’s sample request and organized with a concept named for that purpose. The annotated edition invited counterpoint. The cycle book found its causal thesis during drafting, received a more assertive commercial title and was disciplined by Andrew’s challenge. Authorship was still Marks’s, but partnership and publication pressure materially improved the ideas. ### Work the question How should a reader use Marks’s books without mistaking codification for original chronology? ### Compare with the guide’s reasoning Use the books to see the mature architecture, then return to dated memos and interviews to see when labels and causal claims appeared. The publisher pages establish the editions, while Marks’s accounts show that second-level thinking was named in 2009 and the cycle book’s causal thesis arrived during drafting. ### Sources - Howard Marks, The Most Important Thing (2003-07-01), Printed pp. 1-13, eighteen “most important thing” passages and closing account of the resulting Oaktree creed. https://www.oaktreecapital.com/docs/default-source/memos/2003-07-01-the-most-important-thing.pdf?sfvrsn=91c00f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, It’s not Easy (2015-09-09), Printed pp. 1-3, section “Second-Level Thinking” and the account of drafting the 2009 sample chapter. https://www.oaktreecapital.com/docs/default-source/memos/2015-09-09-its-not-easy.pdf?sfvrsn=47bb0f65_7 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Conversation at Panmure House (2022-06-23), Complete interview transcript, discussion of cycle causality and the “Listening to the Cycle” working title. https://www.oaktreecapital.com/insights/memo/conversation-at-panmure-house Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, The Most Important Thing: Uncommon Sense for the Thoughtful Investor (2011-05-01), Publisher page, description, contents, and listed Chapter 2 excerpt. https://cup.columbia.edu/book/the-most-important-thing/9780231153683/ Evidence note: Publisher record inspected. The full book was not locally available and was not read. - Howard Marks, The Most Important Thing Illuminated (2013-01-01), Publisher page, description and contents. https://cup.columbia.edu/book/the-most-important-thing-illuminated/9780231162845/ Evidence note: Publisher record inspected. The full annotated edition was not locally available and was not read. - Howard Marks, Mastering the Market Cycle: Getting the Odds on Your Side (2018-10-02), Publisher page. https://www.hachette.co.uk/titles/howard-marks/mastering-the-market-cycle/9781473695689/ Evidence note: Publisher record inspected. The full book was not locally available and was not read. ## Howard Marks · Development · 2020. Six weeks of writing turn radical uncertainty into a method The pandemic produced an unusually dense record of Marks updating in public, and the sequence begins with a miss. Nobody Knows II on March 3 separated facts, informed inferences and guesses. Marks then made a concrete guess that coronavirus would resemble another seasonal disease and was unlikely to change life and business fundamentally. That judgment proved wrong. Latest Update on March 19 registered far more severe health and economic consequences and still argued for incremental buying, without claiming the market had reached its bottom. The two documents show uncertainty language did not prevent a bad forecast. They also preserve how quickly he revised it. Which Way Now? on March 31 catalogued a positive and negative case after the S&P 500 had fallen 33.9 percent from February 19 to March 23 and then gained 17.5 percent in three days. Marks disclosed his bias directly. He was more worrier than dreamer and had probably been more defensive than necessary on average. His negative list was longer, yet he did not treat length as probability. He thought the rebound gave too little weight to worsening news and expected prices could decline, while telling readers to be ready to buy if they did. Calibrating on April 6 turned the shifting views into a posture decision. Marks said his March conclusions had waxed and waned with prices and fully expected further amendment. Prior caution had served its purpose, but recognized risk, lower prices and chastened investors meant defense no longer deserved its prior overweight. He moved toward neutral or offense while retaining the possibility of lower markets. The method was gradual adjustment under acknowledged uncertainty, not confidence that the latest view was final. Uncertainty on May 11 moved from scenarios to epistemology. Four interacting systems mattered at once. The virus, economic shutdown, oil shock and government rescue lacked a useful historical joint distribution. Marks examined confirmation bias in himself. Negative reports fit his wary view easily, while he generated counterarguments against positive evidence. That admission makes intellectual humility concrete. It is not modest vocabulary attached to a settled conclusion. It is active suspicion of the way one filters evidence. Uncertainty II arrived seventeen days later because Marks found more to add. He developed path dependence. The future does not sit fully formed waiting for a gifted forecaster to discover it. Decisions by citizens, firms, scientists and governments create it. He also extended the expertise problem. A layperson may lack both expertise and the ability to identify the true expert. The memo ends with a practical concession. Investors cannot predict tail events from past frequency, but they can reduce vulnerability when behavior and leverage make markets fragile. The Anatomy of a Rally on June 18 confronted evidence that challenged his March caution. From the March low to June 8 the S&P 500 had risen almost 45 percent. Marks did not erase his earlier view. He distinguished “overpriced” from “about to fall,” noted that excellent investors had also been humbled, and concluded that liquidity and optimistic expectations had moved faster than fundamentals. The market could continue upward even if that valuation judgment was reasonable. Coming into Focus in October revised the crisis analogy itself. Marks argued that this was not a conventional cycle caused by economic excess. An exogenous health event caused an intentional shutdown, and economic tools could support incomes without curing the disease. The Fed and Treasury also prevented the credit freeze, forced liquidations and lasting investor pain that had created bargains in 1990, 2001 and 2008. A crisis had arrived without producing the usual distressed opportunity set. Oaktree had been prepared, but the policy rescue reopened capital markets too quickly for a long buying window. That outcome pressured the mature playbook. The firm’s caution helped during the first-quarter drawdown, but abundant public support meant competitors also retained money and nerve. Marks returned to defense because expected returns were historically low relative to uncertainty. He did not claim the old cycle would repeat. The writing sequence shows why repeated memos matter. Each new piece preserves what was known at that date, records what surprised him and updates the mechanism without pretending the update was available earlier. ### Work the question What changed across Marks’s 2020 memos? ### Compare with the guide’s reasoning He moved from a wrong seasonal-disease guess to recognizing a severe shock, recalibrating from defense toward offense, auditing his bias, accepting path dependence, explaining a rally that contradicted his caution, and recognizing that policy support removed the usual distressed opportunity. ### Sources - Howard Marks, Nobody Knows II (2020-03-03), Printed pp. 1-7, facts/inferences/guesses framework; p. 6, the seasonal-disease guess; p. 7, partial buying response. https://www.oaktreecapital.com/docs/default-source/memos/nobody-knows-ii.pdf?sfvrsn=108eb165_8 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Latest Update (2020-03-19), Printed pp. 1-6, worsening health and economic update and concluding case for incremental buying rather than calling the bottom. https://www.oaktreecapital.com/docs/default-source/memos/weekly.pdf?sfvrsn=cbf3b065_8 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Which Way Now? (2020-03-31), The Positive Case; the negative case; The Government Programs; Summing Up. https://www.oaktreecapital.com/insights/memo/which-way-now Evidence note: Dated scenario analysis and author judgment during the pandemic shock. It is not an Oaktree transaction ledger. - Howard Marks, Calibrating (2020-04-06), Printed pp. 1-8, the audit of March views; pp. 5-8, move from overweight defense toward offense and buy-on-the-way-down discussion. https://www.oaktreecapital.com/docs/default-source/memos/calibrating.pdf?sfvrsn=3e98b665_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Uncertainty (2020-05-11), All We Don't Know; In Praise of Doubt; confidence and position-sizing discussion. https://www.oaktreecapital.com/insights/memo/uncertainty Evidence note: The discussion of 2008 includes a retrospective practitioner report. The memo distinguishes evidence-based confidence from certainty. - Howard Marks, Uncertainty II (2020-05-28), Complete memo, probability-tree discussion and concluding limits on conviction under pandemic uncertainty. https://www.oaktreecapital.com/docs/default-source/memos/uncertainty-ii.pdf?sfvrsn=d48a65_8 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, The Anatomy of a Rally (2020-06-18), Market chronology; psychology of the rally; final assessment questions. https://www.oaktreecapital.com/insights/memo/the-anatomy-of-a-rally Evidence note: Contemporaneous market assessment after the spring rally. The judgment is not evidence of later excess returns. - Howard Marks, Coming into Focus (2020-10-13), The Power of Interest Rates; crisis comparison; final portfolio-calibration discussion. https://www.oaktreecapital.com/insights/memo/coming-into-focus Evidence note: Retrospective report on Oaktree's pre-crisis posture and March buying, followed by a dated October assessment. It does not disclose portfolio allocations. ## Howard Marks · Development · 2021-2022. Andrew Marks forces a late-career revision on growth, crypto and selling Something of Value is the clearest evidence that Marks’s philosophy remained revisable at age seventy-five. During the pandemic, his son Andrew and family lived with Howard and Nancy Marks for ten weeks. Andrew invested in growth and technology companies, areas Howard had long treated with caution and limited interest. Their sustained argument made Howard identify where his credit background had become a bias rather than an advantage. Marks traced the bias honestly. Parents shaped by the Depression taught him to save for a rainy day. The Nifty Fifty collapse punished high valuations. Fixed income rewarded finding what could go wrong. Technology felt over his head and rarely appeared in the bonds he managed. Those experiences produced a powerful risk filter. They also made mean reversion feel universal and encouraged him to sell as prices rose. Andrew argued that a rare company can keep expanding its opportunity set, so a fixed target price and routine profit taking can destroy the benefits of compounding. Marks accepted several uncomfortable revisions. Low valuation measures are not the same as value. A high multiple is not proof of overpricing. Intangible assets, management quality and future optionality can matter even when they cannot be measured precisely. A stable legacy company can be less predictable than a growing software company if disruption destroys its apparent moat. Most personally, Marks acknowledged that his instinct to “scale out” of winners could be a life-altering mistake. He even imagined that his limitations might have kept him from becoming a venture investor or backing a company such as Amazon. Cryptocurrency made the revision testable before the 2021 family debate. In July 2017 Marks had expressed strong skepticism. Criticism and subsequent conversations led him to write Yet Again? in September. He said he had framed Bitcoin the wrong way and conceded it could function as a currency or medium of exchange, while still declining to invest and retaining doubts about value and price. This was a change in framing rather than an endorsement. Andrew later pressed the issue further. He owned Bitcoin for the family and challenged his father’s confidence. By 2021 Marks still did not declare crypto sound. He said his earlier skeptical view had not worked to date and that he was not informed enough to hold a firm opinion. Curiosity had to accompany skepticism. The sequence matters because it shows revision beginning through public criticism and conversation in 2017, then deepening through Andrew’s challenge in 2021. His most characteristic trait did not automatically deserve authority in a field he had not studied closely. The revision did not turn him into a growth evangelist. He retained price discipline, warned that exuberant markets treat every entrant as a future winner and emphasized how few companies can justify long growth runways. What changed was the burden of proof. Novelty could no longer be dismissed merely because old metrics struggled to value it. In the 2022 Panmure conversation, Marks generalized Andrew’s challenge. If quantitative information about the present is readily available to everyone, it cannot reliably produce superior results. Advantage must come from better qualitative judgment or a better view of the future. In I Beg to Differ, he applied the same logic to unconventionality. Different action is necessary for different results, but it carries the risk of being wrong and looking wrong for a long time. The cautious investor must decide whether avoiding embarrassment also prevents achievement. This chapter matters because it prevents a false picture of humility as a fixed rule. Marks’s earlier experiences gave him good reasons to distrust stories, distant cash flows and rising prices. New evidence revealed that the same pattern recognition could become knee-jerk dismissal. His response was neither surrender nor defense. He narrowed the claim, named the limitation and allowed a partner with different skills to change his framework. ### Work the question What exactly did Andrew Marks change in his father’s framework? ### Compare with the guide’s reasoning He changed the scope of mean reversion and skepticism. Howard retained valuation and risk control, but accepted that exceptional compounders may not fit fixed target prices, that high multiples can coexist with value, and that he lacked enough knowledge to dismiss crypto. ### Sources - Howard Marks, Yet Again? (2017-09-07), Printed pp. 4-8, revised Bitcoin framing after conversations; pp. 9-11, response to criticism that the July memo lacked prescriptions. https://www.oaktreecapital.com/docs/default-source/memos/yet-again.pdf?sfvrsn=3767f765_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Something of Value (2021-01-11), Complete memo and appendix, sections on Marks’s personal value-investing journey and the dialogue with Andrew Marks. https://www.oaktreecapital.com/insights/memo/something-of-value Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Conversation at Panmure House (2022-06-23), Complete interview transcript, discussion of cycle causality and the “Listening to the Cycle” working title. https://www.oaktreecapital.com/insights/memo/conversation-at-panmure-house Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, I Beg to Differ (2022-07-26), Complete memo, sections on contrarianism, selling, and the Something of Value retrospective. https://www.oaktreecapital.com/insights/memo/i-beg-to-differ Evidence note: Read directly in the official Oaktree complete collection or official memo page. ## Howard Marks · Development · 2004-2023. Testing the “five calls” story against the contemporaneous record Taking the Temperature begins with an explicit selection rule. Preparing for a Financial Times interview, Marks chose five memos from 2000 to 2020 because they represented important turning points and because the calls were right. That is a useful teaching set. It is not a forecast scorecard. Marks tells the reader how survivorship entered the sample. His own caveat is strong. Andrew observed that the calls looked good because Marks had made only five in fifty years. Marks accepted the point and said rare extremes permit action with better odds than ordinary markets. The memo also admits that the first call appeared after he had already spent more than thirty years in the industry. In 1979 he did not recognize the error in the famous “Death of Equities” argument. Experience and scars, not book learning alone, eventually gave him the pattern. The pre-crisis trail also began earlier than the best-known 2007 warning. Risk and Return Today in October 2004 described a low and unusually flat capital-market line. With investors fleeing low safe yields, risk aversion had fallen, spreads had narrowed, and alternative markets were crowded. Marks said likely errors had shifted from omission to commission and argued for caution, smaller funds and lower expectations. This was a useful environmental diagnosis. It was also very early. The credit expansion continued for years, making the commercial and psychological cost of maintaining the stance part of the record. The contemporaneous 2007 memos resist an even neater retrospective. The Race to the Bottom correctly diagnosed collapsing lender discipline. It’s All Good made the warning more forceful. Yet the July 30 follow-up says Marks had not expected the cycle to be curtailed by the end of that month. He would not declare the turn complete. By September he was already asking whether negativity had moved too far. These were strong observations joined to repeated uncertainty about timing and magnitude. The Long View in January 2009 adds a more serious qualification. Marks said he had seen the short credit cycle but failed to appreciate the big picture. He had understood 2003-07 as similar to past episodes, only more extreme, and missed the degree to which leverage and the financial system created a different kind of crisis. The later “five calls” story is accurate about selected posture decisions. It should not be expanded into a claim that Marks foresaw the crisis mechanism in full. There They Go Again... Again in July 2017 supplies counterevidence from the long recovery. Marks wrote that his caution since 2011 had not been right and that describing a six-year-early warning as correct might never become defensible. He could not disprove the simpler explanation that a born worrier would eventually look right after repeating caution for long enough. This admission changes the denominator. On Marks’s account, the defensive posture that helped before 2008 also brought years of underparticipation while markets continued upward. The public record does not provide a complete counterfactual portfolio with which to measure the cost. Lines in the Sand in April 2017 adds operating pressure to the record. Oaktree had begun using subscription lines after clients requested them and the practice became common. Marks then raised concerns about how the financing could flatter fund internal rates of return, obscure dollars earned and add liquidity risk. Oaktree started an internal process to create guidelines. The memo records participation followed by reconsideration, rather than a firm warning from outside the practice. It does not document the final rules or prove that the process solved the risks. There is also a denominator problem. The official archive contains 165 dated public memos through April 9, 2026. The five-call narrative selects a tiny subset of a much larger written record. Many other memos deliberately avoid calls, repeat caution, or explore alternatives without resolution. That is consistent with Marks’s method. It means success cannot be evaluated by presenting only the memorable turning points. The fair judgment is neither debunking nor celebration. Marks and Oaktree did translate broad credit-market evidence into useful preparation before 2008, then deployed capital aggressively in the panic. The memos document those facts. They also document early warnings, missed timing, partial models and later reframing. His advantage may lie less in prophecy than in preserving flexibility, waiting for rare asymmetry and updating faster than investors committed to a single story. For a student, the test changes the lesson. Do not copy the five conclusions. Copy the recordkeeping discipline that makes it possible to compare what was believed at each date with what happened next. Keep the unflattering entries in the sample. A development course should reward revisions that improved decisions, rather than making revision disappear so the investor looks consistently prescient. ### Work the question Does the “five calls in fifty years” record establish that Marks can time markets? ### Compare with the guide’s reasoning It establishes that he made a small number of useful posture calls near extremes. It does not establish routine timing skill. The sample was selected partly because the calls worked, the 2007 originals show missed timing, and the 2009 memo admits the pre-crisis model missed the larger mechanism. ### Sources - Howard Marks, Risk and Return Today (2004-10-27), Printed pp. 1-7, sections “The Market at Work - 2004 Version,” “Why a Flatter Line?,” and “Implications for Investing”. https://www.oaktreecapital.com/docs/default-source/memos/2004-10-27-risk-and-return-today.pdf?sfvrsn=53bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Taking the Temperature (2023-07-10), Printed p. 1, five-memo selection rule; pp. 3-4, reserve-fund deployment and $6 billion/$7.5 billion figures; pp. 8-11, method and limits. https://www.oaktreecapital.com/insights/memo/taking-the-temperature Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, It’s All Good (2007-07-16), Complete memo, passages on the flattened risk-return line, thin risk premiums, and reduced reserve capacity. https://www.oaktreecapital.com/docs/default-source/memos/2007-07-16-its-all-good.pdf?sfvrsn=8fbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, It’s All Good . . . Really? (2007-07-30), Complete memo, opening admission that the credit turn arrived sooner than expected and the revised assessment that follows. https://www.oaktreecapital.com/docs/default-source/memos/2007-07-30-its-all-good-really.pdf?sfvrsn=97bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Now It’s All Bad? (2007-09-10), Complete memo, opening reassessment of the credit reversal and closing discussion of whether pessimism had gone too far. https://www.oaktreecapital.com/docs/default-source/memos/2007-09-10-now-its-all-bad.pdf?sfvrsn=9fbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, The Long View (2009-01-09), Complete memo, “The Short View” and “The Long View” passages distinguishing the foreseen credit cycle from the missed systemic mechanism. https://www.oaktreecapital.com/docs/default-source/memos/2009-01-09-the-long-view.pdf?sfvrsn=c3bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Lines in the Sand (2017-04-18), Printed p. 1, client pressure and Oaktree’s adoption of subscription lines; p. 7, the internal guidelines process and its unresolved status. https://www.oaktreecapital.com/docs/default-source/memos/lines-in-the-sand.pdf?sfvrsn=bf5dfa65_2 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, There They Go Again... Again (2017-07-26), Printed p. 1, admission that caution since 2011 had not been right; pp. 19-22, limits and conditional response. https://www.oaktreecapital.com/docs/default-source/memos/there-they-go-again-again.pdf?sfvrsn=56d4f265_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ## Howard Marks · Development · 2022-2024. From a low-return world to a sea change, with a forecast he refuses to bet Sea Change is unusual because Marks calls it his only career expression of a macro opinion of that kind. The December 2022 memo argues that two large shifts shaped his working life. Risk-return thinking made low-rated credit investable in the late 1970s. Paul Volcker’s defeat of inflation then opened four decades of declining interest rates. Investors benefited from company growth, but also from a moving walkway that lifted asset values, reduced financing costs and amplified leveraged returns. Marks’s earlier framework treated low rates as a cyclical condition. The post-2008 experience forced a longer view. From 2009 through 2021, easy money made defaults scarce, capital abundant and risk taking rewarding. Oaktree’s credit and value orientation spent much of that period “in the wilderness.” The firm cut the size of successor distressed funds after the 2008 opportunity and grew less than managers who embraced the boom. That is the institutional cost of waiting for a cycle that public policy repeatedly postponed. Inflation and rapid rate increases in 2021-22 changed the opportunity set. Credit instruments that had yielded too little began offering returns closer to what institutions required. Marks inferred that rates were more likely to average 2 to 4 percent than 0 to 2 percent over the next several years. He immediately narrowed the authority of the statement. Oaktree could hold the opinion, but would not bet client money as though it were certain. Easy Money in January 2024 expanded the causal account. Marks revisited Edward Chancellor’s history of interest and catalogued how low rates distort decisions. They raise asset prices, reduce perceived opportunity cost, push investors toward risk, enable leverage and keep weak borrowers alive. These effects can look like skill while the moving walkway runs. They also create “malinvestment” and fragility that become visible only when financing normalizes. The memo contains a rare numerical guess. Marks proposed that the federal funds rate might average 3.0 to 3.5 percent over the next five to ten years, starting in 2024. He labelled it a guess and again said the firm would not use it as a portfolio foundation. That distinction is worth preserving. The forecast could be wrong without invalidating the observed shift from zero rates. Students should separate the strong historical claim from the weak prospective one. This revision did not replace the cycle model. It changed the baseline around which cycles occur. A manager who learned during a forty-year rate decline could attribute too much of past performance to strategy and assume old valuation ranges will return. Marks’s sea-change thesis asks investors to re-estimate what normal means, which strategies benefited from falling rates and which can work without them. The idea also turns back on Oaktree. A full-return credit world favors the firm’s original capabilities, and its public writing can become a sales argument for them. That incentive does not make the thesis false. It does mean claims about a new credit era should be tested against rates, defaults and realized returns over the stated period, rather than accepted because the diagnosis fits Oaktree’s product set. ### Work the question Which part of the sea-change thesis is observation, and which part is forecast? ### Compare with the guide’s reasoning The move from zero rates to rapid tightening and the prior forty-year decline are observations. The claim that rates will average roughly 2 to 4 percent, or 3.0 to 3.5 percent over five to ten years from 2024, is a labelled forecast that Oaktree says it will not treat as certain. ### Sources - Howard Marks, Sea Change (2022-12-13), Complete memo, sections “Sea Change #1,” “Sea Change #2,” and the closing outlook and 2-to-4-percent range. https://www.oaktreecapital.com/insights/memo/sea-change Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Easy Money (2024-01-09), Printed p. 1, ten years without reader response; p. 14, 3.0-to-3.5-percent guess over the next five to ten years. https://www.oaktreecapital.com/insights/memo/easy-money Evidence note: Read directly in the official Oaktree complete collection or official memo page. ## Howard Marks · Development · 2023-2026. A defensive investor adds winners, succession, AI and private-credit self-critique Fewer Losers, or More Winners? revisits the sentence that launched Marks’s memo career. Avoiding losers still fits conventional bonds, where upside is capped. Oaktree, however, now operates aspirational strategies that require actual winners. Marks uses tennis to distinguish a game won by minimizing errors from one won by producing more winners than errors. Neither style is universally superior. Skill, mandate and risk tolerance determine the right balance. This is a meaningful late revision. The original 1990 formulation made the absence of disasters the foundation of superior long-term results. The 2023 memo retains risk control as Oaktree’s first tenet but concedes that loss avoidance alone cannot deliver every target. Marks also says Oaktree has experienced high yield defaults nearly every year since he began the effort, just fewer than the market and fewer than the yield spread allowed for. Risk control is comparative performance under loss, not a spotless record. The founders’ 2025 anniversary conversation moves the development problem from portfolio management to succession. Marks, Karsh and Stone want the culture and principles to endure, but explicitly reject a rule that leadership can only come from inside. They expect the next generation to evolve the business and perhaps do that job better. Marks says the world now changes every minute and Oaktree must change with it. An enduring institution therefore cannot be a museum of its founders’ habits. Mr. Market Miscalculates in August 2024 records a different kind of correction. Marks acknowledged that he had repeatedly attributed a warning about corporate profit growth to Warren Buffett. When he eventually asked Buffett where he had said it, Buffett replied that he had not. The admission authenticates Marks’s source failure, not the supposed quotation. The problem had surfaced before. In Investing Without People in June 2018, Marks corrected his earlier attribution of another remembered line to Albert Einstein. Much of the memo archive builds arguments through memorable sayings, so persistence in writing requires a parallel discipline of checking whether the line belongs to the person named. Marks published both corrections rather than silently dropping the attributions. Marks’s AI view also developed before 2026. In Investing Without People in June 2018, he opened by saying he was not a technology expert and credited Andrew with challenging parts of his timing argument. He accepted that computers could outperform the vast majority of investors and that many active managers did not justify their fees. He provisionally reserved qualitative, creative and long-horizon judgment for the best humans, and hoped people would retain an edge if intelligent machines converged on the same conclusions. Those were dated conjectures, not established limits on AI. AI Hurtles Ahead in February 2026 moved the boundary further. Marks says he asked Anthropic’s Claude for a tutorial and was awed by its personalized reasoning. He now accepts that AI can absorb and process quantitative information better than most humans and may perform entire knowledge-work tasks. He still identifies limits around novel situations, reliability, qualitative judgment and skin in the game. His recommended investment posture is moderate because the technology can be real while its assets are overpriced. The sequence is a revision from a qualified 2018 position, rather than a sudden first encounter with machine investing. The memo also exposes a source problem. Marks relies heavily on AI systems and advisers to explain AI, including claims about their own abilities. He acknowledges hallucination and the need to check outputs, but some of the most consequential evidence is supplied by interested systems and technology executives. Marks’s reported reaction shows how he presented the update. It is not independent proof of the economic forecasts. The period is February 2026, and the claims should be tested as the tools and labor market change. What’s Going on in Private Credit? in April 2026 brings the argument back to Oaktree’s home ground. Marks traces credit innovation from high yield through leveraged loans, securitization, direct lending and retail vehicles. He argues that capital flooded direct lending, managers competed away standards, software exposure grew and semi-liquid vehicles surprised public investors when redemptions were limited. He then discloses Oaktree’s position. Direct lending was about 20 percent of its performing credit investments and less than 15 percent of overall assets under management at that date, according to the firm. Oaktree grew more slowly than peers and expects the new skepticism to create opportunity. The self-interest is plain enough to name. A manager that stayed smaller now benefits from explaining why competitors expanded too far. Marks partly addresses this by saying Oaktree and Brookfield are not paragons and by admitting annual defaults. The most useful late-career continuity is his willingness to state where Oaktree may profit from the diagnosis. The most useful change is broader. A philosophy born in neglected public bonds now has to evaluate private vehicles, technology risk, retail liquidity and tools that may outperform human analysts. Persistence means continuing the argument after the original market has disappeared. ### Work the question What remains stable in Marks’s late work, and what has changed? ### Compare with the guide’s reasoning Price, skepticism, risk control and the need to survive remain stable. He now grants a larger role to winners, expects successors to evolve Oaktree, revises his view of what machines can do, and applies credit-cycle criticism to private products close to his own firm’s interests. ### Sources - Howard Marks, Investing Without People (2018-06-18), Printed p. 1, nonexpert caveat; pp. 12 and 14-17, Andrew’s challenges, limits of active management, provisional human advantages, and the corrected Einstein attribution. https://www.oaktreecapital.com/docs/default-source/memos/investing-without-people.pdf?sfvrsn=7a5ec465_8 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Fewer Losers, or More Winners? (September 12, 2023), Putting It in Brief; Not Risk Avoidance; The Role of Risk Bearing; What About in Practice?. https://www.oaktreecapital.com/insights/memo/fewer-losers-or-more-winners Evidence note: A conceptual and retrospective memo. It distinguishes bond upside, strategies needing winners, and risk control. It does not establish that a reader can identify investment bargains. - Howard Marks, Mr. Market Miscalculates (2024-08-22), Printed p. 5, correction after Buffett denied making the repeatedly attributed profit-growth warning. https://www.oaktreecapital.com/insights/memo/mr-market-miscalculates Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Bruce Karsh, and Sheldon Stone, 30 Years of Oaktree (2025-04-30), Official eight-page transcript, pp. 1-3 on formation and early funds, pp. 4-6 on partnership and operating choices, and pp. 7-8 on succession. https://www.oaktreecapital.com/insights/insight-podcast/education/30-years-of-oaktree-with-howard-marks-bruce-karsh-and-sheldon-stone Evidence note: Read in full. The founders reconstruct their partnership, early funds, culture, mistakes avoided and succession aims. - Howard Marks, AI Hurtles Ahead (2026-02-26), Official HTML, opening Claude tutorial and sections on investment implications, AI limitations, and risk posture. https://www.oaktreecapital.com/insights/memo/ai-hurtles-ahead Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, What’s Going on in Private Credit? (2026-04-09), Official HTML, sections “Direct Lending and Software,” “What Does the Market Know?,” and “What’s a Manager to Do?”; Oaktree disclosure gives 20-percent and 15-percent exposures. https://www.oaktreecapital.com/insights/memo/whats-going-on-in-private-credit Evidence note: Read directly in the official Oaktree complete collection or official memo page. ## Mark Leonard · Development · 1995–2006 · A permanent owner learns what permanence costs Start Leonard’s development with a problem in the evidence. The earliest letter in the official president’s archive concerns 2007. Accounts of his venture-capital years, the founding period and early ownership mistakes therefore arrive partly through later recollection. A convincing history must keep that time gap visible. Knowing what Constellation became does not tell us what its participants could know when they first committed capital. In his 2016 letter, published April 25, 2017, Leonard traced his preference for small teams partly to his venture experience. He also recalled reaching his own managerial limit around 2005–2006. Short-term monitoring remained possible, but knowledge of products, competitors, customers and managers no longer kept pace. This is a more useful formation story than assuming decentralization sprang fully developed from a founder’s philosophy. The organization needed other people who could make decisions he could no longer make well. He said delegation was delayed by doubts about some managers and eventually required replacements. The source is his later account, not a diary from the transition. See the 2016 letter, pp. 7–10. Ownership brought a separate lesson. In the November 2018 shareholder Q&A, Leonard identified taking unnecessary equity as his largest remembered failure. He described a legitimate purpose, balancing the influence of an existing large investor, alongside the enduring cost of dilution. The 2014 letter had already discussed the 1999 financing and capital that remained unused before some was returned. The capital was available, but availability did not create suitable uses for it. The relevant question was not whether a stronger balance sheet was desirable in general. It was whether the financing’s cost and timing matched an actual need. See the 2014 letter, personal compensation and reliable-capital discussion, and November 2018 Q&A, biggest-failure response. He named another early failure in that Q&A, selling a software business after receiving an exceptional offer. The immediate transaction could look successful while undermining the permanent-owner identity he later valued. He said the company would like to buy it back. Do not convert that admission into a claim that the original sale necessarily destroyed financial value. The account does not supply enough transaction evidence to calculate that counterfactual. It does show a changed objective and a reputational cost measured against it. These experiences also resist a tidy rule such as never raise outside capital or never sell anything. In 2014 Leonard wanted reliable additional funding again, this time through a different instrument and for a larger opportunity set. A painful experience can teach a useful constraint, or it can leave a manager overcorrecting. Leonard himself raised the possibility that early dilution had made Constellation too sensitive to issuing shares. The learner must distinguish the durable concern, paying for capital before it can be used, from the contingent response to one financing. Reconstruct the younger decision-maker with an explicit unknown column. How much independent managerial talent existed? How credible were acquisition opportunities? What bargaining power did the original investor have? The available recollections do not settle all of these. Understanding development means learning both from the decisions described and from the questions that the surviving record cannot answer. ### Work the question Reconstruct one early decision twice: first from the information available at the time, then from Leonard’s later account. What changed in the objective, evidence or available alternatives? Identify one unknown that prevents you from declaring the original decision foolish. ### Compare with the guide’s reasoning A strong response distinguishes financing for independence from financing for immediate deployment, and the price of a business from the value of a permanent-owner reputation. It identifies the retrospective nature of the evidence. Do not treat later wealth or stated regret as a complete counterfactual analysis. ### Sources - Mark Leonard, Constellation Software Inc. 2014 President's Letter (2015-04-06), Invested capital; debt duration; ROIC; TSS contribution to maintenance revenue. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: Company-reported figures include management-defined non-IFRS measures explained in the letter's glossary. - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. - Mark Leonard and named Constellation respondents, November 5, 2018 · License revenue volatility, CSI’s biggest failure, Business systems (2018-11-05), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-nov-5-2018-final-1.pdf Evidence note: Full Q&A read. Leonard names excess early equity and a business sale as failures. Jamal explains license volatility. ## Mark Leonard · Development · 2007 · When discipline begins to suppress growth Read the first three quarterly letters together. Their value lies in the movement between them. In the first-quarter letter, Leonard described a reduction in internal growth initiatives as managers discovered that their returns were less attractive than acquisition returns. He expected a self-correcting process: weaker initiatives would disappear, the remaining opportunities would improve, and resources could eventually move back. That was a hypothesis about organizational behavior, not a demonstrated law. See the 2007 Q1 letter, organic-growth discussion, pp. 1–2. By the next quarter, the tone had changed. Leonard worried that the culling had gone too far and that too few new initiatives were being proposed. He preferred greater worthwhile investment even at the expense of current margins. The disagreement was inside his own evolving account. A manager could follow the return discipline faithfully and still produce an organizational result the company did not want. People respond not only to a project’s expected economics but also to what is measured, criticized and rewarded. A process designed to make investment more rational may make proposing uncertain investments less attractive. See the 2007 Q2 letter, organic growth and expense discussion. By the third quarter, a second pressure had become explicit. A public growth target and growth-sensitive bonuses could tempt managers to buy businesses that did not meet investment standards. Leonard described the counterweight provided by the board and managers with longer horizons. He also considered lowering growth expectations rather than relaxing acquisition standards, although he had not yet decided to do so. The important distinction is between wanting growth and needing to display growth. A target intended to focus effort can become a reason to rationalize a marginal deal. See the 2007 Q3 letter, acquisition and target discussion. Taken separately, these passages can be made to support opposing slogans. Cut weak projects. Invest for growth. Maintain discipline. Accelerate acquisitions. Taken in sequence, they reveal the actual management task: calibrating a system that can underinvest and overpay at the same time. There is no contradiction in wanting more experiments while rejecting expensive acquisitions. Both decisions require comparing incremental expected benefits with incremental costs, but the organizational pressures surrounding them differ. The learner should pay attention to who experiences the adjustment. Lower investment can improve near-term reported margins. It can also leave future products weaker and remove learning opportunities from employees. A bonus formula can preserve company profitability in a downturn while reducing employees’ compensation. None of those consequences automatically makes the policy wrong. They identify the people and periods that a headline margin hides. Now design a response that could have been used in 2007. Keep a record of proposed initiatives as well as approved ones. Track which assumptions fail, how long learning takes, and whether managers have stopped proposing good experiments because early losses damage their rewards. Separate a market opportunity becoming worse from a forecasting process becoming more honest. For acquisitions, require the investment case to survive without the corporate growth target. These are course applications, not claims that Leonard implemented this exact checklist. The historical letters show why such distinctions matter. ### Work the question Explain how the same organization can invest too little in new products and feel pressure to acquire too much. Design a review that detects both problems without making reported growth the objective. ### Compare with the guide’s reasoning Look for separate mechanisms. Project scrutiny can suppress proposals or expose weak economics, while a public growth goal can encourage acquisition exceptions. Use proposed-project data, staged learning and independent acquisition assumptions. State which evidence would distinguish healthy discipline from excessive caution. ### Sources - Mark Leonard, Constellation Software 2007 Q1 president’s letter (2007-05-08), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q1_2007_shareholders_report.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2007 Q2 president’s letter (2007 Q2 reporting period; undated letter), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q2_2007_shareholders_report.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2007 Q3 president’s letter (2007 Q3 reporting period; undated letter), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q3_2007_shareholders_report.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. ## Mark Leonard · Development · 2008–2009 · Good reported results, unresolved economic risk The crisis letters make a demanding accounting case because they do not offer a clean division between good management and bad luck. Constellation’s reported performance could improve while individual markets contracted, currency movements helped margins, compensation adjusted downward, and an acquisition carried unfinished loss-making commitments. If the learner labels the whole period a triumph, the mechanism disappears. In the 2008 Q3 letter, Leonard described the recently acquired Maximus Asset, Justice and Education businesses, abbreviated MAJES. Some large contracts were uneconomic. The purchase price had been negotiated with those problems in mind, but the existence of a discount did not establish that the liabilities had been estimated correctly. His conditional language matters. Buying a troubled asset is a decision about the distribution of possible outcomes, including difficult cases, rather than a declaration that trouble has already been solved. See the 2008 Q3 letter, MAJES and acquisition-financing discussion. The fourth-quarter letter explained why accounting could make the result look better than cash experience justified. For contracts whose costs could not yet be estimated reliably, completion accounting complicated recognition. It would take time to understand the ultimate economics. In the 2009 Q1 letter, Leonard directed an investor toward cash paid and cash generated rather than the more flattering adjusted earnings measures. The course does not supply a final acquisition return because these documents do not establish one. A reader who wants closure must obtain later contract and cash evidence, not substitute an optimistic narrative. See the 2008 Q4 and 2009 Q1 letters, MAJES sections. The same sequence challenges simple countercyclical confidence. Weak industry conditions seemed likely to create opportunities to buy distressed software companies. In the 2009 Q1 letter, Leonard said those expected opportunities had not materialized as anticipated. Healthy owner-managers could wait. Financial distress somewhere in a market does not mean the specific assets you want will be sold on terms you can finance. An investor can be right about recession and wrong about the availability of bargains. Funding therefore belongs inside the investment thesis. In the 2008 letters, Leonard became more willing to consider borrowing, while worrying about financing permanent assets with short-term capital. A business can eventually produce attractive cash flows and still fail its owner if the funding must be repaid before those flows arrive. This is a timing problem, not merely a debate over the amount of leverage. Work through the period with separate records for operating health, reported earnings, realized cash and financing obligations. Then add the distribution of costs. Some favorable margin movement reflected exchange rates. Some reflected compensation adjustment. Some acquisition earnings were not yet realized cash. These distinctions are not a demand for paralysis. They let a decision-maker act while knowing which part of the apparent success is demonstrated, which is temporary, and which remains a forecast. That is the practical discipline the crisis teaches. ### Work the question You see rising adjusted earnings after a troubled acquisition. Write the evidence you would demand before calling it a successful turnaround. Include cash, unresolved contracts, funding maturity and a possible explanation unrelated to operating improvement. ### Compare with the guide’s reasoning A good answer separates purchase economics from current accounting, traces cash paid and generated, specifies unresolved contract liabilities, and tests whether financing survives a slower resolution. Currency or compensation effects can explain margin improvement without proving operating improvement. These letters do not provide a completed MAJES investment return. ### Sources - Mark Leonard, Constellation Software 2008 Q2 president’s letter (2008-08-07), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q2_2008_shareholdersreport.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2008 Q3 president’s letter (2008-11-06), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidentsletter50ed1eb201c049b287838ec5e943432b.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2008 Q4 president’s letter (2009-03-04), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q4_2008presidentsletter.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2009 Q1 president’s letter (2009-05-06), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q1_2009_presidentletter.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. ## Mark Leonard · Development · 2009–2010 · A measure is a model with assumptions Leonard wanted shareholders to understand a complicated acquisition business, but explanation required simplification. The annual letter for 2009 began by acknowledging skepticism toward company-defined measures and included conventional financial measures alongside them. That is the starting tension of this chapter: a useful measure can make a business more intelligible while also concealing assumptions that a skeptical owner needs to test. One assumption concerned acquired intangible assets. Adding back amortization makes economic sense only to the extent that the underlying customer relationships and products retain or grow their earning power. Leonard pointed to maintenance revenue and attrition as evidence about that health. He did not provide the full investment-by-investment forecasts available internally. Shareholders therefore had a proxy, not the same information as management. See the 2009 annual letter, pp. 1–3 and glossary. A second assumption concerned capital intensity. Combining a return measure with organic growth may be informative for businesses that can grow without proportionate additional tangible investment. It is not a universal arithmetic rule for any company. The label ROIC in these letters also uses Constellation’s own numerator and invested-equity-capital history. Importing the number into an unrelated business without reconstructing the definition would turn a contextual tool into a misleading benchmark. The same letter contains a change in growth expectations. After accumulating initiative-level evidence, Leonard concluded that internal projects were not producing the returns originally hoped for. He described a permanent reduction in the amount of organic growth the company would seek. This was more consequential than a quarterly forecast revision. A system of observation was changing the organization’s appetite for a type of investment. The earlier 2007 concern about excessive culling still matters: the data could expose bad projects while the measurement process itself changed what managers chose to attempt. The board’s concern about complexity created another fork. Constellation could continue owning many small businesses in different verticals, or concentrate in fewer larger markets at higher purchase prices. Leonard favored continuing the existing approach while monitoring its capacity to scale. He explicitly described the broad span of control as an experiment. That word should prevent the learner from turning later success into proof that the structure was known to be safe at the time. See the 2009 annual letter, many-vertical strategy discussion. Use the 2010 annual letter as a check on what a proxy can and cannot do. It again reports maintenance and per-share operating measures, but the company was also undergoing a strategic review that could lead to a sale. Strong historic performance did not eliminate ownership and governance pressures. A dashboard can describe an operating record without explaining the incentives of the people who control the next decision. For your own business, choose a central metric only after writing its assumptions. Then choose a different measure capable of embarrassing it. If adjusted earnings look strong, what would declining customer value reveal? If growth looks attractive, what would capital consumption reveal? If every measure confirms the preferred story, test whether you have selected the measures to make that happen. ### Work the question Take Leonard’s combined return-and-growth measure and explain where it is informative, where it breaks, and which independent cross-check you would use. Do not apply the percentage mechanically to another company. ### Compare with the guide’s reasoning Reconstruct Constellation’s definitions and the low-asset-intensity assumption. Test maintenance health, attrition, per-share cash, debt and acquisition spending. Explain that limited public disclosure leaves uncertainty even when the proxies are sensible. The company’s later metric revisions are evidence that no proxy should become doctrine. ### Sources - Mark Leonard, Constellation Software 2007 Q4 president’s letter (2008-03-05), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q4_2007_shareholders_report.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2008 Q1 president’s letter (2008-05-07), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q1_2008_shareholdersreportf49d4032f35645668472bbd0e4d94b0b.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2009 annual president’s letter (2010-03-25), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/shareholder-letter-2009.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2010 annual president’s letter (2011-05-02), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/shareholder-letter-2010.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. ## Mark Leonard · Development · 2011–2012 · When the best-looking year is a warning The 2010 annual letter, published in May 2011, placed readers inside an unsettled process. The board was reviewing strategic alternatives. Leonard believed a sale was likely and described the work as a distraction. He also said he would be sad if the company were sold. That is a documented expression of feeling, not permission to invent his private state or portray everyone else as a villain. Major investors had liquidity and valuation concerns, and the board was acting within a different set of pressures from those experienced by a business-unit manager. The next annual letter supplied the consequence. During the review, managers slowed acquisitions and hesitated to add expense for long-term initiatives. Short-term profitability improved. Cash accumulated. The observable financial results therefore could reward precisely the behavior Leonard thought weakened future prospects. See the 2011 annual letter, opening performance analysis and the discussion of managing the stock versus managing the company. This is the central exercise in reading his development. Do not simply memorize that trust matters. Specify the mechanism. If an owner may soon change the compensation system, financing and management structure, an employee has less reason to sacrifice a current reward for an uncertain future reward. The same employee can remain capable and conscientious while changing behavior rationally. Long-term incentives depend on belief in the institution that will honor them. A formula on paper is not enough. Leonard then revised an earlier conviction. He had believed that attention to business fundamentals would allow the share price to take care of itself. The sale process persuaded him that a price too far below perceived intrinsic value could threaten the operating model. An excessively high price posed a different concern: knowledgeable employee-shareholders could sell and use their skills elsewhere. His new position was not an invitation to maximize the price by promotion. It was an argument that the relationship between market valuation, stable ownership and operating behavior could not be ignored. The record also qualifies his self-confidence. Long-term investors gave valuations he initially thought high and subsequently committed more capital. He described their actions as changing his own assessment. That is learning from another participant’s behavior, not merely instructing less informed outsiders. The founder did not possess every useful perspective. Now compare the contemporary 2011 letter with the explanation written in 2012. The later account can identify mechanisms more clearly because consequences have become visible. It can also favor the narrator’s interpretation. The documents establish his concern, reported investment slowdown and changed view. They do not isolate a precise causal effect of the sale process from the economic recovery, acquisition mix and other influences he also discusses. A strong learner can defend the board’s search for liquidity and the founder’s concern about damage without pretending those interests automatically reconcile. Your decision should address both. How might existing owners gain liquidity while employees retain confidence in long-term commitments? Which short-term profit improvement would you refuse to celebrate until you understood what had stopped being funded? ### Work the question Write a board memo during the sale review. Protect a legitimate investor need for liquidity while identifying how the process could damage future investment. Explain why improving margins might be a warning rather than proof that the review is working. ### Compare with the guide’s reasoning The answer should connect uncertainty about future ownership and rewards to deferred initiatives and acquisitions. Consider liquidity alternatives, communication and credible continuity commitments. Keep economic recovery and acquisition mix separate from the alleged process effect. The later founder account supports a mechanism, not a clean causal experiment. ### Sources - Mark Leonard, Constellation Software 2010 annual president’s letter (2011-05-02), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/shareholder-letter-2010.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2011 annual president’s letter (2012-05-02), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/2011_presidents_letter.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. ## Mark Leonard · Development · 2012–2014 · Growth changes the financing problem By the 2012 annual letter, scaling was a persistent question from directors, shareholders and analysts. Leonard initially connected some of that concern to a rising valuation that required future growth. But he also took the operating problem seriously. More acquisitions required more people capable of evaluating and supporting them, while larger transactions demanded a different financing capacity. Growth was creating a problem that success at a smaller scale had not solved. The letter’s review of internal initiatives is particularly revealing. Years of project measurement had produced adaptations such as smaller early commitments, faster termination when key assumptions failed, and clearer responsibility. Yet it had also reduced the number of new initiatives. Renewed spending was not yet accompanied by conclusive evidence of stronger aggregate growth. This is an unfinished learning process, not a triumphant before-and-after diagram. See the 2012 annual letter, initiative analysis following Table 3. Funding exposed competing commitments. Leonard wanted the ability to acquire during difficult markets without relying on capital that could disappear at the wrong moment. He also recognized that the dividend had helped attract investors during the ownership transition. Removing it would affect people whose support had helped preserve independence. A funding decision therefore contained a relationship obligation as well as a calculation of returns. Later, in the 2013 annual letter, reducing the dividend reappeared as a possibility if suitable financing could not be found. The preference was conditional, not a permanent law. The 2013 letter describes an experiment with analysts’ valuation models. Leonard found that much of the market value could be justified by assumptions about continuing acquisition activity. He was surprised by how much value disappeared in the model when acquisitions stopped. This is a distinction between a company’s existing assets and its ability to deploy future capital. It also creates a danger: if investors already pay for that ability, future good acquisitions may be necessary merely to satisfy expectations. See the 2013 letter, consensus-model sensitivity discussion. The same models favored larger acquisitions under particular assumptions, but smaller acquisitions remained more attractive when available at lower multiples. The implication was conditional substitution, not a declaration that bigger deals were inherently better. Changes to financing could expand what was possible, but could not make an overpriced transaction sound by themselves. The proposed long-duration instrument evolved through conversations with institutions and banks. Leonard described commercial and organizational obstacles to a novel structure, then favored making an attractive offer to existing shareholders. By the 2014 letter, he was again advocating more reliable capital while acknowledging the poor earlier experience with excess equity. The changed instrument and opportunity set explain why revisiting an old decision was not necessarily inconsistency. For practice, separate four questions: how much capital a strategy can use, when opportunities arrive, how dependable the funding is, and what existing owners were promised. A company can have excellent acquisition skill and still lack a financing structure that fits it. It can also obtain ideal financing and then waste it. The argument for reliable capital should survive a scenario where deployment takes longer than hoped. ### Work the question Design a financing recommendation for a permanent acquirer that pays a dividend and faces sporadic large opportunities. State what would make you retain cash, borrow, reduce distributions or refuse the acquisition. ### Compare with the guide’s reasoning A useful recommendation matches funding duration and covenants to uncertain cash timing, considers the cost of idle capital, and respects stated shareholder commitments without treating them as unchangeable. It also tests acquisition economics independently of financing availability and market expectations. ### Sources - Mark Leonard, Constellation Software 2012 annual president’s letter (2013-05-01), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-final.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software Inc. 2013 President's Letter (2014-04-30), Maintenance revenue; valuation sensitivity; TSS-scale acquisition; capital-funding discussion. https://www.csisoftware.com/wp-content/uploads/2026/04/presidentletter_2013.pdf Evidence note: Management report using company models and assumptions. Model outputs are not independently verified forecasts. - Mark Leonard, Constellation Software Inc. 2014 President's Letter (2015-04-06), Invested capital; debt duration; ROIC; TSS contribution to maintenance revenue. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: Company-reported figures include management-defined non-IFRS measures explained in the letter's glossary. ## Mark Leonard · Development · 2014–2015 · Building a company that does not consume its founder The personal passage in the 2014 letter belongs in a serious study of development. Leonard announced that he would stop taking salary and incentive compensation and pay his own expenses. He explained that he wanted to work less intensively and reduce the sense of obligation associated with his previous routine. The letter describes a choice about the shape of a life as well as the alignment of capital. See the 2014 annual letter, personal compensation discussion following Table 3. Do not turn this into a universal prescription for founders. Ownership, accumulated wealth, capable colleagues and a particular organizational structure made the arrangement possible. A young founder without those conditions cannot reproduce its economics by declining a salary. The transferable question is whether the institution depends on a level of personal effort that its leader no longer wants, or cannot continue, to supply. Leonard connected the change to the strength of managers below him. He expected to continue acquisitions, monitoring, research and financing work, while relying more on people who had become strong coaches and culture carriers. His account gives other participants a central role. If the operating model is genuinely decentralized, it should reduce some dependence on the founder. If it only uses that word while every important choice still waits for him, a lighter workload would expose the gap. The 2015 letter returned to the arrangement and reported comfort with it from the board and shareholders. That establishes his account of acceptance, not a controlled measurement of organizational performance caused by reducing hours. A serious reader should ask what decisions were delegated, how outcomes were monitored and which capabilities remained concentrated at headquarters. The same period brought a related compensation problem for employees. Historical bonus structures could become awkward as businesses produced very high returns on old invested capital. Capping rewards could encourage shifting results between periods. Rewriting a trusted plan could undermine willingness to make long-term trade-offs. Requiring successful groups to redeploy capital offered another path, but also changed their jobs. They now needed to become allocators as well as operators. See the 2015 letter, Combined Ratio discussion. This is a recurring development pattern. A practice can help create the success that eventually makes it inadequate. The solution cannot be chosen only by asking whether the old practice worked. Ask whether its incentives still fit the present opportunity set, who has acquired new responsibilities and which promises people relied upon. Changing a compensation arrangement without acknowledging those promises can destroy information and trust that took years to accumulate. For your own exercise, map the personal and institutional dependencies separately. Which activity requires the founder’s judgment today? Which merely reflects habit? Which colleague can inherit the work, and what experience would make that transfer credible? The aim is not maximum leisure or maximum effort. It is a durable division of responsibility that preserves the organization’s ability to learn when its founder changes. ### Work the question Identify a founder task that should remain personal, one that can be delegated now, and one that requires an apprenticeship before delegation. Explain how compensation and decision rights would change without relying on heroic working hours. ### Compare with the guide’s reasoning Separate exceptional personal wealth from a transferable organizational lesson. Match responsibility to demonstrated capability, specify a learning period where needed, and preserve credible reward commitments. Acceptance of the arrangement and subsequent success do not alone prove that reduced founder effort caused better results. ### Sources - Mark Leonard, Constellation Software Inc. 2014 President's Letter (2015-04-06), Invested capital; debt duration; ROIC; TSS contribution to maintenance revenue. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: Company-reported figures include management-defined non-IFRS measures explained in the letter's glossary. - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. ## Mark Leonard · Development · 2015–2017 · The evidence pushes back on the philosophy A course built from Leonard’s most quotable principles could easily overstate his certainty. The 2015 and 2016 letters contain the corrective. He wanted more acquisition capacity and favored small autonomous businesses, but the evidence was incomplete and not always friendly to the preferred explanation. The 2015 letter compared Constellation with selected high-performing conglomerates. Some had expanded from unusually attractive original businesses into larger acquisitions at higher prices. Returns could decline while total value still grew. That pattern helped Leonard frame the pressure created by an expanding pool of capital. Maximizing a historical return percentage is different from finding the best use for the next dollar. It also explains why accepting a lower return on a sound new investment need not imply that the investment is bad. At the same time, his analysis of past hurdle changes suggested that expected acquisition returns moved with the hurdle across whole vintages. The concern was not limited to a few marginal deals becoming eligible. A lowered standard could change negotiation, proposal selection and forecasting throughout the organization. This was company evidence about expected returns, not a universal causal law with every confounder removed. See the 2015 letter, Organic Net Revenue Growth discussion. The 2016 letter put the pressure in personal terms. Leonard described becoming increasingly committed to a large acquisition simply because he was spending time on it. The transaction narrowly failed the hurdle and was not completed. Experience did not make him immune to escalation of commitment. An outside standard and review process still had work to do. The useful lesson is not that a disciplined investor never feels temptation, but that the process must function when temptation is present. See the 2016 letter, p. 4. Then consider his argument for small business units. He acknowledged that the apparent statistical advantage was weak and could reverse when a few outliers were removed. He still favored smaller units on the basis of experience and judgment. That is legitimate decision-making under incomplete evidence if its status remains explicit. It is not legitimate to retell the preference as scientifically established superiority. See the 2016 letter, pp. 6–8. The letter also described his earlier inability to keep up with all the information needed to coach businesses. Delegation was constrained by trust in specific managers, not solved by drawing a different organization chart. Developing future portfolio managers required operating experience, team-building ability and acquisition judgment. Capital was not the only scarce resource. The organization needed people able to deploy it responsibly. The learner’s task is to retain the practical judgment without inflating its evidence. You may need to split a business before a definitive study exists. You should still identify what the split is meant to improve, what it could worsen, and which observations would make you reconsider. Similarly, a hurdle can protect discipline while becoming too restrictive. The next chapter follows Leonard into exactly that tension. ### Work the question Choose one of Leonard’s beliefs in this period and label its support: direct result, observational association, forecast or managerial judgment. Propose a decision that uses it without pretending the evidence is stronger than it is. ### Compare with the guide’s reasoning A good response can defend a small-unit experiment while admitting weak causal evidence, or retain a hurdle while recognizing organizational and forecasting effects. It should specify a reversible test, a comparison and an adverse result that would change policy. Never treat the rejected deal’s unknown future as proof it was correctly rejected. ### Sources - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. - Mark Leonard, Constellation Software Inc. 2017 President's Letter (2018-04-20), Performance metrics; retained-capital policy; perpetual-owner objective; business-unit count. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-april-2018-final.pdf Evidence note: Management report. Statements about scalability and governance are Leonard's judgments, not independent causal findings. ## Mark Leonard · Development · 2018 · Apply the skepticism to Leonard In 2018 Leonard welcomed critics and defended a highly particular governance model. He wanted directors who understood the business deeply and could coach management over a long horizon. He argued that replacing capable directors merely because they had served a set term could sacrifice accumulated knowledge. That is an argument about the value of experience, but it does not by itself establish that every incumbent is competent or that a familiar board will challenge management effectively. See the 2017 annual letter, published April 20, 2018, governance discussion and director-search criteria. Read his references to practice and expertise carefully. The letter uses a popularized hours-of-practice argument to illustrate the difficulty of acquiring contextual knowledge. The course does not treat that arithmetic as a validated requirement for directorship or a universal rule for expertise. Time served and useful learning are different variables. A board can accumulate knowledge, reinforce its own assumptions, or do both. The shareholder Q&A that followed makes this problem concrete. In August, a reader asked whether Leonard’s study of successful conglomerates also examined failures. He acknowledged that it had not done so formally and called the design an example of confirmation bias. That admission changes how the earlier study should be used. It can identify possible practices worth investigating, but it cannot show that those practices distinguish winners from failures without the missing comparison. See the August 3, 2018 Q&A, first question and Leonard’s response. The July Q&A provides another useful correction. Leonard described a culture of skeptical argument and information that others did not have. A follow-up challenged whether these were facts or beliefs. He accepted the uncertainty and qualified the language. A reader who repeats the confident formulation but omits the qualification loses the development. The aspiration to possess unusual insight is not evidence that the insight is true. The October responses further complicate the idea of a single company culture. Leonard resisted promoting one approved culture from headquarters and described multiple subcultures that could prosper or fail. In the same month, he questioned whether a central engagement-measurement program would improve performance and emphasized closer attention to managers who could not retain good people. These are his judgments and interpretation of research. They should not become unsupported universal claims that engagement measurement is useless or that informal awareness always works. Now test him fairly. The strongest critique is not that he was biased and therefore wrong about everything. Everyone selects evidence. The important questions are whether the selection concealed a necessary comparison, whether contrary evidence was welcomed, and whether the next decision changed. Likewise, the strongest defense of long board tenure must specify how the board will detect declining competence or complacency. A rule that exempts successful organizations from scrutiny can become dangerous precisely when past success gives it credibility. This chapter supplies a model for studying every Titan. Keep the valuable mechanism, expose the missing comparison, and ask what evidence could defeat the preferred explanation. Admiration should make that work more careful, not less. ### Work the question Write the strongest defense and strongest critique of Leonard’s case for long board tenure. Then use the conglomerate-study admission to explain what additional evidence would improve his reasoning without discarding everything he learned. ### Compare with the guide’s reasoning Distinguish contextual expertise from tenure itself. Test challenge, succession and actual contribution. For the conglomerate study, add comparable failures and examine whether the proposed practices discriminate between outcomes. Leonard’s admission improves transparency but does not retroactively repair the original selection. ### Sources - Mark Leonard, Constellation Software Inc. 2017 President's Letter (2018-04-20), Performance metrics; retained-capital policy; perpetual-owner objective; business-unit count. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-april-2018-final.pdf Evidence note: Management report. Statements about scalability and governance are Leonard's judgments, not independent causal findings. - Mark Leonard and named Constellation respondents, August 3, 2018 · HPC’s, Blackboard, Customer Acquisition Economics, Margin Trajectory, Organic Growth Profile, Tax Rate, SaaS vs On-Premise (2018-08-03), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-august-2018-final.pdf Evidence note: Full Q&A read. Leonard admits selection bias in the conglomerate study; other responses cover margins, growth, taxes and SaaS. - Mark Leonard and named Constellation respondents, July 25, 2018 · Employee Retention, Consensus, IFTODH, Bias, Incentive Compensation (2018-07-25), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-july-25-2018-final.pdf Evidence note: Full Q&A read. Leonard discusses skeptics, formulaic incentives, downside borne by homebuilding managers and uncertainty in supposedly private facts. - Mark Leonard and named Constellation respondents, October 4, 2018 · Employee attraction, retention, motivation and engagement (2018-10-04), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-oct-4-2018-final.pdf Evidence note: Full Q&A read. Leonard questions engagement measurement and discusses opportunity, empathy and supervisor quality. - Mark Leonard and named Constellation respondents, October 9, 2018 · Moat, SaaS, TAM, corporate culture (2018-10-09), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-oct-9-2018-final-1.pdf Evidence note: Full Q&A read. Leonard discusses SaaS value capture and plural cultures. The club-software example belongs to Barry Symons. ## Mark Leonard · Development · 2018–2021 · A reversal with a boundary The February 2021 letter is unusually explicit about a change of mind. Leonard said a director had long argued that Constellation should deploy more of its cash rather than distribute it, even if that required lower hurdle rates. Leonard had resisted. He now accepted a different direction. To understand the reversal, begin before the conversion, while his earlier position was still being defended. In the September 2018 Q&A, he anticipated that persistent free-cash-flow growth could eventually require lower hurdles, but hoped to postpone the change. In the February 2019 Q&A, he defended a special dividend and retaining the existing hurdle discipline. His reasons included the observed tendency of expected returns to move with the hurdle, the company’s historical performance, and the need for excess returns that could be shared with employees. A director was also dissatisfied with progress finding opportunities outside vertical-market software. This was an ongoing debate with identifiable opportunity costs, not a sudden discovery in 2021. The 2021 policy retained traditional hurdles for the small and medium-sized acquisition activity while pursuing more very large transactions and developing competence beyond the existing sphere. It also ended routine reliance on special dividends except in compelling circumstances. The boundary matters. Saying that Leonard abandoned discipline is too broad; saying that nothing changed is also wrong. He changed the acceptable use of incremental capital in a specific part of the opportunity set. See the complete February 15, 2021 letter. A learner should be able to argue both sides at the earlier decision date. Retaining a high hurdle might preserve bargaining discipline, prevent optimistic forecasting and maintain employee participation in exceptional returns. It might also leave capital idle or distribute it to owners with poorer opportunities. Lowering a hurdle could improve total value while reducing an average return metric. It could also normalize weaker underwriting. The decision depends on the next-best use of capital and on whether the organization can contain the behavioral spillovers it fears. The letter reframed headquarters’ role from directly owning software businesses toward stewardship of investors’ capital, as operating groups increasingly handled the established acquisition model. This linked an investment-policy change to an organizational change. A policy that fits a small specialist allocator may cease to fit a larger institution whose operating units have become allocators themselves. Do not attach a successful ending that the announcement does not establish. In the April 2022 CFO interview, Jamal Baksh described difficulty finding large outside-sector opportunities and cautioned against assuming them in valuation models. That is another executive’s account, not Leonard speaking. It provides a useful check on treating a declared intention as an accomplished new capability. See Baksh’s interview, pp. 3–4. Your answer should state exactly what changed, what stayed in place, and what result would indicate that the boundary failed. That is how a reversal becomes a teachable decision rather than a character judgment. Consistency is valuable when the conditions persist. Revision is valuable when the opportunity costs change. Neither word settles whether this particular decision was sound. ### Work the question Write a 2019 defense of returning cash, then a 2021 argument for the revised policy. Identify the new boundary around large deals and a warning sign that lower hurdles are contaminating the rest of the organization. ### Compare with the guide’s reasoning The defense should include behavioral spillovers and employee incentives, while the revision compares idle or distributed capital with realistic incremental investments. Traditional small- and medium-deal hurdles stayed in place. Treat outside-sector competence as something to develop and verify, not an asset created by announcing it. ### Sources - Mark Leonard and named Constellation respondents, September 19, 2018 · Additional sectors for investment, Buybacks, ROIC, Value Investing (2018-09-19), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-september-2018-final.pdf Evidence note: Full Q&A read. Leonard qualifies his buyback position and distinguishes beliefs from certainty; Jamal supplies the invested-capital reconciliation. - Mark Leonard and named Constellation respondents, February 20, 2019 · Special Dividend (2019-02-20), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q-a-february-2019-divrelated-final.pdf Evidence note: Full Q&A read. Leonard defends the 2019 special dividend, magnetic hurdles and employee sharing of excess returns. - Mark Leonard, Constellation Software Inc. 2021 President's Letter (2021-02-15), Complete letter, especially dividend policy, large VMS acquisitions, and head-office role. https://www.csisoftware.com/wp-content/uploads/2026/04/letter-to-shareholders-february-15-2021.pdf Evidence note: Records a board policy decision and management intentions as of publication. The reported 16 percent invitation figure does not have an unambiguous denominator in the letter. It does not establish the results of later investments. - Jamal Baksh with Tegus, May 26, 2022 · CFO Interview (2022-04-06; posted May 26, 2022), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/April-6-2022-Tegus-interview-with-CFO.pdf Evidence note: Full seven-page company-reviewed interview read. Baksh is the speaker throughout; this is not Leonard’s testimony. It explains VMS Ventures, delegated acquisitions and limits to outside-sector investment. ## Mark Leonard · Development · 2021–2026 · Can the institution outlast the central person? A development history should continue beyond the most famous letter. The later evidence asks whether the practices can survive changes in ownership arrangements, growth opportunities and the founder’s role. It also forces careful attribution. The available record includes Leonard’s signed Q&A answers, other executives’ explanations and company announcements. Those are different kinds of evidence. In the September 2021 Q&A, Leonard explained extending an option horizon in the Topicus ownership arrangements. His concern was behavioral. As an exercise date approaches, someone whose future ownership is uncertain may favor short-term value drivers. Extending the horizon was intended to align participants with a longer ownership period. This echoes the earlier sale-process lesson: an incentive system includes the expected duration of the relationship, not just this year’s bonus formula. See the Q&A’s first response, signed MHL. In the same document, he resisted reducing management selection to flattering lists of personality traits. He preferred evidence about the team’s customer and employee record. For a weak business, intervention might range from coaching to replacing management, often drawing on people already inside the acquired company. Permanent ownership did not mean never intervene. It changed the time available and the preferred way to build capability. The claim remains a stated practice; outsiders would need specific business records to judge how consistently it was applied. Baksh’s April 2022 interview adds institutional detail. It describes acquisition authority moving downward and a venture mechanism intended to fund larger internal ideas without placing all their early costs inside a small unit’s existing bonus economics. These explanations are useful because they connect governance design to earlier problems in growth investment. They belong to Baksh, and the course does not attribute his words or figures to Leonard. The interview is also a company-reviewed account, rather than independent proof that every mechanism works. The September 25, 2025 company announcement establishes the next transition. Leonard resigned as president for stated health reasons and Mark Miller was appointed. The March 27, 2026 announcement said Leonard would not stand for board re-election and would continue as an adviser, particularly on the Permanent Engaged Minority Shareholder strategy. The documents establish announced roles and intentions. Their expressions of confidence do not establish long-run succession performance, and they do not authorize speculation about private medical circumstances. Minority ownership was not wholly absent from the earlier story. The 2008 Q1 letter discussed the opportunities and limitations of minority investments. The later advisory focus should therefore prompt a comparison across time, not an invented claim that Leonard discovered minority stakes in 2026. What changed in scale, influence, governance and deployment need? The announcements alone cannot answer all of that. End this path with an institutional audit. Which processes can another person execute? Which depend on Leonard’s judgment or relationships? Who can challenge the next leader, train future allocators and recognize a failed assumption? The test of a durable system is not whether its language remains the same after succession. It is whether the next generation can preserve what works, revise what does not, and explain those decisions with evidence. ### Work the question Design a succession review that could be conducted without relying on the founder’s reputation. Identify evidence for preserved capability, evidence of drift, and one practice the successor should be allowed to change. ### Compare with the guide’s reasoning Examine delegated decisions, underwriting outcomes, development of managers, incentive horizons and customer and employee retention. Separate announced continuity from demonstrated results. Changes can be appropriate if they address new conditions while preserving clear accountability. Attribute Baksh and company statements separately from Leonard. ### Sources - Mark Leonard, Constellation Software 2008 Q1 president’s letter (2008-05-07), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q1_2008_shareholdersreportf49d4032f35645668472bbd0e4d94b0b.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Constellation Software; named respondents vary, September 17, 2021 · Members Agreement, Acquisitions, NCI and Dividends (2021-09-17), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-september-2021-final.pdf Evidence note: Full Q&A read. Leonard discusses Topicus option horizons and judging managers through customer and employee records. Jamal answers accounting and tax questions. - Jamal Baksh with Tegus, May 26, 2022 · CFO Interview (2022-04-06; posted May 26, 2022), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/April-6-2022-Tegus-interview-with-CFO.pdf Evidence note: Full seven-page company-reviewed interview read. Baksh is the speaker throughout; this is not Leonard’s testimony. It explains VMS Ventures, delegated acquisitions and limits to outside-sector investment. - Constellation Software, Mark Leonard resignation and Mark Miller appointment (2025-09-25), Announcement body and separately attributed statements.. https://www.csisoftware.com/constellation-software-inc.-announces-the-resignation-of-mark-leonard-and-appointment-of-mark-miller-as-president-of-constellation-software/ Evidence note: Company announcement establishes the leadership change and stated health reason. Expressions of confidence are expectations, not proof of succession outcomes. - Constellation Software, Mark Leonard decision not to stand for board re-election (2026-03-27), Announcement body and separately attributed statements.. https://www.csisoftware.com/constellation-software-inc-announces-mark-leonards-decision-to-not-stand-for-re-election-to-board-of-directors/ Evidence note: Company announcement gives the end-of-term plan and continuing advisory role focused on PEMS. The company’s retrospective praise is not an independent assessment. ## Warren Buffett · Turn accounting into an owner question A reported profit is a starting point. An owner still has to ask how much cash can leave the business without weakening the operation. Buffett made this adjustment concrete in Berkshire's 1986 letter after the Scott Fetzer acquisition. Purchase accounting created higher asset values and new depreciation and amortization charges. The acquired operation had not changed overnight, yet its reported earnings changed because the accounting basis changed. A valuation method that treated the old and new legal forms as different economic machines would confuse the record with the thing being recorded. Buffett's owner-earnings calculation starts with reported earnings, adds back depreciation, depletion, amortization and selected non-cash charges, then subtracts the average capital spending required to maintain long-term competitive position and unit volume. Required working capital also belongs in the subtraction when maintaining the business demands more of it. The hard part is the maintenance estimate. It is not printed as a clean audited fact, and managers can flatter the answer by calling necessary upkeep optional growth. Buffett explicitly accepted an imprecise estimate because a precise accounting total could answer the wrong question. The distinction works in both directions. Purchase-accounting amortization may reduce reported earnings without reducing the operation's cash-generating power. At the same time, adding back all depreciation can greatly overstate what owners may take. A manufacturer can defer replacement for a year and display strong cash, but machinery, stores, software or reserves eventually decay. In the 1986 Scott Fetzer comparison, Berkshire estimated the maintenance-capital subtraction near $8.3 million, the old accounting basis total for depreciation, depletion, amortization and specified non-cash charges. The comparable new-basis add-back was $19.9 million. These were 1986 measures, not earnings totals. Buffett warned that many businesses face the opposite situation, where maintenance capital exceeds the accounting add-back. Use the idea as a reconciliation rather than a magic metric. Begin with the income statement. Identify non-cash charges and explain why each one does or does not represent economic wear. Estimate maintenance capital across a full replacement cycle, not a convenient quarter. Add working-capital needs. Then run a range, because the answer depends on judgments about capacity, competition and service quality. A low estimate should be tested against physical evidence such as equipment age, outages, lost customers and deferred projects. Record the estimate made at the time so later spending can test it. The boundary matters. Owner earnings do not make growth investment free, and they do not tell you whether the remaining cash deserves a high valuation multiple. A business can produce cash today while its competitive position erodes. The calculation answers a narrower question. It estimates distributable cash after preserving the current earning engine. Durability, growth opportunity, financing risk and price still require separate judgments. ### Work the question In a fictional exercise for the twelve months ending December 31, 2026, a company reports $30 million of earnings, $12 million of depreciation, and $8 million of capital spending, all in United States dollars. What further evidence would you need before calling $34 million its owner earnings? ### Compare with the guide’s reasoning You would need to separate maintenance from growth capital, estimate required working capital, inspect whether current spending is temporarily deferred, and test whether the assets and competitive position can be maintained. If all $8 million is maintenance and working capital is unchanged for that twelve-month period, $34 million is the arithmetic result, but the evidence may justify a larger subtraction. ### Sources - Warren E. Buffett, 1986 Chairman's Letter (1986 reporting period; published 1987), Purchase-Price Accounting Adjustments and the owner-earnings discussion near the end of the letter. https://www.berkshirehathaway.com/letters/1986.html Evidence note: Explains Buffett's valuation adjustment using Scott Fetzer. Maintenance capital expenditure is explicitly an estimate, not an audited line item. ## Warren Buffett · Recognize when efficient investment cannot improve the business Berkshire's textile history separates managerial effort from business economics. When Buffett's partnership took control in 1965, the textile operation had $22 million of accounting net worth. The merged textile companies had produced $530 million of sales and a $10 million aggregate loss over the prior nine years. Buffett later wrote that Ken Chace and Garry Morrison managed well, employees cooperated, and product lines, machinery and distribution were repeatedly changed. The operation still failed to earn adequate returns. Blaming weak people would have hidden the more difficult diagnosis. The capital trap appeared in proposals that looked sensible one at a time. New textile equipment promised lower variable cost and passed ordinary return tests. Competitors could buy similar equipment. Once enough mills did so, the cost reduction became the basis for lower industry prices. Each operator invested to avoid falling behind, yet the group earned little more on a larger capital base. The relevant counterfactual was not an old machine forever. It was committing more capital merely to keep a place in a commodity contest where rivals made the same move. Berkshire did not close the mills at the first weak year. Buffett's stated conditions for continued support included important community employment, candid management, cooperative labour and an expectation of modest cash returns. The fourth condition failed. After a moderately profitable 1979, the textile operation consumed major cash. By mid-1985 Buffett judged continuing losses likely. He preferred a sale to a buyer that would keep operating, even for lower proceeds, but reported no interest. Berkshire decided in July 1985 to close the operation, and the letter says the work was largely completed by year-end. This was a delayed exit with acknowledged human costs, not a frictionless reallocation slogan. The equipment sale, completed through disposals before and during an early 1986 auction, exposed another error. Machinery that had cost about $13 million and carried an $866,000 book value produced gross proceeds of $163,122, while removal and other costs made the net less than zero. Replacement cost was not realizable value. An asset is valuable because of the cash it can help earn in its actual competitive setting, not because recreating it would be expensive. Liquidation value can also fall as many similar operators exit together. A useful decision procedure has four parts. First, calculate cash returns after the investment required to remain competitive. Second, ask what happens when rivals copy the project. Third, distinguish a temporary downturn from a structure that repeatedly passes benefits to customers. Fourth, state the obligations to employees and communities alongside the capital cost. A low-return operation can still merit support for bounded reasons. What it cannot receive is an unlimited claim on the rest of the company's cash merely because the assets, history or people matter. ### Work the question When every competitor can install the same cost-saving equipment, what evidence would show that your proposed upgrade creates value rather than resets the industry's price floor? ### Compare with the guide’s reasoning Look for an advantage competitors cannot copy on similar terms, a binding capacity constraint, differentiated customer value, or a cost reduction that remains with the owner after likely price responses. Model the collective response and the no-invest, invest, sell and close alternatives. ### Sources - Warren E. Buffett, 1985 Chairman's Letter (March 4, 1986), Shutdown of Textile Business. https://www.berkshirehathaway.com/letters/1985.html Evidence note: Contemporary account written after the July 1985 closure decision. It reports Buffett's own diagnosis and hindsight about earlier delay. - Warren E. Buffett, 2014 Chairman's Letter (2014 reporting period; published 2015), Berkshire, Past, Present and Future, especially the textile, See's and Dexter passages. https://www.berkshirehathaway.com/letters/2014ltr.pdf Evidence note: Fifty-year retrospective. It supplies later cumulative figures and Buffett's revised interpretation, not a contemporaneous record of the earlier decisions. The Dexter share value is measured at the letter's February 2015 publication. ## Warren Buffett · Value the cash a franchise can release See's Candies helped Buffett distinguish accounting assets from an earning franchise. Blue Chip Stamps bought See's in 1972 for $25 million when it had about $8 million of net tangible assets and earned about $2 million after tax. The premium did not come from hidden machinery. Buffett attributed it largely to a consumer reputation formed through repeated experiences, which allowed customer value rather than production cost alone to influence price. He called the capitalized value of returns above ordinary rates economic goodwill. The 1983 letter used an inflation comparison. Imagine two businesses each earning $2 million. See's needed $8 million of tangible assets, while the hypothetical ordinary business needed $18 million. If the price level and nominal earnings doubled, each would probably need more inventory, receivables and equipment merely to preserve real unit volume. The asset-light operation would need about $8 million more under the example. The other would need about $18 million. Both could report doubled nominal earnings, but one would release much more cash to its owner. Later results sharpened the mechanism without proving it was predictable in 1972. The 2007 letter reported that See's sales volume rose from 16 million pounds in 1972 to 31 million in 2007, roughly 2 percent annual growth, while sales reached $383 million and pre-tax profit $82 million. Required capital rose from $8 million to $40 million. Berkshire reported $1.35 billion of cumulative pre-tax earnings through 2007. The 2014 retrospective updated cumulative pre-tax earnings to $1.9 billion and added investment to $40 million. Those distributions could be invested elsewhere. The lesson is more specific than buying brands. A favorable name matters only if it supports customer choice, price, repeat demand or lower selling cost, and if maintaining that advantage does not absorb the resulting earnings. A famous brand with constant promotional spending, fashion risk or heavy store investment may be capital hungry. A regulated franchise can also earn high returns, but regulation may limit price. A low-cost producer may possess economic goodwill without a consumer brand. The mechanism must be identified, not inferred from a label. Test a candidate with a cash bridge. Measure incremental earnings over a period and the incremental tangible capital required to produce them. Ask how price changes affected volume and whether customers had acceptable substitutes. Identify spending that preserves reputation. Then examine whether excess cash actually left the business or was needed to defend it. High historical return on tangible capital can attract entry and imitation. The strongest case requires a reason the advantage can endure. Compare contribution after brand-maintenance spending across several price and volume changes, not only a favorable year. See's is evidence of one successful instance and of Buffett's later interpretation. It is not a guarantee that every beloved product can raise price safely. ### Work the question In a fictional 2026 comparison using annual United States dollar figures, two businesses each earn $4 million. One needs $12 million of tangible capital and the other needs $40 million. What must you establish before preferring the first? ### Compare with the guide’s reasoning Establish that the first business can maintain earnings, customer value and competitive position with the smaller capital base. Also compare price, growth opportunities, risk and the maintenance spending hidden outside tangible assets. Capital efficiency alone does not settle valuation. ### Sources - Warren E. Buffett, 1983 Chairman's Letter (1983 reporting period; published 1984), Goodwill and its Amortization, especially the See's comparison. https://www.berkshirehathaway.com/letters/1983.html Evidence note: Contemporary discussion of accounting and economic goodwill using historical See's figures. The acquisition rationale is partly retrospective. - Warren E. Buffett, 2007 Chairman's Letter (2007 reporting period; published 2008), Businesses, The Great, the Good and the Gruesome. https://www.berkshirehathaway.com/letters/2007ltr.pdf Evidence note: Retrospective comparison of See's, FlightSafety, capital-intensive growth, and Dexter. The Dexter share value is measured at the letter's February 2008 publication, not in the 2007 reporting year or in 1993. - Warren E. Buffett, 2014 Chairman's Letter (2014 reporting period; published 2015), Berkshire, Past, Present and Future, especially the textile, See's and Dexter passages. https://www.berkshirehathaway.com/letters/2014ltr.pdf Evidence note: Fifty-year retrospective. It supplies later cumulative figures and Buffett's revised interpretation, not a contemporaneous record of the earlier decisions. The Dexter share value is measured at the letter's February 2015 publication. ## Warren Buffett · Count the business surrendered in a stock deal The Dexter Shoe acquisition shows why shares are a business interest, not costless deal currency. Berkshire's 1993 letter presented a strong contemporary case. Dexter made more than 7.5 million pairs a year, operated 77 retail outlets, held about 15 percent of United States golf-shoe output, and had received supplier awards from two large retailers. Berkshire's earlier shoe acquisitions had exceeded expectations. The letter acknowledged the common belief that domestic footwear could not compete with low-wage imports, then argued that Dexter's management and workforce had done so. Berkshire acquired Dexter on November 7, 1993, using 25,203 Class A shares. The sellers did not want cash. Buffett described advantages to them such as diversification, liquidity, tax deferral and continued operating autonomy. These were real features of the consideration, but they did not answer the acquirer's central question. What fraction of Berkshire's future value was being surrendered, and was the target worth more? A quoted share price converts the exchange into dollars. Intrinsic value asks what business claim leaves the continuing owners. The later record reversed the optimistic assessment. In 2001 Buffett separated three mistakes. Berkshire bought Dexter, paid with stock, and delayed operational changes when problems became apparent. He said Dexter had prospered before and for several years after purchase despite harsh import competition, but he wrongly expected that resilience to continue. The letter for the 2007 reporting year, published in February 2008, put the purchase value at $433 million and estimated that the Berkshire shares represented $3.5 billion at the time of writing. The 2014 letter, published in February 2015, put their value at about $5.7 billion at that later writing date while saying Dexter's value had gone to zero. These later share values measure opportunity cost at different dates. They were not known dollar losses in 1993. The case also resists a shallow rule against stock deals. Cash has an opportunity cost, debt adds fixed claims, and sellers may require shares. Issuing overvalued shares can benefit continuing owners if the target value and risks are assessed honestly. The procedure is to value both sides independently. Estimate the target under base, adverse and favorable cases. Estimate the acquirer's intrinsic value per share, multiply by shares issued, and show dilution as a percentage of the combined owner claim. Repeat the analysis without projected synergies and under competitive stress. Payment form and target quality are separate decisions. A good target can be overpaid for, and a fairly priced target can be financed dangerously. After closing, update the competitive thesis using evidence that could trigger repair, sale or closure. Dexter's history shows that recognizing the original mistake and responding to the operating decline are different jobs. It also shows why a rising acquirer share price can make an old stock issuance more costly than its closing-date dollar label suggests. The share count preserves that ownership transfer even when the original deal value fades from memory. ### Work the question In a fictional acquisition proposed on September 30, 2026, an acquirer offers one million shares trading at $500 each for a target valued at $520 million, all in United States dollars. Why is comparing $500 million with $520 million insufficient? ### Compare with the guide’s reasoning The market quote may differ from the acquirer's intrinsic value, target valuation may omit adverse cases, and the exchange transfers a percentage of future ownership. Value both businesses, dilution, synergies, taxes, financing alternatives and downside before deciding. ### Sources - Warren E. Buffett, 1993 Chairman's Letter (1993 reporting period; published 1994), Dexter Shoe. https://www.berkshirehathaway.com/letters/1993.html Evidence note: Contemporary, favorable account after the November 7, 1993 merger. It is useful evidence of the case made then, not proof that all deliberations were disclosed. - Warren E. Buffett, 2001 Chairman's Letter (2001 reporting period; published 2002), Non-insurance operating highlights, shoe operations and Dexter. https://www.berkshirehathaway.com/2001ar/2001letter.html Evidence note: Retrospective self-assessment identifying the purchase, use of stock, and delayed operational response as three separate errors. - Warren E. Buffett, 2007 Chairman's Letter (2007 reporting period; published 2008), Businesses, The Great, the Good and the Gruesome. https://www.berkshirehathaway.com/letters/2007ltr.pdf Evidence note: Retrospective comparison of See's, FlightSafety, capital-intensive growth, and Dexter. The Dexter share value is measured at the letter's February 2008 publication, not in the 2007 reporting year or in 1993. - Warren E. Buffett, 2014 Chairman's Letter (2014 reporting period; published 2015), Berkshire, Past, Present and Future, especially the textile, See's and Dexter passages. https://www.berkshirehathaway.com/letters/2014ltr.pdf Evidence note: Fifty-year retrospective. It supplies later cumulative figures and Buffett's revised interpretation, not a contemporaneous record of the earlier decisions. The Dexter share value is measured at the letter's February 2015 publication. ## Warren Buffett · Separate growth from good reinvestment Growth can make a weak business more demanding rather than more valuable. Berkshire's 2007 letter sorted businesses by the relationship between earnings and additional capital. At one end, See's could raise earnings over time while requiring relatively little new tangible capital. In the middle, FlightSafety delivered a valuable service and possessed a competitive advantage, but growth required expensive simulators. From Berkshire's 1996 purchase through 2007, FlightSafety's pre-tax operating earnings rose from $111 million to $270 million. Net fixed assets rose by $509 million, after accounting for depreciation and capital expenditure. Buffett called the result good but far from See's-like. At the damaging end, a business grows quickly, repeatedly demands capital and earns little on it. Revenue growth then enlarges the funding problem. The categories are not moral rankings and they do not imply that capital-intensive businesses should close. Utilities and training operations can produce satisfactory returns, serve essential needs and absorb very large investment. The owner needs to price the business and its capital plan for the returns actually available. The reinvestment test works at the margin. Historical return on capital blends old projects, accounting bases and assets acquired at different prices. For a new dollar retained today, estimate the incremental cash earnings after maintenance, tax and working capital. Specify the period before capacity is productive. Include the residual value and the chance that competitors, regulators or customers capture the benefit. Compare that expected return with repaying debt, distributing cash, repurchasing shares, or investing elsewhere. Durability belongs beside return. Buffett's 2007 discussion described a strong business as one with an enduring barrier that protects returns, but it also warned that advantages based only on a brilliant manager or a product vulnerable to rapid change are harder to forecast. See's had a long-developed customer franchise. FlightSafety's advantage rested partly on the importance of training quality, yet new aircraft models required continuing simulator investment. The same headline return can therefore imply different future cash. A portfolio adds another constraint. The best internal project should compete with opportunities elsewhere, while liquidity and debt promises limit how much can be committed at once. Ranking projects only within a division can preserve a weak premise. A hurdle rate should therefore reflect financing, risk and the next-best use available to the whole owner. There is no single capital-intensity cutoff. A data centre, railway or utility can create value when demand, utilization, financing and allowed returns support the investment. A software service can destroy value while appearing asset light if customer acquisition and product renewal consume cash that accounting treats as current expense. The practical discipline is to trace retained cash into identifiable projects, then compare forecast incremental cash with what later arrived. The purpose is accountability. It prevents managers from defending every expansion with revenue growth and prevents investors from rejecting all heavy investment without examining its price and return. ### Work the question In a fictional five-year capital plan dated December 31, 2026, a division can add $20 million of annual pre-tax earnings after investing $250 million in United States dollars. What comparison determines whether growth helps owners? ### Compare with the guide’s reasoning Estimate the incremental cash return, timing, durability and downside on the $250 million, then compare it with the company's cost of capital and alternative uses. Also separate maintenance spending from the growth project and test whether competitors or customers capture the benefit. ### Sources - Warren E. Buffett, 2007 Chairman's Letter (2007 reporting period; published 2008), Businesses, The Great, the Good and the Gruesome. https://www.berkshirehathaway.com/letters/2007ltr.pdf Evidence note: Retrospective comparison of See's, FlightSafety, capital-intensive growth, and Dexter. The Dexter share value is measured at the letter's February 2008 publication, not in the 2007 reporting year or in 1993. - Warren E. Buffett, 1986 Chairman's Letter (1986 reporting period; published 1987), Purchase-Price Accounting Adjustments and the owner-earnings discussion near the end of the letter. https://www.berkshirehathaway.com/letters/1986.html Evidence note: Explains Buffett's valuation adjustment using Scott Fetzer. Maintenance capital expenditure is explicitly an estimate, not an audited line item. ## Jeff Bezos · Declare the tradeoffs before the pressure arrives Amazon's 1997 shareholder letter did more than announce long-term thinking. It stated how management intended to choose when attractive goals conflicted. Amazon would prioritize long-term market leadership over short-term profit or market reaction, invest aggressively in customers, brand and infrastructure, measure programs, stop those with unacceptable returns, and increase investment in those that worked. It would prefer the present value of future cash flows to a better-looking accounting result. The letter also said this was Amazon's philosophy rather than a universal one. The policy was tied to evidence available in 1997. Sales rose from $15.7 million in 1996 to $147.8 million, an 838 percent increase. Cumulative customer accounts rose from 180,000 to 1.51 million. Repeat-customer orders exceeded 58 percent in the fourth quarter, up from more than 46 percent a year earlier. Amazon expanded distribution space from 50,000 to 285,000 square feet, increased inventory to more than 200,000 titles, and ended with $125 million of cash and investments after its initial public offering and a $75 million loan. These measures showed demand and repeat use, while the infrastructure figures showed the cost and execution burden of serving it. The decision was not simply growth at any price. The letter paired bold investment with analytical review, cost consciousness, long-term profitability and capital management. It named competition, operational strain, expansion risk and the need for continuing investment. That combination is the useful part. A time horizon does not excuse weak unit economics. A customer metric does not automatically become shareholder value. Management needs a proposed chain from customer benefit to repeat behavior, operating efficiency and cash, plus evidence that could break the chain. The 1998 follow-up reported sales of $610 million, 6.2 million cumulative customer accounts and repeat orders above 64 percent in the fourth quarter. It also reported $31 million of operating cash flow and $28 million of net fixed-asset additions, while inventory rose from $9 million to $30 million. Those results are consistent with continued customer growth and capital efficiency in that period. They do not prove that aggressive reinvestment caused the outcome, nor do they reveal the return on every program. To adapt the policy, write a decision constitution before a contested investment. Name the objective, the financial measure, the customer evidence, the review interval and the stop condition. Specify which near-term result may worsen and why. Then state a liquidity boundary that cannot be crossed. This turns patience into a testable commitment. It also lets owners decide whether the declared tradeoff fits their own time horizon before results arrive. Without review and stop rules, long-term language can protect waste. Without a longer horizon, a firm may reject investments whose customer and cash effects take time to appear. ### Work the question What turns a long-term investment policy from an excuse into a decision system? ### Compare with the guide’s reasoning It needs a causal link to customer and cash outcomes, program-level measures, review dates, stop or scale criteria, and a liquidity boundary. It should also name the short-term metric management is willing to sacrifice and why. ### Sources - Jeffrey P. Bezos, 1997 Letter to Shareholders (1997 reporting period; published 1998), It's All About the Long Term, Obsess Over Customers, Infrastructure, and Goals for 1998. https://www.aboutamazon.com/news/company-news/amazons-original-1997-letter-to-shareholders Evidence note: Contemporary statement of management policy and 1997 results. Management's claims about market leadership and future economics are its judgments at the time. - Jeffrey P. Bezos, 1998 Letter to Shareholders (1998 reporting period; published 1999), A Recap of 1998, Our Customers, and Goals for 1999. https://s2.q4cdn.com/299287126/files/doc_financials/annual/Shareholderletter98.pdf Evidence note: Contemporary follow-up reporting customer, revenue, repeat-order, cash-flow, inventory and infrastructure figures. The outcome does not isolate the effect of the 1997 policy. ## Jeff Bezos · Map the loop and test every link Amazon's 2001 letter described a reinforcing operating cycle. Lower cost could fund lower prices. Lower prices and better selection or convenience could attract more demand. Greater volume could spread fixed costs and improve purchasing or operating efficiency, creating room for further price reduction. The attraction of this model is compounding. The danger is drawing arrows that have not been measured. Each link can fail, arrive late, or benefit another participant more than the company. The 2001 evidence came after a severe market correction and operational retrenchment. Amazon reported its first profitable quarter using the letter's pro forma measures, and it emphasized free cash flow, operating leverage and limiting share dilution. Marketplace represented 15 percent of United States orders in the fourth quarter. Inventory turns improved from 12 to 16 for the full year. These facts made the cycle more than a growth slogan, yet they still represented one company's evolving system and selected measures. The 2002 letter examined a specific alleged tradeoff. Traditional stores often spent more to provide personal service, while discounters removed service to lower price. Amazon argued that software features, product information and reviews were largely fixed costs that could improve with scale while shrinking as a percentage of sales. It reported an American Customer Satisfaction Index score of 88, a 13 percent improvement in customer-service contacts per order, and a price check on 100 bestselling books. In that basket, Amazon said it was cheaper on 72, equal on 25 and initially more expensive on three, which it then reduced. The letter identifies the list as the other retailer's 100 bestsellers for 2002 but gives no exact store-visit date. Free cash flow under Amazon's stated definition was $135 million in 2002, compared with negative $170 million in 2001. The evidence has limits. A bestseller basket does not establish all prices. Satisfaction does not isolate which feature caused repeat use. Fixed software costs are not permanently fixed, because systems need engineers, security and capacity. Scale can also create congestion, bureaucracy or service failures. A cycle becomes useful when every arrow carries a measure and a lag. Price may affect conversion this week, repeat purchase over months and supplier terms after a volume threshold. Build the map from a customer's changed behavior backward. State who receives value and who pays for it. For each link, choose a leading measure, an economic measure and a failure condition. Separate a stock from a flow. A large installed customer base is a stock, while new repeat orders are a flow. Finally, look for balancing loops such as returns, working capital, seller dissatisfaction or delivery congestion. The goal is not to admire a flywheel. It is to find the weakest link before capital is committed and to update the model when evidence does not travel around the loop. ### Work the question A team says lower prices will drive scale and scale will lower cost. What must it specify before this is an investable model? ### Compare with the guide’s reasoning It must quantify the price response, contribution per order, fixed and variable costs, capacity and working-capital needs, timing of scale economies, who captures them, and thresholds that would show a link is failing. ### Sources - Jeffrey P. Bezos, 2001 Letter to Shareholders (2001 reporting period; published 2002), Opening discussion of price, cost reduction, growth, fixed costs and cash flow. https://ir.aboutamazon.com/files/doc_financials/annual/2001_shareholderLetter.pdf Evidence note: Management's contemporary explanation of a reinforcing operating cycle after reporting its first profitable quarter on the letter's pro forma measures. - Jeffrey P. Bezos, 2002 Letter to Shareholders (2002 reporting period; published 2003), Opening discussion of customer experience, price, satisfaction, the 100-book comparison, and free cash flow. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2002_shareholderLetter.pdf Evidence note: Contemporary company evidence. The 100-book basket, satisfaction score and management's free-cash-flow definition have specific scopes and do not prove a universal price rule. ## Jeff Bezos · Make growth pass through cash and share count The 2004 Amazon letter argued that earnings growth can destroy value when the investment required exceeds the present value of the cash it produces. Its fictional transportation machine made the arithmetic visible. Each machine cost $160 million, carried 100,000 trips a year for four years, and produced $10 million of accounting earnings at full capacity in the first year. Doubling capacity repeatedly made earnings rise from $10 million to $80 million across four years. Cumulative earnings were $150 million. Cash told a different story. The example added back depreciation, then subtracted machinery purchases. Free cash flow was negative $110 million, negative $60 million, negative $120 million and negative $240 million in the four years, for cumulative negative $530 million. Capital expenditure totalled $1.28 billion. EBITDA also rose rapidly, but it excluded the machines required to generate the trips. Under the assumptions, slower growth reduced the cash deficit. The example was intentionally simple and even its no-growth case had a negative present value at the stated 12 percent discount rate. The point was not that growth is bad. It was that revenue and earnings cannot replace investment economics. Amazon then connected the lesson to free cash flow per share. Cash from operations less fixed-asset purchases was the company's stated free-cash-flow measure. Per-share analysis also required attention to stock awards, options and convertible debt. In 2004 Amazon reported $477 million of free cash flow. Its footnote separately displays rounded operating cash flow of $567 million and rounded fixed-asset purchases of $89 million. Those displayed components subtract to $478 million, so the reported $477 million depends on underlying amounts not shown at the same precision. The footnote included capitalized internal software and website development in fixed assets. Amazon also reported $480 million of year-end inventory on nearly $7 billion of sales, aided by collecting from customers before supplier payments were due. Free cash flow still needs interpretation. Supplier payment timing can improve current cash while creating an obligation. Leasing can move investment away from a capital-expenditure line. Stock compensation dilutes owners even when it does not use current cash. A growing business may rationally invest ahead of demand, so a negative year is not a verdict. Conversely, positive cash produced by underinvesting can precede decline. A consistent definition across periods matters as much as the headline total. Use a multi-year bridge. Start with operating profit after cash tax. Add and subtract working-capital changes by cause. Separate maintenance, committed growth and optional growth capital. Include leases, acquisitions and dilution when they are part of the economic funding. Forecast a base, downside and capacity-constrained case, then discount the cash per current and likely future share. The exercise forces two questions. How much cash can this model eventually release, and how much ownership must be issued or retained to reach that point? ### Work the question In the 2004 letter's fictional four-year transportation-machine exercise, why can 100 percent annual earnings growth coexist with value destruction? ### Compare with the guide’s reasoning If each increment of earnings requires capital whose cost exceeds the present value of the resulting cash, growth enlarges the loss. Working capital, capital expenditure, financing and dilution must be included. ### Sources - Jeffrey P. Bezos, 2004 Letter to Shareholders (April 2005), Our Most Important Financial Measure, Free Cash Flow Per Share, including the transportation-machine example. https://ir.aboutamazon.com/files/doc_financials/annual/2004_Annual_report.pdf Evidence note: Uses a deliberately simplified fictional machine to separate earnings growth from cash generation, followed by Amazon's 2004 company figures. ## Jeff Bezos · Know what the model cannot decide Amazon's 2005 letter divided important choices by what evidence could resolve. Fulfilment-centre location and inventory placement used historical demand, product dimensions, vendor performance, transport cost and customer location. Assumptions remained, but quantitative analysis did most of the work. The firm could compare alternatives because repeated operations produced relevant data and the decision could be modelled before committing. Other choices lacked a complete historical sample. Amazon said its short-term price-elasticity analysis usually favored raising rather than lowering prices. A price reduction increased units, but generally not enough in the current week or quarter to repay the lost price. Management could not numerically estimate what repeated price reductions would do to customer relationships and free cash flow over five or ten years. It chose to return efficiency and scale gains to customers, treating the long-run effect as a judgment rather than pretending the short-run model answered it. Free Super Saver Shipping and Prime involved similar reasoning. The single product detail page created another conflict. In 2000 Amazon let third-party sellers compete with its own retail offers on valuable product pages. Internal buyers worried about cannibalization and excess inventory if a seller won the order. The customer could receive a better price or availability. By 2005, third-party units had risen from 6 percent to 28 percent of total units while Amazon retail revenue had tripled. This later evidence showed that the two channels had grown together in that period. It did not reveal the missing counterfactual or prove that every platform should expose its own offer to rivals. Judgment does not mean intuition without discipline. State which part the model answers, which part it cannot, and why new data cannot be obtained without partly making the decision. Use base rates from genuinely comparable settings. Bound the loss if the thesis is wrong. Identify a leading indicator that should move before the long-run cash result, and pre-commit to a review. A price cut might be reversible. A nationwide fulfilment network or irreversible brand promise is less so. Different parts of one commitment can also receive different methods. Model site capacity closely, pilot the customer offer where possible, and reserve the broad rollout for evidence from the pilot. The distinction also guards against false quantification. A spreadsheet can calculate an assumed ten-year retention benefit to two decimal places without making the assumption reliable. At the same time, declaring a choice strategic cannot exempt it from arithmetic. Scenario ranges should expose the break-even retention or cash effect needed to justify the decision. Amazon's stated practice joined a quantitative operating culture with willingness to make contested judgments. The transferable skill is matching the decision method to the evidence, while keeping both the measurable downside and the unmeasured thesis visible. ### Work the question When should a team rely on judgment despite a model pointing the other way? ### Compare with the guide’s reasoning When the model is reliable for a narrower horizon or repeated operation but omits a material effect that cannot be measured before acting. The team should name that omission, bound downside, seek comparable evidence, stage the commitment where possible, and set a review trigger. ### Sources - Jeffrey P. Bezos, 2005 Letter to Shareholders (2005 reporting period; published 2006), Opening discussion of math-based and judgment-based decisions, price reductions, and the single detail page. https://ir.aboutamazon.com/files/doc_financials/annual/AMZN2005AnnualReport.pdf Evidence note: Contemporary account of Amazon's decision policy with retrospective results for the Marketplace choice made in 2000. Long-run price effects were explicitly not measurable in advance. - Jeffrey P. Bezos, 2014 Letter to Shareholders (2014 reporting period; published 2015), Opening business-quality test and the Marketplace, Prime, FBA and AWS sections. https://ir.aboutamazon.com/files/doc_financials/annual/AMAZON-2014-Shareholder-Letter.pdf Evidence note: Retrospective account of several initiatives after they had shown traction. Management's business-quality assessments and causal explanations remain judgments. ## Jeff Bezos · Manage controllable inputs without protecting failed projects Amazon's 2009 planning account illustrates a shift from commanding outputs to assigning inputs. For 2010, management had 452 detailed goals with owners, deliverables and dates. It said 360 directly affected customer experience. Revenue appeared eight times and free cash flow four times, while net income, gross profit, margin and operating profit did not appear. These were the goals senior leaders considered most important to monitor, rather than every goal set across the company. Bezos said leaders still took financial results seriously but believed controllable inputs were the most effective focus. This does not mean money disappears from management. Revenue and profit are outcomes of price, selection, availability, delivery, cost and customer behavior. Teams can act on those inputs today. A goal to reduce delivery defects or add in-stock selection is more operational than a command to grow sales. The causal link still has to be tested. If customer-oriented activity never improves retention, contribution or cash, its label cannot save it. Some financial constraints also operate as inputs, including budget, unit cost and available capital. The 2014 letter shows how inputs can connect businesses. Amazon's Marketplace initially followed unsuccessful Auctions and zShops attempts. The single detail page allowed third parties to add selection and compete for orders. Fulfilment by Amazon then made seller inventory eligible for Prime, which could add seller sales and make Prime more valuable through greater selection. The account identifies a proposed loop among sellers, members and Amazon. It also acknowledges that Marketplace, Prime and AWS began as risky bets and needed continued investment after success. The 2015 letter added two controls for experimentation. First, a portfolio with rare very large wins can tolerate many bounded failures, but the stated ten-percent chance of a hundred-times payoff was an illustration, not Amazon's measured hit rate. Second, reversible decisions should usually move faster than decisions that are difficult to reverse. The classification depends on real switching cost, legal commitment, customer harm, data migration, capital and reputation. Calling an initiative reversible does not make it so. Portfolio logic also requires common loss limits and honest accounting for abandoned work. Otherwise one visible winner can be used to excuse unrelated failures. A practical system links four layers. State the financial objective and constraint. Choose controllable customer and operating inputs with named owners. Run bounded experiments where the downside is recoverable. Review whether input movement produces the expected intermediate and financial outputs. Stop, revise or scale based on the chain, not on the prestige of the project. Compare input quality as well as quantity. More selection can help customers, while duplicate or unreliable listings can make the same input measure worse. This avoids two opposite errors. Managing only quarterly outputs can encourage shortcuts, while celebrating inputs and experimentation can let unsuccessful programs survive without evidence. ### Work the question How can a company focus goals on customer inputs without losing financial discipline? ### Compare with the guide’s reasoning Tie each input to a causal hypothesis and intermediate measure, keep explicit budget and cash constraints, assign owners and dates, and review whether the expected customer and financial outputs follow. Stop or revise when the chain fails. ### Sources - Jeffrey P. Bezos, 2009 Letter to Shareholders (April 2010), Opening results and the 452 goals for 2010. https://s2.q4cdn.com/299287126/files/doc_financials/annual/AMZN_Shareholder-Letter-2009-(final).pdf Evidence note: Reports the content of one annual planning set. It shows Amazon's chosen management emphasis, not proof that financial outputs can be ignored. - Jeffrey P. Bezos, 2014 Letter to Shareholders (2014 reporting period; published 2015), Opening business-quality test and the Marketplace, Prime, FBA and AWS sections. https://ir.aboutamazon.com/files/doc_financials/annual/AMAZON-2014-Shareholder-Letter.pdf Evidence note: Retrospective account of several initiatives after they had shown traction. Management's business-quality assessments and causal explanations remain judgments. - Jeffrey P. Bezos, 2015 Letter to Shareholders (2015 reporting period; published 2016), Failure, long-tailed business returns, and Type 1 versus Type 2 decisions. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2015-Letter-to-Shareholders.PDF Evidence note: Retrospective management account. The ten-percent and one-hundred-times illustration explains an asymmetric payoff shape, not a measured base rate for Amazon experiments. ## Howard Marks · A market price is evidence, not a verdict In March 2008, Marks described a credit system reversing under its own financing structure. Falling asset prices reduced borrowers' equity, which raised leverage ratios, prompted margin calls, forced sales, and produced still lower prices. This was more than a mood swing. The same price that normally helps investors coordinate could become an output of urgent liquidity needs. A forced seller was answering a different question from a patient owner. The seller needed cash now. The owner wanted to know what future payments the claim could support. Treating the forced-sale quote as a complete verdict confused those two questions. That distinction does not make price irrelevant. It changes the work required. A low quote can reveal impaired cash flows, a fragile borrower, or information the buyer has missed. It can also reveal that potential buyers lack financing or refuse to take risk. Marks's 2008 argument was that supply and demand had temporarily overwhelmed fundamental analysis in some markets. The learner's task is therefore diagnostic. Ask what transaction produced the quote, who had to act, what financing constraint operated, and whether the holder can wait for the asset's economics to emerge. A price dislocation is only investable if the buyer can survive the path between purchase and realization. The mechanism creates two separate estimates. One is fundamental value under stated operating and default assumptions. The other is the range of prices that market structure may produce before those assumptions are resolved. A sensible decision needs both. Buying merely because price fell delegates judgment to the previous high. Refusing to buy merely because price is falling delegates judgment to the current crowd. Marks's September memo made skepticism symmetrical. Investors who had failed to question good news during the boom could make the same error by failing to question bad news during the panic. There is a hard boundary. An apparently distressed price may be accurate because the asset is permanently impaired. Marks's October buying report depended on his colleagues being right about value, not on contrarianism alone. A cheap-looking claim, financed with unstable debt, can still destroy the buyer before value is realized. The transferable procedure is to write down the cash-flow thesis, the evidence that could falsify it, the liquidity path, and the financing terms before treating a falling price as opportunity. Then distinguish a changed value estimate from a changed ability to hold. That separation is what allows price to inform judgment without replacing it. A useful decision memo should record both ranges before the trade and update them independently afterward. If the market price falls while the value evidence is unchanged, the prospective return may improve. If both fall, the apparent discount may not change at all. This record also limits hindsight. It shows whether the buyer recognized a liquidity mechanism before the rebound or merely supplied one after seeing the outcome. ### Work the question In this fictional 2026 exercise, a bond falls from US$92 to US$58 per US$100 of par value while several levered funds face redemptions. What would you need to know before calling US$58 a bargain, and which observations would cause you to conclude that value rather than liquidity had deteriorated? ### Compare with the guide’s reasoning Separate the issuer's payment capacity from the sellers' need for cash. Rebuild the value range from cash flows, covenants, seniority, and default scenarios. Investigate who is selling, whether trades are forced, and whether comparable assets are clearing. Then test your own funding and holding period. A bargain conclusion should fail if issuer economics, claim priority, or recovery evidence worsens enough to justify the price. ### Sources - Howard Marks, The Tide Goes Out (2008-03-18), Leveraging and Inflating; Mark-to-Market Accounting; Should Does Not Equal Will; Leverage and Risk. https://www.oaktreecapital.com/insights/memo/the-tide-goes-out Evidence note: Contemporaneous analysis written before the worst phase of the financial crisis. Examples and conclusions express the author's assessment and are not an independent performance study. - Howard Marks, Nobody Knows (2008-09-19), Does the Market Know?; The End of the Financial System; final bear-market-stage discussion. https://www.oaktreecapital.com/insights/memo/nobody-knows Evidence note: Contemporaneous memo after Lehman's bankruptcy. It records a decision premise and a search for bargains, but no security-level trade list. - Howard Marks, The Limits to Negativism (2008-10-15), Final discussion of third-stage pessimism and Oaktree purchases. https://www.oaktreecapital.com/insights/memo/the-limits-to-negativism Evidence note: Contemporaneous practitioner report. It describes purchase direction and averaging down, but not positions, allocations, or later investment results. ## Howard Marks · Survival changes the opportunity set Marks's 2008 leverage analysis turns risk from a volatility statistic into a path problem. Leverage magnifies gains and losses, but it also adds an asymmetry. A favorable outcome can increase wealth. A sufficiently adverse interim move can end the strategy through a margin call, covenant breach, or loss of financing before the favorable outcome arrives. The relevant question is not only whether the asset will eventually pay. It is whether the chosen ownership structure can live long enough to collect. His example of Carlyle Capital made the arithmetic concrete. The fund held about $21.7 billion of securities against roughly $670 million of equity. Those figures imply assets equal to about 32.4 times equity and debt equal to about 31.4 times equity. Marks rounded the latter relationship to 31 times in the memo. A decline near 3 percent could consume the equity before considering the complications of funding withdrawal. The underlying assets had high ratings. The financing made the position fragile. This supports a general decision distinction, though not a universal numerical rule. Asset risk and structure risk must be evaluated separately, then recombined through stress scenarios. Survival is not the same as maximum defensiveness. If a portfolio holds only cash against every imaginable disaster, it may preserve nominal capital while abandoning its purpose. Marks asks how much capital should be devoted to improbable events because there is no operational worst case short of total loss. A useful answer starts with consequences. Which events would merely hurt reported returns? Which would force selling? Which would impair the institution's obligations or prevent it from acting when bargains appear? The most important protection is aimed at the last two categories. This reframes dry powder. Cash is not automatically prudent, and full investment is not automatically aggressive. What matters is whether the portfolio combines durable financing, assets whose downside has been considered, and enough flexibility to respond when others lose theirs. In the 2020 retrospective, Marks said Oaktree's cautious pre-crisis posture reduced remediation work and left some drawdown funds able to buy in March. That is a practitioner account rather than controlled evidence, but it illustrates the mechanism. Preparation created decision capacity at the moment prices changed. A transferable test has four steps. Identify the adverse path rather than only the terminal case. Map every forced-action trigger, including collateral calls, redemptions, covenants, and operating cash needs. Estimate how much room exists before each trigger. Finally, compare the expected gain from added exposure with the loss of future choice if the path is worse than expected. The objective is not to avoid every loss. It is to prevent a temporary or tolerable loss from taking away the ability to decide. Stress tests should therefore report time as well as loss. A 25 percent decline that arrives after five years of cash generation may be manageable. The same decline next week may trigger lenders and redemptions. Adding dates, collateral terms, and decision rights to each scenario exposes risks that a terminal value table can hide. ### Work the question In this fictional 2026 exercise, two portfolios own the same senior loans. One is unlevered and closed-end. The other is financed overnight with debt equal to six times equity. Explain why their risk is different even if your default forecast is identical. ### Compare with the guide’s reasoning The leveraged portfolio can face funding withdrawal and forced sale after price declines that do not change ultimate loan payments. The closed-end portfolio can wait. The second portfolio therefore adds refinancing, collateral, and path risk to the same asset risk. Its owner must analyze the trigger points and not rely only on the terminal default forecast. ### Sources - Howard Marks, The Tide Goes Out (2008-03-18), Leveraging and Inflating; Mark-to-Market Accounting; Should Does Not Equal Will; Leverage and Risk. https://www.oaktreecapital.com/insights/memo/the-tide-goes-out Evidence note: Contemporaneous analysis written before the worst phase of the financial crisis. Examples and conclusions express the author's assessment and are not an independent performance study. - Howard Marks, Coming into Focus (2020-10-13), The Power of Interest Rates; crisis comparison; final portfolio-calibration discussion. https://www.oaktreecapital.com/insights/memo/coming-into-focus Evidence note: Retrospective report on Oaktree's pre-crisis posture and March buying, followed by a dated October assessment. It does not disclose portfolio allocations. ## Howard Marks · Contrarian action needs a mechanism Being different is not an investment thesis. Marks's 2008 memos are often reduced to buying when others are fearful, but the record contains a more demanding sequence. By September, the financial system faced failures, rescues, frozen funding, and the possibility of a broader collapse. Marks admitted he could not forecast the outcome. He nevertheless rejected panic as a usable strategy because a true system-ending outcome offered few protective actions that would remain effective, while actions designed for that outcome could be ruinous if ordinary economic life continued. That premise did not prove securities were cheap. It allowed analysis to continue. Marks then asked whether widely recognized risks were already reflected in price, whether pessimism had become indiscriminate, and which claims on continuing businesses could be bought below a defensible estimate of value. By October, he reported that Oaktree's purchase lists were long on most days, sales were scarce, and available cash was being deployed while prices fell. The later 2020 memo says Bruce Karsh invested an average of $450 million per week over the final 15 weeks of 2008. That implies $6.75 billion across the period, but the total is arithmetic derived from a retrospective report, not a contemporaneous audited schedule. The decision was paced rather than certain. Marks recalls Karsh alternating between concern that Oaktree was moving too slowly and too quickly. Doubt was part of the control system. It supported repeated review of pace without forcing paralysis. Averaging down only made sense while the value thesis survived new evidence and the organization retained liquidity. If falling prices reflected new impairment, buying more would compound an error. The historical outcome cannot turn this into a formula. The memos do not disclose each investment, the allocation across clients, or the counterfactual return from waiting. They also came from the firm describing its own behavior. What can transfer is the structure of the decision. State the catastrophe premise you are accepting or rejecting. Explain why the proposed hedge works in that state. Estimate value independently of the crowd. Divide deployment into revisable increments. Define what would stop further purchases. Preserve enough capacity to act if prices become more attractive. This form of contrarianism is conditional. It does not reward disagreement for its own sake. It asks whether a dominant belief has produced a price that overstates the probability or consequence of one path. The investor earns the right to differ by identifying the mechanism, surviving if early, and naming the evidence that would make the crowd right. The stopping rule deserves the same attention as the entry rule. It can be tied to a broken covenant, a revised recovery range, loss of reliable funding, or a portfolio concentration ceiling. It should not be tied only to another price decline, because price is the disputed signal. This makes the decision capable of changing for a reason rather than for relief from discomfort. ### Work the question In a fictional 2026 market shock, write a four-part case for buying into panic that would still be coherent if prices fell another 30 percent next month. ### Compare with the guide’s reasoning State why the asset can keep paying under a defined adverse case, why current sellers may be constrained, how the position is financed through a further decline, and what evidence would invalidate value. Then set a staged pace and stopping rule. A thesis that depends on an immediate rebound is not robust to the proposed test. ### Sources - Howard Marks, Nobody Knows (2008-09-19), Does the Market Know?; The End of the Financial System; final bear-market-stage discussion. https://www.oaktreecapital.com/insights/memo/nobody-knows Evidence note: Contemporaneous memo after Lehman's bankruptcy. It records a decision premise and a search for bargains, but no security-level trade list. - Howard Marks, The Limits to Negativism (2008-10-15), Final discussion of third-stage pessimism and Oaktree purchases. https://www.oaktreecapital.com/insights/memo/the-limits-to-negativism Evidence note: Contemporaneous practitioner report. It describes purchase direction and averaging down, but not positions, allocations, or later investment results. - Howard Marks, Uncertainty (2020-05-11), All We Don't Know; In Praise of Doubt; confidence and position-sizing discussion. https://www.oaktreecapital.com/insights/memo/uncertainty Evidence note: The discussion of 2008 includes a retrospective practitioner report. The memo distinguishes evidence-based confidence from certainty. ## Howard Marks · Judge your confidence as well as your forecast A probability estimate can look disciplined while hiding a weak evidence base. In May 2020, Marks separated two judgments that decision makers often collapse. The first is an opinion about what will happen. The second is an assessment of how dependable that opinion is. A 60 percent estimate built from stable frequencies and repeated observations is different from a 60 percent estimate assembled from a unique shock, interacting systems, and analogies that do not fit. The distinction matters because position size converts belief into consequence. If the evidence is thin, the range of plausible probabilities should be wide. A decision can still be made, but it should place less weight on the fragile estimate and preserve more room to revise. Marks applied this caution most strongly to broad variables such as economies, markets, currencies, and interest rates, where many interacting actors and feedback loops frustrate precise prediction. He did not argue that all knowledge is impossible. He argued that confidence should match the domain and the evidence. There is another trap. Consensus forecasts are often approximately right because recent patterns often continue, but the expected continuation is normally embedded in prices. A surprising forecast is more valuable only if it is also correct, which is hard. The practical alternative is not to abandon judgment. It is to stop demanding that one macro forecast carry the whole decision. Analyze the asset's contractual terms, competitive position, financing, and price across several macro paths. Ask whether the investment requires a narrow forecast or can work under a range. Intellectual humility also needs a boundary. Low confidence can become a performance of caution that avoids accountability. A decision maker who says anything could happen and never specifies odds, exposure, or revision rules has not managed uncertainty. The stronger practice records a base case, adverse cases, confidence level, and the evidence that would change each. It also asks whether an opposing analyst has access to different facts or merely a different temperament. The decision procedure is to audit both layers. First, list the claims required for the investment to work. Second, classify the evidence behind each claim as direct fact, inference from a relevant base rate, or speculation. Third, estimate how an error in each claim affects value and liquidity. Fourth, size the decision to the weakest consequential claim. This converts humility from a tone of voice into a constraint on exposure. It also makes updating visible. When new evidence arrives, the decision maker can revise the claim, the confidence placed on it, or both. Calibration can be reviewed after the fact without pretending that one result settles it. Group past forecasts by stated confidence and compare how often they occurred, while keeping unique events separate from repeatable classes. The purpose is not to manufacture precision. It is to detect a consistent gap between the certainty expressed and the evidence that was actually available. ### Work the question In a fictional 2026 case, you assign a 65 percent chance to a rapid industry recovery, but the industry has never experienced the current shock. How should that affect a capital decision? ### Compare with the guide’s reasoning Treat 65 percent as a fragile judgment rather than a frequency-backed estimate. Test the investment across slower paths, reduce dependence on the recovery date, and size exposure so an error does not force action. Record what observations would raise or lower confidence instead of treating the single probability as settled. ### Sources - Howard Marks, Uncertainty (2020-05-11), All We Don't Know; In Praise of Doubt; confidence and position-sizing discussion. https://www.oaktreecapital.com/insights/memo/uncertainty Evidence note: The discussion of 2008 includes a retrospective practitioner report. The memo distinguishes evidence-based confidence from certainty. ## Howard Marks · Prepare, act, and recalibrate The 2020 memos show a decision process changing as price and evidence changed. On March 31, after a 33.9 percent S&P 500 decline followed by a sharp three-day rise, Marks laid out both a positive and negative case. The health path, policy response, economic damage, defaults, and market psychology could interact in ways that recent history did not resolve. He judged that prices allowed too little room for worse news and emphasized readiness to respond to another decline. Read alone, that judgment might sound like a forecast to stay out. The October account adds the operating context. Oaktree had entered the shock more cautious than usual while still trying to remain invested. Marks says that posture reduced forced remediation and allowed drawdown funds with available capital to buy when bargains were most available in March. The record does not give allocations or claim that every strategy did the same thing. It does show that a cautious prior position and selective buying could coexist. By the June 18 memo, the S&P 500 had closed on June 17 at 3,113, about 8 percent below its February high of 3,386. Marks concluded that listed prices reflected optimistic expectations and that prospective upside did not adequately compensate for disappointment or multiple contraction. In October, he described the March bargain period as brief. Near-zero policy rates had lowered prospective returns and encouraged investors to take more risk. His list of portfolio choices contained costs in every direction. Cash sacrificed return, added risk offered no certainty, and special niches added illiquidity and manager risk. The lesson is not a dated asset-allocation recommendation. It is a cadence. Preparation occurs before the shock through financing, liquidity, and portfolio quality. Action occurs when price changes the prospective return enough to compensate for uncertainty. Recalibration occurs after price moves, even if the underlying story remains unsettled. The same security can move from unattractive to attractive and back without the decision maker claiming to forecast the whole economy. A practical review can be run at every material price change. Update the value range without changing assumptions merely to follow the market. Record which risks are now reflected in price and which are still ignored. Compare the next dollar of exposure with the optionality it consumes. Check whether prior purchases have created concentration or funding pressure. Then choose among adding, holding, trimming, or preserving capacity. This process prevents a good initial decision from becoming a permanent identity. A buyer in March does not owe the market bullishness in June. The review should also compare opportunities, not only each asset with cash. A recovered public security may offer less compensation than a still-dislocated private claim, yet the latter may consume more liquidity. Ranking the next dollar by prospective return, downside, evidence quality, and optionality makes recalibration a portfolio decision rather than a judgment about whether the last trade was right. ### Work the question In a fictional 2026 case, a security you bought after a 35 percent fall recovers almost to its old price while the fundamental range remains wide. What should you recalculate before deciding to add, hold, or trim? ### Compare with the guide’s reasoning Recalculate prospective return from today's price, not the purchase price. Update the value distribution, downside under adverse paths, risks already reflected, concentration, funding capacity, and the value of preserving cash. The earlier bargain decision does not determine the current one. ### Sources - Howard Marks, Which Way Now? (2020-03-31), The Positive Case; the negative case; The Government Programs; Summing Up. https://www.oaktreecapital.com/insights/memo/which-way-now Evidence note: Dated scenario analysis and author judgment during the pandemic shock. It is not an Oaktree transaction ledger. - Howard Marks, The Anatomy of a Rally (2020-06-18), Market chronology; psychology of the rally; final assessment questions. https://www.oaktreecapital.com/insights/memo/the-anatomy-of-a-rally Evidence note: Contemporaneous market assessment after the spring rally. The judgment is not evidence of later excess returns. - Howard Marks, Coming into Focus (2020-10-13), The Power of Interest Rates; crisis comparison; final portfolio-calibration discussion. https://www.oaktreecapital.com/insights/memo/coming-into-focus Evidence note: Retrospective report on Oaktree's pre-crisis posture and March buying, followed by a dated October assessment. It does not disclose portfolio allocations. ## Mark Leonard · Permanent ownership distributes capital work Constellation's model created a scaling problem inside its success. Small vertical-market software businesses could generate cash while requiring limited tangible investment. If head office remained the only buyer and reviewer, capital would accumulate faster than a small central team could deploy it. Leonard's 2015 letter describes the response. Acquisition work was pushed toward business units, with analysis and structuring support available to managers who would allocate capital part time. At December 31, 2015, the company reported 182 business units, 158 business-unit managers, and six operating groups. Delegation was not a declaration that every manager could buy anything. It separated local knowledge from portfolio control. Business-unit leaders knew customers, products, competitors, and adjacency opportunities. Senior acquisition leaders could challenge assumptions, compare proposals with company base rates, and concentrate on large or borderline decisions. Leonard imagined increasing annual acquisition capacity from about 30 to 100, but that was an aspiration, not a proven production rate. The later letter reports 40 acquisitions in 2016. Permanent ownership changed the economics of this distribution. A buyer expecting to hold indefinitely could tolerate higher transaction effort on small acquisitions if the return remained attractive over a long period. It could also preserve product and customer knowledge in stand-alone units instead of forcing every target into a common operating system. Leonard argued that autonomy and responsibility helped attract and motivate managers. That is management's view, not causal proof. Autonomy can also duplicate functions, reduce purchasing scale, and make weak practices harder to detect. The control loop therefore matters more than the organization chart. A delegated allocator needs a defined mandate, access to comparable outcomes, a review of assumptions before commitment, and later evidence on what happened. Large commitments require tighter ownership because one mistake can overwhelm the learning value. Small borderline commitments may be useful experiments when their loss is bounded and the organization actually studies them. A transferable design starts by locating information. Put the decision close to the person with specific knowledge, but set the capital limit at the level that can absorb an error. Define which assumptions require independent challenge. Assign one named owner for the whole decision, including integration and outcome. Feed results back into the base rates used by the next allocator. This produces distributed judgment rather than distributed permission. The distinction protects the speed and motivation of autonomy while preserving the firm's ability to learn across many units. A distributed system also needs a rule for knowledge that crosses unit boundaries. The parent can maintain comparable data, specialist support, and a forum for challenge without taking over the local decision. Central resources should be judged by whether managers voluntarily use them and whether forecast errors narrow. That gives shared capability a measurable purpose instead of allowing it to become permanent overhead. ### Work the question A group wants local managers to source and approve small acquisitions. Design the minimum control loop that preserves local knowledge without turning approval into a blank cheque. ### Compare with the guide’s reasoning Set capital and risk limits, require a named investment owner, specify assumptions that receive independent challenge, provide comparable base rates, and schedule a post-acquisition review. Escalate large or unusual deals. Feed actual outcomes into future decisions so delegation creates shared learning. ### Sources - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. - Mark Leonard, Constellation Software Inc. 2017 President's Letter (2018-04-20), Performance metrics; retained-capital policy; perpetual-owner objective; business-unit count. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-april-2018-final.pdf Evidence note: Management report. Statements about scalability and governance are Leonard's judgments, not independent causal findings. ## Mark Leonard · A hurdle rate shapes the pipeline A hurdle rate appears to be a number at the end of a model. Leonard's letters treat it as an organizational signal that changes which opportunities are pursued, how forecasts are argued, and how much cash is allowed to remain idle. In the 2015 letter, CSI compared the weighted expected IRRs of acquisition vintages from 1995 to early 2015 with the hurdle in force. Leonard reports that expected returns tended to move toward a newly changed hurdle, including after the rate was lowered. He called this a magnetic effect. The evidence is internal and expected IRR is partly forecast, so it cannot prove what realized returns would have been under another policy. The mechanism is still useful. A lower threshold does more than approve a few marginal deals. It changes search incentives and negotiation posture across the pipeline. Teams spend time on opportunities that previously would have been rejected. Optimistic assumptions face less pressure. Sellers and brokers may learn that the buyer can pay more. A higher hurdle can protect discipline, but it can also leave capital unused and prevent scale from developing. The control must be evaluated with both false positives and false negatives. The 2016 letter provides a concrete decision. Leonard worked on a large transaction and felt commitment grow as time and effort accumulated, even though the information did not improve. The investment fell just short of the hurdle. CSI tried to negotiate a structure that would add a few points of projected IRR, failed, and walked away. The exact target, hurdle, and model are undisclosed. What is visible is the role of the pre-existing rule. It gave the team a stopping point when sunk effort created pressure to proceed. CSI did not apply one mechanical response to every deal. The letter says a small borderline proposal might proceed as a bounded lesson for local staff or senior reviewers. A very large acquisition required a single experienced leader to own diligence, structure, negotiation, integration, and outcome. Size altered the cost of error and therefore the acceptable learning method. A practical hurdle policy should state the return measure, period, financing assumption, and treatment of terminal value. It should also define a sensitivity range, evidence classes, and escalation by loss size. Review rejected deals as well as completed ones, or the firm will only learn from what it bought. Finally, watch pipeline behavior after changing the threshold. If forecast returns cluster precisely at the new minimum, that may reveal adaptation by analysts rather than a sudden change in opportunity quality. The rate also needs an owner and a review date. If nobody is accountable for its portfolio effect, deal teams can treat it as an obstacle while the board treats it as protection. A periodic review should compare deployed capital, rejected opportunities, forecast drift, realized cash flows, and concentration before deciding whether the original threshold still serves the objective. ### Work the question Your acquisition forecasts cluster just above a newly lowered hurdle. Give two competing explanations and an investigation that could distinguish them. ### Compare with the guide’s reasoning The opportunity set may have improved, or teams may have adjusted assumptions and bids to clear the new minimum. Compare assumption distributions, purchase multiples, rejected deals, and later operating outcomes before and after the change. Review who changed forecasts and whether external deal quality moved independently. ### Sources - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. - Mark Leonard, Constellation Software Inc. 2021 President's Letter (2021-02-15), Complete letter, especially dividend policy, large VMS acquisitions, and head-office role. https://www.csisoftware.com/wp-content/uploads/2026/04/letter-to-shareholders-february-15-2021.pdf Evidence note: Records a board policy decision and management intentions as of publication. The reported 16 percent invitation figure does not have an unambiguous denominator in the letter. It does not establish the results of later investments. ## Mark Leonard · A deal becomes a lesson only after review Acquisition volume can create the appearance of experience without creating learning. Leonard's 2016 letter says CSI tried to schedule post-acquisition reviews about a year after the initial investment. The process had begun at head office roughly four years earlier and had recently moved toward the operating groups. The purpose was to compare the decision with what happened while memories and records were still usable. The timing reveals a tradeoff. Review too early and integration noise may dominate the economics. Review too late and the people, forecasts, and decision context may be gone. A one-year checkpoint cannot settle a perpetual owner's full return, but it can test leading assumptions. Did customer attrition match the case? Were margins real after separating temporary cost cuts? Did the integration owner complete the actions used to justify price? Which forecast errors came from missing facts, and which came from motivated reasoning? Leonard also warns that a routine report can survive after its value disappears. He preferred operating groups to keep a review method only if it helped them act. That caveat matters. A central template may standardize comparisons while imposing work on units whose risks differ. Complete local freedom may preserve relevance while destroying the common language needed for base rates. The design problem is to standardize a small set of decision variables and let local teams add the evidence specific to their markets. Review must include process quality and outcome quality separately. A well-reasoned acquisition can encounter a low-probability loss. A careless acquisition can benefit from a favorable market. If the firm rewards only outcomes, managers learn to hide uncertainty and seek luck. If it rewards only process, it may excuse a model that repeatedly misses. A useful review asks whether the original information supported the choice, whether actual outcomes changed the base rates, and whether the decision rules need revision. The feedback must reach future decisions. Leonard says CSI accumulated distributions for key operating metrics and could see whether a proposed assumption sat near the middle or tail of its own history. That does not make the past a law. It makes exceptional forecasts visible and forces the sponsor to explain why this case differs. The transferable loop is decision memo, dated assumptions, named owner, scheduled review, error classification, and base-rate update. Without the final two steps, postmortems become stories. With them, the organization gains a memory that can challenge both novice optimism and senior enthusiasm. Rejected proposals belong in the same memory. If a company studies only completed deals, it cannot estimate the opportunity cost of caution or see whether rejected forecasts were systematically stronger than approved ones. A light follow-up on named rejects makes the hurdle testable from both sides. It also prevents a portfolio of winners from becoming the only evidence used to defend the process. ### Work the question Design a one-year acquisition review that can distinguish a bad process from an unlucky outcome. ### Compare with the guide’s reasoning Preserve the original facts, forecasts, alternatives, and rationale. Compare each forecast with actual leading indicators. Classify errors as missing information, weak inference, execution failure, or external surprise. Judge whether the choice was reasonable on the evidence then, and separately update base rates using what happened. ### Sources - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. ## Mark Leonard · Autonomy has an operating design and a bill Leonard's account of autonomy is more specific than leaving managers alone. CSI generally kept acquired businesses as stand-alone units, expected local managers to choose product direction, and placed many units inside operating groups that shared knowledge and capital support. In 2016, CSI reported that 35 of its 40 acquisitions involved businesses with fewer than 100 employees. Volaris and TSS sometimes divided larger units into smaller ones, while Harris kept multiple businesses in the same industry independent rather than merging them. This design deliberately surrendered obvious economies of scale. Separate sales, research, human resources, and administration can duplicate cost. Leonard's argument was that smaller units gained clearer responsibility, closer market focus, and more room for entrepreneurial managers. The 2016 letter explicitly says the human-scale idea was not universally accepted inside CSI and lacked compelling data. It was an ongoing experiment, not a law. The failure modes point to the controls. Autonomy without a measurable objective can protect weak performance. Local optimization can block a product that requires several units to cooperate. Shared services can quietly expand until the operating group becomes a second head office. Central intervention can also grow after one visible mistake and impose bureaucracy on every sound unit. The learner should treat the boundary between local and central work as a recurring allocation decision. Leonard's 2015 letter shows a related tension in incentives. When high-return units generated more cash than they could easily reinvest, CSI sometimes required them to retain capital and find acceptable acquisitions or growth initiatives. This lowered reported ROIC while increasing the managers' capital-allocation responsibility. The firm rejected capped bonuses partly because managers could shift results across periods and because changing plans could damage trust. These are CSI's reported reasons, not proof that the chosen system dominates elsewhere. A transferable autonomy charter names four things. First, specify decisions the unit owns, such as product, pricing, hiring, and bounded investments. Second, specify the few outcomes the parent monitors, including customer retention, cash conversion, and return on incremental capital. Third, define escalation triggers based on magnitude, legal exposure, or repeated miss, rather than executive preference. Fourth, price the bill. Track duplicated cost and missed cross-unit value alongside speed, manager retention, and customer fit. Autonomy earns continuation through evidence, and centralization must meet the same burden. The ownership design also affects what happens after a miss. A central team that rescues every struggling unit can weaken local accountability, while a parent that never intervenes can allow customers and employees to absorb avoidable damage. The charter should state a sequence for diagnosis, support, leadership change, and, if necessary, consolidation. That keeps intervention available without turning one exception into a standing central mandate. Clear dates and decision rights also make any intervention reviewable rather than personal. ### Work the question In a fictional 2026 case stated in US dollars, a parent company can save US$4 million annually by merging three acquired product teams, but local managers expect slower releases and higher customer attrition. How should it decide? ### Compare with the guide’s reasoning Quantify both the certain shared-cost saving and the uncertain losses in release speed, retention, and accountability. Test a bounded shared service or one reversible consolidation before imposing the model broadly. Set escalation and reversal criteria. Require centralization to prove value just as autonomy must. ### Sources - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. - Mark Leonard, Constellation Software Inc. 2017 President's Letter (2018-04-20), Performance metrics; retained-capital policy; perpetual-owner objective; business-unit count. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-april-2018-final.pdf Evidence note: Management report. Statements about scalability and governance are Leonard's judgments, not independent causal findings. ## Mark Leonard · Scale can change the capital-allocation problem CSI's capital policy evolved as its ability to generate cash outgrew its traditional acquisition channel. The 2015 letter reports that Average Invested Capital rose from US$739 million in 2014 to US$965 million in 2015, a rounded 31 percent increase. Leonard still expected deployment growth to slow. The firm had maintained its traditional hurdles for most acquisitions and used lower rates for some larger transactions, while acknowledging that deploying more capital at lower returns would probably reduce overall ROIC. By the 2017 letter, CSI's stated policy was to invest retained capital when targeted hurdles could be met and otherwise build cash as long as shareholders and the board permitted. That posture protected the threshold but left a governance question. At what point does unused cash become evidence that the old opportunity channel cannot absorb the firm's output? The February 2021 letter records a policy change. The board decided to stop special dividends except in compelling circumstances while leaving the regular quarterly dividend in place for then. Small and midsized VMS acquisitions would continue at traditional hurdles. Head office would build a dedicated team for very large VMS deals, consider lower hurdles for that segment, and explore an investment competence outside VMS. Leonard reported that less than 10 percent of the cash available to shareholders had gone to large VMS transactions and only three such acquisitions had been made in 26 years. Over the preceding five years, CSI knew of roughly 80 percent of the large VMS businesses sold, while brokers invited it into only 16 percent of the sale processes. The letter does not define whether that denominator covers all large VMS sale processes or only the roughly 80 percent CSI knew about. The choice accepted a visible tradeoff. Lowering hurdles for large acquisitions might improve access and absorb more cash, but Leonard expected one or two such deals a year to reduce return on investors' capital. Expanding beyond VMS added a different risk. The firm could confuse skill in one setting with a general acquisition capability. The letter states an intended search and organizational response. It does not establish that the later investments succeeded. The transferable procedure is to diagnose the constraint before changing the hurdle. If cash accumulates because sourcing coverage is weak, a better team may solve the problem without paying more. If the market segment is structurally too small, returning capital may be better than forcing deployment. If larger deals have lower but still attractive prospective returns, use a separate policy that reflects their risk, ownership demands, and effect on portfolio returns. Expansion into a new domain should begin with a written competence thesis, bounded commitments, and evidence gates. Scale justifies a policy review. It does not make capital deployment an end in itself. The board can make the alternatives comparable by stating the return and risk expected from each dollar retained. A special dividend has a visible opportunity cost if an attractive deal later appears. Retention has an equally real cost when managers lower standards to avoid admitting that shareholders have the better use. Naming both costs makes the policy falsifiable rather than ceremonial. ### Work the question A company cannot deploy all its cash at its historical hurdle. What diagnosis should precede a lower threshold? ### Compare with the guide’s reasoning Determine whether the constraint is sourcing capacity, process access, market size, pricing, or a real decline in opportunity quality. Compare the return from improved coverage, larger deals, new domains, and returning capital. Use separate controls for options with different risk rather than lowering one rate across the portfolio. ### Sources - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. - Mark Leonard, Constellation Software Inc. 2017 President's Letter (2018-04-20), Performance metrics; retained-capital policy; perpetual-owner objective; business-unit count. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-april-2018-final.pdf Evidence note: Management report. Statements about scalability and governance are Leonard's judgments, not independent causal findings. - Mark Leonard, Constellation Software Inc. 2021 President's Letter (2021-02-15), Complete letter, especially dividend policy, large VMS acquisitions, and head-office role. https://www.csisoftware.com/wp-content/uploads/2026/04/letter-to-shareholders-february-15-2021.pdf Evidence note: Records a board policy decision and management intentions as of publication. The reported 16 percent invitation figure does not have an unambiguous denominator in the letter. It does not establish the results of later investments. ## Comparison · A loan and an experiment need different tests Imagine two commitments that can both lose money. In one you lend to an established business. In the other you fund an experiment that may create an entirely new source of demand. Describing both as risky tells you very little. You need to know the shape of the payoff, the source of the information advantage, and the consequences of being wrong. In a plain loan, the contractual upside is limited. If the borrower does spectacularly well, your entitlement does not ordinarily grow like an equity stake. Preventing credit losses can therefore dominate the selection problem. An early experiment can have a different shape. Several small failures might be consistent with a sensible program if a feasible success can repay the whole program and if failures reveal useful information. The mere presence of failures does not establish that such an experiment program is disciplined. Neither does a clean historical record establish that a lender understood every risk it avoided. Marks's September 2023 memo explicitly distinguishes conventional bond investing from strategies that need substantial winners. He also distinguishes risk control from refusing all uncertain commitments. That distinction prevents the apparent conflict between avoiding mistakes and accepting experimental failures from becoming a slogan contest. The guide's application is to identify the payoff and mandate before selecting the evaluation rule. Write an experiment's maximum planned commitment and the observation it is meant to produce. Ask whether a negative result can arrive soon enough to prevent the larger commitment. A pilot is weak protection if you must sign the full long-term contract before observing the pilot. A small accounting budget can also conceal a large exposure if it affects customers who cannot be restored, discloses information that cannot be recovered, or interrupts an essential service. Then examine dependence. A portfolio of experiments is not diversified merely because it has many project names. They may all rely on the same distribution partner, model provider, sales assumption, or source of finance. Conversely, spending more on an existing business does not automatically make a commitment safe. Familiarity can hide a deteriorating market or make dissent socially costly. Draw the shared dependencies and specify which scenario could damage several positions at once. A useful decision can be cautious about ruin while ambitious about opportunity. Reserve enough capacity to meet obligations, contain the trial where containment is possible, and identify the evidence required for expansion. If the upside is contractually capped, ask why you should bear a large downside. If the upside is open-ended, ask what would prevent competitors, customers, suppliers, or financing costs from absorbing it. An impressive product demonstration does not answer that economic question. The modern scenario work asks you to make these distinctions without borrowing a famous person's certainty. An AI service can be an experiment, a sustaining investment, an acquisition target, or a source of risk to existing revenue. Its label does not determine the appropriate policy. Your answer should identify the commitment, the loss-bearing capacity, the information to be gained, and the conditions under which an initial trial earns further resources. ### Work the question A colleague says that a leader who celebrates failed experiments contradicts an investor who emphasizes avoiding losses. Reconstruct a setting in which both approaches are defensible, then name a condition that would make each approach dangerous. ### Compare with the guide’s reasoning Distinguish the payoff of ordinary credit from the payoff of an experimental equity-like commitment, and distinguish loss control from a promise of no losses. Experimentation is dangerous when failures are not bounded or informative, or when the organization cannot survive them. Avoiding losses is dangerous when it becomes refusal to bear any uncertainty while still demanding returns that require risk. These are conditional comparisons, not claims that every leader uses these policies successfully. ### Sources - Howard Marks, Fewer Losers, or More Winners? (September 12, 2023), Putting It in Brief; Not Risk Avoidance; The Role of Risk Bearing; What About in Practice?. https://www.oaktreecapital.com/insights/memo/fewer-losers-or-more-winners Evidence note: A conceptual and retrospective memo. It distinguishes bond upside, strategies needing winners, and risk control. It does not establish that a reader can identify investment bargains. - Jeffrey P. Bezos, Amazon 2016 shareholder letter (2016 reporting year; published 2017), Resist Proxies and High-Velocity Decision Making.. https://www.aboutamazon.com/news/company-news/2016-letter-to-shareholders Evidence note: Use the signed Bezos letter, not the hosting page’s staff byline. Distinguish his reported practices from proof that they caused Amazon’s results. ## Comparison · Delegate decisions without losing the evidence Delegation concerns a particular decision, not a general personality preference for freedom or control. A local team may know a customer relationship better than headquarters, while headquarters knows aggregate financing commitments that no single team can see. The useful design connects each kind of knowledge to the decisions it can improve. In his 2014 letter, Leonard expresses a preference for smaller business units and describes information sharing and experimentation across them. He also records that some operating leaders were less convinced. That detail matters. A public description of decentralization does not establish unanimous agreement, identical practice in every unit, or proof that centralization can never help. In Amazon's 2016 letter, Bezos likewise distinguishes genuine disagreement about an option from misalignment about the objectives. The latter may require escalation, even in an organization that values speed. For an unfamiliar organization, draw the decision boundary. Which choices can a unit make on its own? What expenditure, service exposure, or contractual change needs another party's agreement? Which information must the unit disclose afterward? A boundary that no one can explain will produce either hidden centralization, because everyone informally asks permission, or hidden exposure, because commitments accumulate without a shared view. Now test incentives. A unit paid on revenue can grow by accepting customers who consume more service than they fund. A center paid on reported margin can cut a shared capability whose deterioration appears only after the measurement period. These are constructed mechanisms, not allegations about the four organizations. The exercise is to make the conflict observable. Track a result relevant to the objective alongside a measure of the capability or obligation that could be sacrificed to improve it. An operating comparison needs more care than a league table. Two units can differ in customer mix, contract age, inherited product condition, geography, and accounting treatment. A higher margin can be evidence of a better process, an easier starting position, or deferred work. Before requiring imitation, identify what changed and what plausibly caused the difference. Invite the receiving unit to state the conditions under which the practice would transfer. A valid local objection should be testable; local autonomy should not immunize an ineffective habit. Set an escalation rule for consequences that cross the boundary. A data incident, a financing commitment, or a shared infrastructure failure may affect many units. Local knowledge remains valuable, but the right to expose other people to losses needs a separate justification. The decision owner should know when their authority ends. Headquarters should know what it must decide, not simply demand to be copied on every exchange. The resulting operating design should be small enough to use. Name the owner, the boundary, the evidence, and the exception. Then run a case in which the local unit and the center sincerely disagree. If the only resolution is another meeting, the design is incomplete. If the resolution is always that one side wins regardless of the evidence, the organization has replaced judgment with hierarchy or habit. ### Work the question A specialist software unit wants to replace part of its customer-support process with an AI service. Design the boundary between its authority and the parent company’s authority. Include a reason to approve a limited test and a reason to escalate even a cheap test. ### Compare with the guide’s reasoning The unit can contribute customer knowledge and run a bounded service experiment. The parent may need to approve shared-data exposure, contractual liability, common infrastructure changes, and commitments beyond the unit’s capacity. Low spend is not proof of low consequence. State what customers can experience, how service is restored, who monitors failure, and which observation ends the test. The scenario is constructed, not a report about Constellation or Amazon. ### Sources - Mark Leonard, Constellation Software 2014 president’s letter (2014 reporting year; published 2015), Pages 1–5, particularly intangible-asset condition, reliable capital, and business-unit decentralization.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: The URL reflects later file hosting, not a 2026 letter. Business figures in the original refer to their stated historical periods. Author explanations remain author explanations. - Jeffrey P. Bezos, Amazon 2016 shareholder letter (2016 reporting year; published 2017), Resist Proxies and High-Velocity Decision Making.. https://www.aboutamazon.com/news/company-news/2016-letter-to-shareholders Evidence note: Use the signed Bezos letter, not the hosting page’s staff byline. Distinguish his reported practices from proof that they caused Amazon’s results. ## Comparison · Change the conditions, then rebuild the argument A modern adaptation should preserve the decision problem while changing the conditions that matter. Replacing the word factory with AI is not enough. The learner needs a reason to reconsider the original choice. That reason might be a new substitute, a different financing contract, a shift in customer bargaining power, a capability that has become easier to copy, or a consequence that cannot be reversed. All 2026 and future business situations in this course are constructed unless a passage explicitly identifies a verified external fact. A fictional scenario can be realistic without being a forecast. Its role is to make an assumption visible and let you test the reasoning under a specified change. The future stages are not claims about what will happen to Amazon, Berkshire, Oaktree, Constellation, or the broader economy. Start the transfer by naming the historical mechanism. If a business benefited from difficult customer switching, identify the actual source of that difficulty. It could be data migration, staff training, operational downtime, contractual integration, or uncertainty about the replacement. Cheaper software development changes none of these automatically. In a constructed 2026 setting, an AI-assisted competitor might reduce one of them. Your task is to determine which one, what remains, and how a customer would experience the difference. Then rebuild the cost structure. A service that reduces employee hours may add vendor fees, review work, remediation, security controls, and transition costs. A system that performs well in a demonstration may encounter a different mix of cases after deployment. These are possibilities to examine, not assertions about a specific product. The right comparison is the whole service under comparable demand and quality requirements. A faster response that fails to solve the problem is not the same output at a lower cost. Next ask who captures the benefit. Customers may obtain a lower price, suppliers may charge more for a scarce input, employees may spend their time on more valuable work, or the owner may retain a larger cash surplus. Several effects can occur together. A technical efficiency does not itself tell you the distribution. If the competitor can implement the same improvement, your temporary benefit may differ from your long-term advantage. If a critical customer refuses the new process, the adoption constraint may outweigh the cost saving. Finally examine the information and the decision clock. A future stage can bring an adverse observation that changes the original estimate. It can also bring a favorable result whose cause is ambiguous. Do not automatically double down on a success or retreat from a failure. Ask whether the observation distinguishes the original competing explanations. Update the relevant assumption, then recompute the feasible actions with the resources and commitments now remaining. The four studies supply questions, not simulated personalities. You are not being asked to impersonate Buffett's opinion of a 2026 security or Leonard's view of an invented software company. The historical sources cannot verify such opinions. You are being asked to use a stated argument, identify its boundary, and take responsibility for a present decision under assumptions you can explain. ### Work the question Take one historical case after completing it. Change exactly one important condition for a fictional 2026 adaptation. Explain why that condition could change the choice, what evidence you would gather, and which part of the original reasoning survives. ### Compare with the guide’s reasoning A strong adaptation changes a mechanism, a constraint, or the evidence available. It does not merely replace company names. If lower production cost is the change, examine whether willingness to pay, switching friction, distribution, or liability changes too. Keep the new condition labeled as fictional. Do not attribute your forecast or decision to the historical author. ### Sources - Jeffrey P. Bezos, Amazon 2016 shareholder letter (2016 reporting year; published 2017), Resist Proxies and High-Velocity Decision Making.. https://www.aboutamazon.com/news/company-news/2016-letter-to-shareholders Evidence note: Use the signed Bezos letter, not the hosting page’s staff byline. Distinguish his reported practices from proof that they caused Amazon’s results. - Mark Leonard, Constellation Software 2014 president’s letter (2014 reporting year; published 2015), Pages 1–5, particularly intangible-asset condition, reliable capital, and business-unit decentralization.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: The URL reflects later file hosting, not a 2026 letter. Business figures in the original refer to their stated historical periods. Author explanations remain author explanations. - Howard Marks, Fewer Losers, or More Winners? (September 12, 2023), Putting It in Brief; Not Risk Avoidance; The Role of Risk Bearing; What About in Practice?. https://www.oaktreecapital.com/insights/memo/fewer-losers-or-more-winners Evidence note: A conceptual and retrospective memo. It distinguishes bond upside, strategies needing winners, and risk control. It does not establish that a reader can identify investment bargains. ## Comparison · Make one allocation you can defend This capstone is an original fictional case. It does not describe any of the four companies. The decision date is September 2026. All amounts are US dollars, and all figures, probabilities if you add them, contract terms, and later developments are assumptions for the exercise. There is no single option whose selection proves mastery. You oversee a small group of profitable specialist service businesses. The group has $6 million in cash at the decision date. Its board requires a $2 million operating reserve. Your stress case also identifies $1 million of additional cash that may be needed during the next twelve months for delayed collections and required service expenditure. Assume those needs do not overlap with the operating reserve. On that basis, $3 million is available for discretionary commitments. Do not silently spend the same dollar twice. An owner offers a specialist software business for $3.2 million paid at closing. You estimate another $0.4 million will be required at closing to support working capital and essential transition work. These are separate uses of cash. The seller reports recurring demand, but you have not yet independently checked the renewal cohorts or the amount of unpaid maintenance work. A competing product promises easier migration. Its effectiveness is unproven in this fictional case. At the same time, one of your existing units proposes a $0.6 million, six-month experiment in AI-assisted service delivery. Another $0.4 million would be committed only after an agreed customer and operating test. Those amounts are total cash outlays for the specified stages, not annual expenses. You can make the first-stage commitment without promising the second. The unit has described an impressive demonstration but has not yet tested the proposed service on a representative mix of customer problems. You can retain cash, pursue either opportunity on different terms, or decline both. Assume no committed financing is presently available. You may investigate financing, but you must specify its terms before treating it as money you can spend. You may negotiate staged acquisition consideration, but the seller has not accepted it. Your first task is to identify which proposals are feasible as stated. The acquisition's immediate cash need is $3.6 million, exceeding the $3 million available under the stated constraints. An attractive return estimate cannot remove that shortfall. Write a recommendation that names what you will do now, what you will investigate, and what you refuse to promise. Identify the customer evidence, sustaining expenditure, liability, and financing information that could reverse it. Explain the difference between buying an existing stream of cash and funding an experiment to discover a new one. Use the individual studies to sharpen your questions, but do not substitute a famous person's name for a reason. Now introduce a hypothetical update in March 2027. The experiment has cost its full first-stage $0.6 million. It reduced time spent on simple requests, but difficult cases require more human review than the demonstration suggested. The software seller offers a lower closing payment in exchange for a larger contingent payment later. A major customer of your existing business delays payment. The exercise intentionally does not specify the delayed amount, revised acquisition terms, or experiment economics. Identify the missing facts and a temporary action that protects the group's obligations while those facts are obtained. Inventing favorable values to finish the recommendation is a failure of evidence discipline. Assess the original recommendation and the revision separately. Give each of the following dimensions a self-assessed 0, 1, or 2: source reconstruction, alternatives, economic arithmetic, uncertainty, downside and obligations, operating feasibility, evidence quality, and revision. Zero means missing or materially mistaken, one means partially developed, and two means explicit and defensible. The possible total is 16 by construction. It is a guide rubric, not a validated test or an expertise percentile. A persuasive assessment names the passage in your own answer that earns the rating and the correction needed where it does not. Return to this case after working through an unfamiliar real decision. Replace the fictional inputs with evidence you can disclose, preserve uncertainty, and ask another person to challenge your causal chain and financing assumptions. The ability to explain a familiar case is useful. The ability to transfer the reasoning, withstand criticism, and revise responsibly is the more demanding test. ### Work the question Write the September 2026 recommendation, then a separate March 2027 revision. Show the available-cash calculation, identify the infeasible proposal as stated, and explain what further evidence is required. Score each version against the eight-dimension rubric and justify every score. ### Compare with the guide’s reasoning Available discretionary cash is $6m − $2m − $1m = $3m. Acquisition cash needed at closing is $3.2m + $0.4m = $3.6m. The immediate shortfall is $0.6m before any experiment. Retaining cash, investigating a bounded experiment, or negotiating different acquisition terms can each be defensible with the right reasons. A simultaneous commitment to the acquisition as stated and the experiment breaches the stated constraints. In the update, reserve cash for obligations, quantify the delayed collection and complete service costs, examine the contingent payment, and revise without pretending the missing amounts are known. ### Sources - Warren E. Buffett, Berkshire Hathaway 2012 shareholder letter (2012 reporting year; published 2013), Dividends, including reinvestment, acquisitions, repurchases, and the conditions behind the payout argument.. https://www.berkshirehathaway.com/letters/2012ltr.pdf Evidence note: Written by the allocator. His assessments of acquisition value and alternative outcomes are his judgments, not independent causal measurements. - Howard Marks, Fewer Losers, or More Winners? (September 12, 2023), Putting It in Brief; Not Risk Avoidance; The Role of Risk Bearing; What About in Practice?. https://www.oaktreecapital.com/insights/memo/fewer-losers-or-more-winners Evidence note: A conceptual and retrospective memo. It distinguishes bond upside, strategies needing winners, and risk control. It does not establish that a reader can identify investment bargains. - Mark Leonard, Constellation Software 2014 president’s letter (2014 reporting year; published 2015), Pages 1–5, particularly intangible-asset condition, reliable capital, and business-unit decentralization.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: The URL reflects later file hosting, not a 2026 letter. Business figures in the original refer to their stated historical periods. Author explanations remain author explanations. - Jeffrey P. Bezos, Amazon 2016 shareholder letter (2016 reporting year; published 2017), Resist Proxies and High-Velocity Decision Making.. https://www.aboutamazon.com/news/company-news/2016-letter-to-shareholders Evidence note: Use the signed Bezos letter, not the hosting page’s staff byline. Distinguish his reported practices from proof that they caused Amazon’s results. ## Advanced practice · Build a record of changing judgment Finishing a chronological path should produce something more useful than agreement with its subject. Build a short development dossier that makes a change in judgment inspectable. The dossier is your work, and the method below is the course’s teaching design. It is not a quotation or a method claimed to have been followed by all four people. Begin with a dated belief. State it narrowly enough that evidence could challenge it. Avoid labels such as long-term thinker or disciplined investor. A useful statement names a choice and its conditions: retain capital at an established hurdle, keep supporting an operating business, subsidize a service while testing customer behavior, or alter exposure when prices change. Identify whether the source was written at the time or is a later recollection. Next reconstruct the pressure. A belief does not change merely because a later writer has found a better slogan. Capital grows, competitors react, employees learn what a bonus rewards, customers reject a product, funding disappears, or a partner sees something the central person missed. Name the mechanism and the people who experienced it. An outcome can be financially tolerable for an owner and painful for employees. Both facts can belong in the same account. Then separate the update from the result. Did the decision-maker change a forecast, the objective, an operating process, or the range of activities considered acceptable? Those are different changes. Leonard’s 2021 letter, for example, changed the policy for particular deployment opportunities while retaining other hurdles. Calling it simply more aggressive erases the boundary. Buffett’s retrospective Berkshire discussion likewise needs to be separated from what could be known when the original decisions were made. Create a contrary-evidence paragraph. Explain the best reason an informed person might reject your preferred interpretation. Perhaps the apparent improvement came from the cycle. Perhaps a successful experiment had an unusually favorable environment. Perhaps a result described as patience was the cost of delaying an exit. The aim is not to invent an objection for balance. It is to find the observation that could make your explanation wrong. Finally transfer the mechanism to a different setting. Change a meaningful condition such as financing duration, customer switching costs, decision reversibility or the ability to measure results. Explain which part of the historical lesson survives and which part no longer applies. The constructed cases in this course provide settings in which to practice that transfer without claiming to predict what a Titan would do today. Keep your first answer. After reading the later period, write a separate revision that identifies the new evidence and exactly what it changed. A polished final explanation is less informative than a visible record of the earlier assumption. Your progress is the quality of that revision: better distinctions, better evidence and a clearer account of uncertainty. Word count, agreement with the famous decision and confidence of tone are not substitutes. ### Work the question Choose one person and produce a dossier with a dated belief, pressure, response, outcome, contrary evidence and transfer. End with the strongest remaining uncertainty. Keep the earlier answer before writing the revision. ### Compare with the guide’s reasoning A strong dossier identifies a real update rather than assigning an unchanging personality. It separates contemporary and retrospective evidence, includes costs borne by someone other than the owner, and changes the transfer conclusion when an important condition changes. Each factual statement should point to a named source and period. If the evidence does not show a private motive or completed outcome, say so. ### Sources - Warren E. Buffett, 2014 Chairman's Letter (2014 reporting period; published 2015), Berkshire, Past, Present and Future, especially the textile, See's and Dexter passages. https://www.berkshirehathaway.com/letters/2014ltr.pdf Evidence note: Fifty-year retrospective. It supplies later cumulative figures and Buffett's revised interpretation, not a contemporaneous record of the earlier decisions. The Dexter share value is measured at the letter's February 2015 publication. - Jeffrey P. Bezos, 2015 Letter to Shareholders (2015 reporting period; published 2016), Failure, long-tailed business returns, and Type 1 versus Type 2 decisions. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2015-Letter-to-Shareholders.PDF Evidence note: Retrospective management account. The ten-percent and one-hundred-times illustration explains an asymmetric payoff shape, not a measured base rate for Amazon experiments. - Howard Marks, Fewer Losers, or More Winners? (September 12, 2023), Putting It in Brief; Not Risk Avoidance; The Role of Risk Bearing; What About in Practice?. https://www.oaktreecapital.com/insights/memo/fewer-losers-or-more-winners Evidence note: A conceptual and retrospective memo. It distinguishes bond upside, strategies needing winners, and risk control. It does not establish that a reader can identify investment bargains. - Mark Leonard, Constellation Software Inc. 2017 President's Letter (2018-04-20), Performance metrics; retained-capital policy; perpetual-owner objective; business-unit count. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-april-2018-final.pdf Evidence note: Management report. Statements about scalability and governance are Leonard's judgments, not independent causal findings. - Mark Leonard, Constellation Software Inc. 2021 President's Letter (2021-02-15), Complete letter, especially dividend policy, large VMS acquisitions, and head-office role. https://www.csisoftware.com/wp-content/uploads/2026/04/letter-to-shareholders-february-15-2021.pdf Evidence note: Records a board policy decision and management intentions as of publication. The reported 16 percent invitation figure does not have an unambiguous denominator in the letter. It does not establish the results of later investments. ## Advanced practice · The examination · Defend, challenge, revise Use this examination after working through the development paths and cases. Its purpose is to reveal what you can reconstruct and apply without leaning on the author’s reputation. The standard is demanding, but it is a course rubric, not a validated professional credential or a measured percentile of expertise. First, write a chronology from memory. Select a consequential development for each person and give the earlier position, the pressure that tested it, the later response and the uncertainty that remained. After committing the answer, compare it with the course’s named source records. Mark errors in dates, authorship and reporting periods explicitly. Do not silently repair them before judging your recall. A correct general theme does not excuse attributing another executive’s statement to the featured person. Second, defend a decision you initially disliked. Use only information that could have been available at the decision date. A defense is not an endorsement. It shows that you understand the constraint, objective and alternative sufficiently well to avoid explaining the past as if the ending were obvious. Then give the best case against the same decision. If both arguments use identical assumptions, examine whether you have actually found a disagreement. Third, compare two people on one specific problem. Use capital allocation, organizational authority, experimentation or risk as the starting point, then narrow the question. Their preferred answers may differ because of business economics, financing, incentives or information. Do not resolve the comparison by blending their names into a universal philosophy. Explain the condition under which one approach would become more appropriate than the other. Fourth, solve a constructed case without opening its debrief. Record a choice, a causal explanation, the largest uncertainty, the people who bear the downside, and a trigger for review. When the case changes, revise the decision only if the new information warrants it. Sometimes the correct update is to retain the decision with lower confidence. Sometimes it is to change the policy while preserving the objective. A dramatic reversal is not inherently evidence of learning. Fifth, audit your own argument. Find one place where you treated an observed association as a cause, one place where an aggregate hid a distribution, and one place where a metric depended on an untested assumption. If you find none, ask someone else or your chosen AI to attack the reasoning using the source packet. Evaluate the criticism against the evidence. Neither an AI’s confident grade nor your own confidence should decide the result. Assess each answer on five dimensions: source accuracy, causal explanation, treatment of alternatives, uncertainty, and adaptation to changed conditions. An answer is not yet ready if it misattributes a source, imports a later outcome into an earlier decision, or treats fictional case numbers as historical facts. A strong answer can survive a counterargument and state where its evidence stops. Return to the relevant reading when it cannot. The intended endpoint is reliable judgment you can use and explain, not a collection of favorable opinions about exceptional people. ### Work the question Complete the five tasks before reading this answer. Choose your weakest response, explain why it is weak, and rewrite it using a specific source or a changed assumption. ### Compare with the guide’s reasoning Judge the work against explicit evidence and reasoning. Source errors or hindsight require correction even if the recommendation sounds sensible. The strongest comparison states why the approaches differ and what condition would change the choice. The strongest revision identifies the information responsible for the update and preserves uncertainty that has not been resolved. ### Sources - Warren E. Buffett, 2014 Chairman's Letter (2014 reporting period; published 2015), Berkshire, Past, Present and Future, especially the textile, See's and Dexter passages. https://www.berkshirehathaway.com/letters/2014ltr.pdf Evidence note: Fifty-year retrospective. It supplies later cumulative figures and Buffett's revised interpretation, not a contemporaneous record of the earlier decisions. The Dexter share value is measured at the letter's February 2015 publication. - Jeffrey P. Bezos, 2015 Letter to Shareholders (2015 reporting period; published 2016), Failure, long-tailed business returns, and Type 1 versus Type 2 decisions. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2015-Letter-to-Shareholders.PDF Evidence note: Retrospective management account. The ten-percent and one-hundred-times illustration explains an asymmetric payoff shape, not a measured base rate for Amazon experiments. - Howard Marks, Fewer Losers, or More Winners? (September 12, 2023), Putting It in Brief; Not Risk Avoidance; The Role of Risk Bearing; What About in Practice?. https://www.oaktreecapital.com/insights/memo/fewer-losers-or-more-winners Evidence note: A conceptual and retrospective memo. It distinguishes bond upside, strategies needing winners, and risk control. It does not establish that a reader can identify investment bargains. - Mark Leonard, Constellation Software Inc. 2017 President's Letter (2018-04-20), Performance metrics; retained-capital policy; perpetual-owner objective; business-unit count. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-april-2018-final.pdf Evidence note: Management report. Statements about scalability and governance are Leonard's judgments, not independent causal findings. - Mark Leonard, Constellation Software Inc. 2021 President's Letter (2021-02-15), Complete letter, especially dividend policy, large VMS acquisitions, and head-office role. https://www.csisoftware.com/wp-content/uploads/2026/04/letter-to-shareholders-february-15-2021.pdf Evidence note: Records a board policy decision and management intentions as of publication. The reported 16 percent invitation figure does not have an unambiguous denominator in the letter. It does not establish the results of later investments. ## Three-stage cases ### Warren Buffett · When support becomes an unlimited claim on capital Skill: Distinguish a repairable operation from a structurally poor reinvestment #### The New England mills · Historical reconstruction Period: Mid-1985 Berkshire's textile operation employs people in communities that depend on the mills. Management and labour have cooperated and repeatedly changed products, equipment and distribution. The operation was moderately profitable in 1979 but has since consumed major amounts of cash. New equipment can reduce unit cost, yet domestic and foreign competitors can buy similar equipment and foreign producers retain a large labour-cost advantage. No buyer willing to continue operations has emerged. The historical letter does not provide a board transcript or a complete forecast available in July 1985. Decision prompt: Choose a course and write the evidence that makes it preferable. State what obligation to employees you would fund and what future fact would reverse your decision. Options: - Close the textile operation and fund an orderly wind-down. - Continue limited support for a defined period while seeking a buyer or differentiated niche. - Make the proposed cost-saving capital investment and remain in commodity textiles. - Close the weakest capacity and reinvest only in products with evidence of durable differentiation. Debrief: Berkshire decided in July 1985 to close and had largely completed the work by year-end. Buffett wrote that he would have preferred a continuing buyer even for lower proceeds, but interest was nil. Disposals in late 1985 and an auction in early 1986 showed how little the assets were worth in that setting. Equipment with about $13 million of original cost and $866,000 of book value produced gross proceeds of $163,122, and net proceeds after costs were below zero. Buffett also judged that he had quit too late. The outcome supports testing industry-wide responses and realizable asset value. It does not remove the human cost or prove that every cyclical manufacturer should close. Sources: - Warren E. Buffett, 1985 Chairman's Letter (March 4, 1986), Shutdown of Textile Business. https://www.berkshirehathaway.com/letters/1985.html Evidence note: Contemporary account written after the July 1985 closure decision. It reports Buffett's own diagnosis and hindsight about earlier delay. - Warren E. Buffett, 2014 Chairman's Letter (2014 reporting period; published 2015), Berkshire, Past, Present and Future, especially the textile, See's and Dexter passages. https://www.berkshirehathaway.com/letters/2014ltr.pdf Evidence note: Fifty-year retrospective. It supplies later cumulative figures and Buffett's revised interpretation, not a contemporaneous record of the earlier decisions. The Dexter share value is measured at the letter's February 2015 publication. #### The specialty fabric subsidiary · Constructed 2026 scenario Period: December 1, 2026 All dollar figures are fictional United States dollar assumptions. A group owns a specialty fabric subsidiary with 140 employees. For the twelve months ending September 30, 2026, revenue is $48 million and EBITDA is $1.6 million. Maintenance capital spending is estimated at $3.0 million a year and working capital rises $0.4 million at the current sales level. Before group tax and financing, owner cash after maintenance and working capital is therefore negative $1.8 million. A $12 million automated line is forecast to save $3.0 million a year, but competitors have announced comparable upgrades. The planning team expects industry price reductions to pass $2.4 million of the annual saving to buyers. A shutdown costs $6 million. An asset buyer offers $1 million but will not continue operations. Forecasts cover the next twelve months and remain uncertain. Decision prompt: Decide on the 2026 capital request. Show the arithmetic, address employees and competitive response, and specify a review or exit trigger. Options: - Reject the line and begin the $6 million wind-down. - Invest $12 million in the automated line and compete on cost. - Fund a one-year bridge only for named differentiated products and seek contracts before capital spending. - Accept the $1 million asset offer and separately fund employee transition. Debrief: The baseline owner-cash bridge is $1.6 million EBITDA minus $3.0 million maintenance capital minus $0.4 million working capital, or negative $1.8 million for twelve months before group tax and financing. If the upgrade works exactly as forecast and $2.4 million of the $3.0 million saving becomes lower market price, only $0.6 million remains before considering the $12 million outlay and further maintenance. That is a 5 percent simple annual benefit on the initial outlay before risk, tax and time value. A close decision should compare the $6 million wind-down with expected continuing losses and employee commitments. A bridge can be defensible if it buys specific evidence, such as signed premium-price contracts. General hope that a niche will appear is not evidence. Sources: - Warren E. Buffett, 1985 Chairman's Letter (March 4, 1986), Shutdown of Textile Business. https://www.berkshirehathaway.com/letters/1985.html Evidence note: Contemporary account written after the July 1985 closure decision. It reports Buffett's own diagnosis and hindsight about earlier delay. - Warren E. Buffett, 2014 Chairman's Letter (2014 reporting period; published 2015), Berkshire, Past, Present and Future, especially the textile, See's and Dexter passages. https://www.berkshirehathaway.com/letters/2014ltr.pdf Evidence note: Fifty-year retrospective. It supplies later cumulative figures and Buffett's revised interpretation, not a contemporaneous record of the earlier decisions. The Dexter share value is measured at the letter's February 2015 publication. #### The recycled-fibre mandate · Hypothetical future stress test Period: 2031 planning cycle This is a fictional future stress test, and all dollar figures are United States dollar assumptions. Assume the 2026 board funded a one-year bridge and later kept one small production line. In 2031, a regional rule requires specified public uniforms to contain certified recycled fibre. A four-year supply tender offers annual revenue of $18 million. Expected EBITDA is $4.5 million a year before $1 million of annual maintenance capital. Winning requires a $9 million dedicated line and $3 million of additional working capital at signing. The full $3 million of working capital is assumed to be recovered at the end of year four. The line has an estimated $2 million resale value after four years. Certification can be copied, and two rivals are bidding. Contract penalties apply if recycled input supply fails. The figures exclude tax and financing. Decision prompt: Revise your 2026 reasoning for the 2031 tender. Decide whether the contract changes the business structure or only creates a temporary protected pocket. Calculate the undiscounted cash before tax and state the evidence you need on renewal and supply risk. Options: - Bid and invest $12 million, treating the contract as a bounded project. - Bid only with a partner that shares capital and recycled-input risk. - License the certification process to a stronger producer rather than manufacture. - Decline because the mandate does not create a durable advantage. Debrief: On the assumptions, four years of EBITDA total $18 million. Subtract $4 million of maintenance capital, the $9 million line and $3 million of working capital, then add the assumed $3 million working-capital recovery and $2 million resale value at the end. Undiscounted pre-tax project cash is $7 million before penalties, financing and time value. That can justify a bounded bid even if the broader textile business remains unattractive. The case turns on contract enforceability, input supply, copying, penalty exposure and what happens after year four. A new protected demand pocket should cause revision, but it does not erase the earlier commodity diagnosis. Sources: - Warren E. Buffett, 1985 Chairman's Letter (March 4, 1986), Shutdown of Textile Business. https://www.berkshirehathaway.com/letters/1985.html Evidence note: Contemporary account written after the July 1985 closure decision. It reports Buffett's own diagnosis and hindsight about earlier delay. - Warren E. Buffett, 2014 Chairman's Letter (2014 reporting period; published 2015), Berkshire, Past, Present and Future, especially the textile, See's and Dexter passages. https://www.berkshirehathaway.com/letters/2014ltr.pdf Evidence note: Fifty-year retrospective. It supplies later cumulative figures and Buffett's revised interpretation, not a contemporaneous record of the earlier decisions. The Dexter share value is measured at the letter's February 2015 publication. #### Reasoning rubric - Separates employee and community obligations from an unlimited capital commitment. - Uses cash after maintenance and working capital rather than EBITDA alone. - Models likely competitor response and who captures cost savings. - Treats closure, sale, bridge and reinvestment as explicit counterfactuals. - Revises when a bounded contract changes the facts without turning the historical case into a universal rule. ### Warren Buffett · The acquisition price hidden inside the shares Skill: Value both the target and the ownership currency surrendered #### Dexter Shoe · Historical reconstruction Period: Before the November 7, 1993 merger Dexter produces more than 7.5 million pairs of shoes a year, mainly in Maine, operates 77 retail outlets, and makes about 15 percent of United States golf-shoe output. Two major retailers recognized it as a supplier for 1992. Berkshire's earlier H. H. Brown and Lowell shoe acquisitions have exceeded expectations. The domestic industry is widely thought vulnerable to imports from low-wage countries, but Dexter has competed successfully so far. The sellers want Berkshire shares rather than cash. The proposed consideration is 25,203 Berkshire Class A shares, later described as $433 million at the transaction value. The public record does not disclose a full pre-deal valuation model or private deliberations. Decision prompt: Decide whether to issue the shares. Explain how you would value Berkshire's ownership currency, stress-test Dexter's import advantage, and distinguish seller benefits from benefits to continuing Berkshire owners. Options: - Acquire Dexter for the proposed 25,203 Class A shares. - Decline because the durability and stock consideration do not provide a margin of safety. - Offer fewer shares at closing plus contingent consideration tied to durable cash results. - Defer and seek evidence on landed import cost, customer switching and stress-case cash flow. Debrief: Berkshire completed the share merger on November 7, 1993. The contemporary letter praised Dexter's economics and management. Later letters reported that its competitive strength disappeared under foreign competition. Buffett's 2001 review separated the purchase, stock payment and delayed response as three mistakes. The 2007 letter, published in February 2008, valued the original consideration at $433 million and the shares at about $3.5 billion at writing. The 2014 letter, published in February 2015, said the shares were worth about $5.7 billion at that later writing date while Dexter's value had gone to zero. Those later figures reveal opportunity cost, but hindsight does not identify the exact probability that should have been assigned in 1993. Sources: - Warren E. Buffett, 1993 Chairman's Letter (1993 reporting period; published 1994), Dexter Shoe. https://www.berkshirehathaway.com/letters/1993.html Evidence note: Contemporary, favorable account after the November 7, 1993 merger. It is useful evidence of the case made then, not proof that all deliberations were disclosed. - Warren E. Buffett, 2001 Chairman's Letter (2001 reporting period; published 2002), Non-insurance operating highlights, shoe operations and Dexter. https://www.berkshirehathaway.com/2001ar/2001letter.html Evidence note: Retrospective self-assessment identifying the purchase, use of stock, and delayed operational response as three separate errors. - Warren E. Buffett, 2007 Chairman's Letter (2007 reporting period; published 2008), Businesses, The Great, the Good and the Gruesome. https://www.berkshirehathaway.com/letters/2007ltr.pdf Evidence note: Retrospective comparison of See's, FlightSafety, capital-intensive growth, and Dexter. The Dexter share value is measured at the letter's February 2008 publication, not in the 2007 reporting year or in 1993. - Warren E. Buffett, 2014 Chairman's Letter (2014 reporting period; published 2015), Berkshire, Past, Present and Future, especially the textile, See's and Dexter passages. https://www.berkshirehathaway.com/letters/2014ltr.pdf Evidence note: Fifty-year retrospective. It supplies later cumulative figures and Buffett's revised interpretation, not a contemporaneous record of the earlier decisions. The Dexter share value is measured at the letter's February 2015 publication. #### The industrial sensor acquisition · Constructed 2026 scenario Period: October 15, 2026 All dollar figures are fictional United States dollar assumptions as of October 15, 2026. Independent work values a public industrial-software acquirer's pre-transaction stand-alone equity at $780 to $920 per share. The seller asks for 1.2 million shares. The ownership currency therefore has an estimated intrinsic equity value of $936 million to $1.104 billion. The target's stand-alone value for 100 percent of its equity after net debt is estimated at $720 million to $900 million. The case assumes zero net debt and no other claims. A possible distribution synergy has an incremental after-tax present value of $180 million on the same date if it works, with a 40 percent probability assigned by the deal team and zero value otherwise. Risk-adjusted expected synergy is $72 million. The target's largest product supplies 55 percent of revenue and faces a new open standard in 2028. The seller rejects all-cash terms. Decision prompt: Choose a structure and show the low and high value exchanged. Explain how the 2028 standard affects the margin of safety and which claim needs independent technical evidence. Options: - Issue 1.2 million shares at closing. - Offer 800,000 shares plus up to 400,000 contingent on cash and retention after the standard arrives. - Buy a minority position with commercial rights and postpone control. - Decline because the consideration exceeds supported value in too much of the range. Debrief: On the common October 15, 2026 equity-value basis, the full-stock offer surrenders an estimated $936 million to $1.104 billion of intrinsic value. Adding the risk-adjusted $72 million after-tax synergy to the target equity range produces $792 million to $972 million. The ranges overlap, so the answer is not automatic. At the acquirer's high value and target's low value, the gap is $312 million before considering the open-standard downside. An earnout can place some risk on the seller, though it can also distort behavior and create disputes. The 40 percent probability is a deal-team judgment and should not become independent evidence merely because it appears in a model. Sources: - Warren E. Buffett, 1993 Chairman's Letter (1993 reporting period; published 1994), Dexter Shoe. https://www.berkshirehathaway.com/letters/1993.html Evidence note: Contemporary, favorable account after the November 7, 1993 merger. It is useful evidence of the case made then, not proof that all deliberations were disclosed. - Warren E. Buffett, 2001 Chairman's Letter (2001 reporting period; published 2002), Non-insurance operating highlights, shoe operations and Dexter. https://www.berkshirehathaway.com/2001ar/2001letter.html Evidence note: Retrospective self-assessment identifying the purchase, use of stock, and delayed operational response as three separate errors. - Warren E. Buffett, 2007 Chairman's Letter (2007 reporting period; published 2008), Businesses, The Great, the Good and the Gruesome. https://www.berkshirehathaway.com/letters/2007ltr.pdf Evidence note: Retrospective comparison of See's, FlightSafety, capital-intensive growth, and Dexter. The Dexter share value is measured at the letter's February 2008 publication, not in the 2007 reporting year or in 1993. - Warren E. Buffett, 2014 Chairman's Letter (2014 reporting period; published 2015), Berkshire, Past, Present and Future, especially the textile, See's and Dexter passages. https://www.berkshirehathaway.com/letters/2014ltr.pdf Evidence note: Fifty-year retrospective. It supplies later cumulative figures and Buffett's revised interpretation, not a contemporaneous record of the earlier decisions. The Dexter share value is measured at the letter's February 2015 publication. #### The rescue capital decision · Hypothetical future stress test Period: 2032 capital review This is a fictional future stress test, and all dollar figures are United States dollar assumptions. Assume the acquisition closed with contingent stock. The open standard became dominant and the sensor division now produces $55 million of annual pre-tax owner cash before any new project. All present values use an after-tax, unlevered basis as of December 31, 2032. Independent work gives the division a $180 million present value if it operates without redesign. That value already incorporates the expected decline of the current $55 million annual pre-tax cash. A redesign requires $260 million immediately and produces incremental gross present value, excluding the current cash and redesign outlay, of $420 million in a success case or $70 million in a failure case. Management assigns probabilities of 55 percent and 45 percent, so expected incremental gross value is $262.5 million before the $260 million outlay. A buyer offers a comparable $310 million value net of assumed tax and separation costs. Closing would cost $90 million. The available issue price is $1,600 of net proceeds per share, assumed equal to estimated intrinsic value, so raising $260 million would require 162,500 shares. Decision prompt: Make the 2032 decision without trying to recover the 2026 purchase price. Compare the $182.5 million expected keep value after redesign with the $310 million sale and the negative closure value. State what evidence could make the probability-weighted redesign estimate misleading. Options: - Fund the redesign from cash and keep the division. - Issue 162,500 shares to fund the redesign. - Sell the division for $310 million. - Close and pay the $90 million wind-down cost. Debrief: Expected incremental gross redesign value is 0.55 times $420 million plus 0.45 times $70 million, or $262.5 million. Subtract the $260 million redesign outlay to get a $2.5 million expected project net present value. Add that to the $180 million no-redesign division value for an expected keep value of $182.5 million. The current $55 million annual pre-tax cash is already inside the no-redesign value and must not be added again. The comparable sale offer is $310 million net of the scenario's assumed tax and separation costs, while closure has a negative $90 million value before any asset recovery. At the assumed issue price, stock funding transfers 162,500 shares for $260 million of net proceeds. The original acquisition cost is sunk. A defensible decision must challenge the coarse outcome cases, probabilities, buyer certainty and strategic dependencies while preserving the common valuation date and basis. Sources: - Warren E. Buffett, 1993 Chairman's Letter (1993 reporting period; published 1994), Dexter Shoe. https://www.berkshirehathaway.com/letters/1993.html Evidence note: Contemporary, favorable account after the November 7, 1993 merger. It is useful evidence of the case made then, not proof that all deliberations were disclosed. - Warren E. Buffett, 2001 Chairman's Letter (2001 reporting period; published 2002), Non-insurance operating highlights, shoe operations and Dexter. https://www.berkshirehathaway.com/2001ar/2001letter.html Evidence note: Retrospective self-assessment identifying the purchase, use of stock, and delayed operational response as three separate errors. - Warren E. Buffett, 2007 Chairman's Letter (2007 reporting period; published 2008), Businesses, The Great, the Good and the Gruesome. https://www.berkshirehathaway.com/letters/2007ltr.pdf Evidence note: Retrospective comparison of See's, FlightSafety, capital-intensive growth, and Dexter. The Dexter share value is measured at the letter's February 2008 publication, not in the 2007 reporting year or in 1993. - Warren E. Buffett, 2014 Chairman's Letter (2014 reporting period; published 2015), Berkshire, Past, Present and Future, especially the textile, See's and Dexter passages. https://www.berkshirehathaway.com/letters/2014ltr.pdf Evidence note: Fifty-year retrospective. It supplies later cumulative figures and Buffett's revised interpretation, not a contemporaneous record of the earlier decisions. The Dexter share value is measured at the letter's February 2015 publication. #### Reasoning rubric - Values issued shares using an intrinsic-value range and ownership percentage, not only the market quote. - Separates target quality, transaction price and financing risk. - Tests competitive durability with an adverse case available at the decision date. - Distinguishes later opportunity-cost evidence from facts known in 1993. - Treats the later rescue as a fresh allocation decision and ignores sunk acquisition cost. ### Jeff Bezos · Reinvest while the customer evidence is moving Skill: Translate long-term intent into measures, limits and revision rules #### Amazon after its first public year · Historical reconstruction Period: End of 1997 planning for 1998 Amazon's 1997 sales are $147.8 million, up from $15.7 million. Cumulative customer accounts reach 1.51 million from 180,000, and repeat-customer orders exceed 58 percent in the fourth quarter, up from more than 46 percent. Distribution space expands from 50,000 to 285,000 square feet. Inventory exceeds 200,000 titles. Cash and investments total $125 million after the initial public offering and a $75 million loan. Large, well-funded competitors are entering. Music, overseas service and more infrastructure are possible investments. The company is still incurring net losses, and the public letter does not supply a program-level forecast for each option. Decision prompt: Set the 1998 allocation policy. Choose an approach, name the customer and cash measures, and specify what result would make you slow or stop investment. Options: - Prioritize customer, brand and infrastructure growth despite weaker near-term profit. - Slow expansion and target near-term accounting profitability. - Reinvest aggressively only in programs that pass explicit customer, unit-economic and liquidity gates. - Enter several categories and countries quickly to maximize market coverage. Debrief: Amazon declared a long-term policy that prioritized market leadership, customers and infrastructure while also promising analytical program review, cost consciousness and capital management. It planned music, international improvements and continued systems investment. In 1998, sales reached $610 million, customer accounts 6.2 million and fourth-quarter repeat orders more than 64 percent. Amazon reported $31 million of operating cash flow and $28 million of net fixed-asset additions, while inventory rose to $30 million. These outcomes are consistent with the policy during that year, but they do not isolate the return on each investment or prove that the same policy fits a company with weaker retention or funding. Sources: - Jeffrey P. Bezos, 1997 Letter to Shareholders (1997 reporting period; published 1998), It's All About the Long Term, Obsess Over Customers, Infrastructure, and Goals for 1998. https://www.aboutamazon.com/news/company-news/amazons-original-1997-letter-to-shareholders Evidence note: Contemporary statement of management policy and 1997 results. Management's claims about market leadership and future economics are its judgments at the time. - Jeffrey P. Bezos, 1998 Letter to Shareholders (1998 reporting period; published 1999), A Recap of 1998, Our Customers, and Goals for 1999. https://s2.q4cdn.com/299287126/files/doc_financials/annual/Shareholderletter98.pdf Evidence note: Contemporary follow-up reporting customer, revenue, repeat-order, cash-flow, inventory and infrastructure figures. The outcome does not isolate the effect of the 1997 policy. #### The specialist marketplace · Constructed 2026 scenario Period: September 30, 2026 All dollar figures are fictional United States dollar assumptions. An online specialist marketplace has $120 million of trailing-twelve-month revenue, 35 percent growth, 420,000 active customers and a 54 percent twelve-month repeat-purchase rate. Baseline free cash flow after current capital spending is $6 million a year. Cash is $30 million and debt is zero. Management proposes $18 million of additional distribution and personalization spending over two years, paid evenly. It expects two-day availability to rise from 62 percent to 82 percent of listed demand and repeat purchase to reach 60 percent, but has no controlled evidence for the combined program. If baseline free cash flow stays $6 million and added spending is $9 million a year, the simple pre-working-capital effect is negative $3 million a year for two years. Decision prompt: Choose the 2026 investment policy. Define a staged test, minimum liquidity and a causal chain from availability to repeat use and free cash flow. Options: - Approve the full $18 million two-year program now. - Approve a staged rollout with geography-level evidence and release gates. - Fund distribution only because availability is the more direct constraint. - Preserve cash and require stronger unit economics before expansion. Debrief: The stated simple cash effect is $6 million baseline free cash flow minus $9 million of annual added spending, or negative $3 million in each of two years before additional working capital and any program benefit. The $30 million cash balance is not a two-year guarantee because demand growth may consume inventory and receivables. A staged rollout can compare eligible regions or categories, but network spillovers and customer mix complicate inference. Strong reasoning names availability, conversion, repeat behavior, contribution and cash in sequence, then sets a minimum cash balance and a stop condition if the early links move without the later ones. Sources: - Jeffrey P. Bezos, 1997 Letter to Shareholders (1997 reporting period; published 1998), It's All About the Long Term, Obsess Over Customers, Infrastructure, and Goals for 1998. https://www.aboutamazon.com/news/company-news/amazons-original-1997-letter-to-shareholders Evidence note: Contemporary statement of management policy and 1997 results. Management's claims about market leadership and future economics are its judgments at the time. - Jeffrey P. Bezos, 1998 Letter to Shareholders (1998 reporting period; published 1999), A Recap of 1998, Our Customers, and Goals for 1999. https://s2.q4cdn.com/299287126/files/doc_financials/annual/Shareholderletter98.pdf Evidence note: Contemporary follow-up reporting customer, revenue, repeat-order, cash-flow, inventory and infrastructure figures. The outcome does not isolate the effect of the 1997 policy. #### The buying-agent gateway · Hypothetical future stress test Period: 2030 strategy review This is a fictional future stress test, and all dollar figures are United States dollar assumptions. Assume the marketplace completed a staged build and reached $210 million of annual revenue. In 2030, software buying agents initiate 45 percent of category searches. Orders arriving through agents convert at 18 percent versus 9 percent for direct visits, but the agent charges 7 percent of order value and does not share customer identity. Direct customers have a 63 percent twelve-month repeat rate. Agent-sourced buyers can be recognized only if they opt into the marketplace account after purchase, and 14 percent currently do. The marketplace can spend $12 million over one year on an open product-data service and post-purchase membership offer, or accept the channel as a lower-data source of profitable volume. Decision prompt: Revise the earlier customer-relationship policy for 2030. Decide how much to invest in direct recognition, show which customer metric still matters, and state what evidence would justify paying the agent fee. Options: - Invest $12 million in open product data and a voluntary direct membership path. - Optimize for agent conversion and accept lower customer identity. - Restrict agent access to protect the direct relationship. - Use agents for acquisition while setting contribution and opt-in gates by category. Debrief: The agent channel offers higher stated conversion but charges 7 percent of order value and weakens direct recognition. A customer count that merges identified repeat buyers with anonymous agent orders would hide the change. A mixed approach can be defensible if contribution after the fee is positive and voluntary account opt-in, repeat purchase and service quality improve enough to support the $12 million investment. Restriction can also be rational in categories where advice, trust or repeat data drive economics. The future case changes the route to the customer. It requires revising the 1997 measures rather than repeating customer growth as a slogan. Sources: - Jeffrey P. Bezos, 1997 Letter to Shareholders (1997 reporting period; published 1998), It's All About the Long Term, Obsess Over Customers, Infrastructure, and Goals for 1998. https://www.aboutamazon.com/news/company-news/amazons-original-1997-letter-to-shareholders Evidence note: Contemporary statement of management policy and 1997 results. Management's claims about market leadership and future economics are its judgments at the time. - Jeffrey P. Bezos, 1998 Letter to Shareholders (1998 reporting period; published 1999), A Recap of 1998, Our Customers, and Goals for 1999. https://s2.q4cdn.com/299287126/files/doc_financials/annual/Shareholderletter98.pdf Evidence note: Contemporary follow-up reporting customer, revenue, repeat-order, cash-flow, inventory and infrastructure figures. The outcome does not isolate the effect of the 1997 policy. #### Reasoning rubric - Uses contemporary customer, infrastructure and liquidity evidence without hindsight. - Pairs long-term investment with program measures, stop conditions and a cash boundary. - Separates company-level growth from the return on an individual program. - Traces customer improvement through repeat behavior and contribution to cash. - Revises customer metrics when an intermediary changes access to identity and repeat behavior. ### Jeff Bezos · When the measurable price effect is not the whole decision Skill: Use quantitative evidence without pretending it measures an unavailable long-term effect #### The 2005 pricing policy · Historical reconstruction Period: 2005 pricing policy, reconstructed from the 2005 letter Amazon's short-term elasticity analysis generally predicts that a price reduction will increase units but not enough within a week or quarter to replace the lost price. Management cannot estimate numerically what repeated price reductions will do over five or ten years. Free Super Saver Shipping and Prime impose visible short-term costs. The earlier single-detail-page decision offers relevant but incomplete evidence. By 2005, third-party units have risen from 6 percent of total units in 2000 to 28 percent while Amazon retail revenue has tripled. Both channels grew, but the counterfactual is unknown. The public letter does not provide a program-level forecast for future price reductions or document a specific annual review meeting. Decision prompt: Choose the price policy for the next planning period using only the evidence known in 2005. Explain which result the elasticity model can answer, which result requires judgment, and how you would bound and review the policy. Options: - Follow short-term elasticity and raise prices where current revenue improves. - Return scale efficiencies through broad lower prices and accept near-term loss. - Use category-level price and seller tests with long-term customer measures and inventory limits. - Hold broad prices and run bounded category tests before changing policy. Debrief: The 2005 letter says Amazon had chosen to reduce prices continuously and significantly as efficiency and scale made reductions possible. Management accepted the short-term result shown by its elasticity work because it judged that lower prices would strengthen the long-run customer relationship and free cash flow. Free Super Saver Shipping, Prime and the earlier single detail page were cited as related judgment calls. The 2002 letter had reported a narrower basket in which Amazon was cheaper on 72 of 100 bestselling books and equal on 25. That basket and the known Marketplace growth were evidence available in 2005, but neither could calculate the future policy's five-to-ten-year return. The selected historical record does not provide a later controlled outcome for the 2005 policy. Sources: - Jeffrey P. Bezos, 2002 Letter to Shareholders (2002 reporting period; published 2003), Opening discussion of customer experience, price, satisfaction, the 100-book comparison, and free cash flow. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2002_shareholderLetter.pdf Evidence note: Contemporary company evidence. The 100-book basket, satisfaction score and management's free-cash-flow definition have specific scopes and do not prove a universal price rule. - Jeffrey P. Bezos, 2005 Letter to Shareholders (2005 reporting period; published 2006), Opening discussion of math-based and judgment-based decisions, price reductions, and the single detail page. https://ir.aboutamazon.com/files/doc_financials/annual/AMZN2005AnnualReport.pdf Evidence note: Contemporary account of Amazon's decision policy with retrospective results for the Marketplace choice made in 2000. Long-run price effects were explicitly not measurable in advance. - Jeffrey P. Bezos, 2014 Letter to Shareholders (2014 reporting period; published 2015), Opening business-quality test and the Marketplace, Prime, FBA and AWS sections. https://ir.aboutamazon.com/files/doc_financials/annual/AMAZON-2014-Shareholder-Letter.pdf Evidence note: Retrospective account of several initiatives after they had shown traction. Management's business-quality assessments and causal explanations remain judgments. #### The service-platform price cut · Constructed 2026 scenario Period: June 30, 2026 All dollar figures are fictional United States dollar assumptions. A service platform has 100,000 accounts paying an average $1,000 a year, for a $100 million annualized recurring-revenue run rate. Variable service cost is $200 per account, so the current annualized contribution run rate before fixed cost is $80 million. A 10 percent price cut to $900 is estimated to raise accounts 6 percent within twelve months to 106,000. At that ending account count, the annualized run rates are $95.4 million of revenue and $74.2 million of contribution at the same unit cost, a $5.8 million contribution decline from the current run rate. The model omits customer-arrival and billing timing, so these are not actual first-year cash results. It also does not estimate effects on later churn, referrals, customer trust or future product adoption. Decision prompt: Decide whether and how to cut price. Show the annualized ending run-rate arithmetic, identify the arrival and billing schedule needed to calculate first-year cash, and propose evidence that could justify the $5.8 million run-rate contribution decline. Options: - Cut all account prices by 10 percent for at least two years. - Run a bounded new-customer and renewal-cohort test with pre-set contribution and retention gates. - Keep price and invest an equivalent amount in measurable service improvements. - Raise selected prices where short-term willingness to pay supports it. Debrief: The current annualized revenue run rate is 100,000 times $1,000, or $100 million. Current annualized contribution is 100,000 times $800, or $80 million. At the proposed ending account count, annualized revenue is 106,000 times $900, or $95.4 million. Annualized contribution is 106,000 times $700, or $74.2 million, a $5.8 million decline from the current run rate. Actual first-year recognized revenue, contribution and cash require the missing customer-arrival, churn and billing schedule. A longer test needs cohort retention, expansion, referrals, service use and support cost. A universal cut creates weak comparison data, while a cohort test can be contaminated by customer communication and competitive response. Judgment remains necessary, but the unmeasured long-term benefit must be large enough to repay a visible near-term cost. Sources: - Jeffrey P. Bezos, 2002 Letter to Shareholders (2002 reporting period; published 2003), Opening discussion of customer experience, price, satisfaction, the 100-book comparison, and free cash flow. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2002_shareholderLetter.pdf Evidence note: Contemporary company evidence. The 100-book basket, satisfaction score and management's free-cash-flow definition have specific scopes and do not prove a universal price rule. - Jeffrey P. Bezos, 2005 Letter to Shareholders (2005 reporting period; published 2006), Opening discussion of math-based and judgment-based decisions, price reductions, and the single detail page. https://ir.aboutamazon.com/files/doc_financials/annual/AMZN2005AnnualReport.pdf Evidence note: Contemporary account of Amazon's decision policy with retrospective results for the Marketplace choice made in 2000. Long-run price effects were explicitly not measurable in advance. - Jeffrey P. Bezos, 2014 Letter to Shareholders (2014 reporting period; published 2015), Opening business-quality test and the Marketplace, Prime, FBA and AWS sections. https://ir.aboutamazon.com/files/doc_financials/annual/AMAZON-2014-Shareholder-Letter.pdf Evidence note: Retrospective account of several initiatives after they had shown traction. Management's business-quality assessments and causal explanations remain judgments. #### Transparent agent comparison · Hypothetical future stress test Period: 2031 annual pricing decision This is a fictional future stress test, and all dollar figures are United States dollar assumptions. The platform now has 120,000 accounts paying $900 a year. Automation has reduced variable service cost to $120 per account. Current annualized run-rate revenue is $108 million and contribution before fixed cost is $93.6 million. Buying agents can compare service quality and total price instantly. A credible competitor charges $700. Management estimates that cutting to $720 would increase accounts 18 percent within one year to 141,600. At that ending account count and constant $120 unit cost, annualized run-rate revenue would be $101.952 million and contribution $84.96 million. The model omits customer-arrival and billing timing, so these are not actual first-year results. It also does not estimate multi-year churn or a competitor response. Decision prompt: Revise the 2026 choice for a market with transparent automated comparison. Decide whether to match price, differentiate, segment or accept share loss. Distinguish the ending run rate from first-year cash and identify one model assumption most likely to fail. Options: - Cut average price to $720 and pursue the estimated account growth. - Keep $900 and invest in service attributes that agents can verify. - Offer a $720 standardized plan through agents and retain a higher-service direct plan. - Keep price, accept account losses and maximize near-term contribution. Debrief: At $900, annualized contribution is 120,000 times $780, or $93.6 million. At $720 with the ending count of 141,600 accounts, annualized contribution is 141,600 times $600, or $84.96 million, an $8.64 million run-rate decline. Annualized run-rate revenue falls by $6.048 million. Actual first-year cash requires a customer-arrival and billing schedule that the scenario does not supply. The future channel may make price more visible, but it may also make verified quality easier to reward. Segmentation can reveal willingness to pay and protect a service tier, though customers may migrate and complexity may raise cost. The choice depends on retention beyond year one, response by the $700 competitor, agent ranking rules and whether quality differences are measurable. The 2005 lesson supports naming what the short-term model omits. It does not dictate a price cut. Sources: - Jeffrey P. Bezos, 2002 Letter to Shareholders (2002 reporting period; published 2003), Opening discussion of customer experience, price, satisfaction, the 100-book comparison, and free cash flow. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2002_shareholderLetter.pdf Evidence note: Contemporary company evidence. The 100-book basket, satisfaction score and management's free-cash-flow definition have specific scopes and do not prove a universal price rule. - Jeffrey P. Bezos, 2005 Letter to Shareholders (2005 reporting period; published 2006), Opening discussion of math-based and judgment-based decisions, price reductions, and the single detail page. https://ir.aboutamazon.com/files/doc_financials/annual/AMZN2005AnnualReport.pdf Evidence note: Contemporary account of Amazon's decision policy with retrospective results for the Marketplace choice made in 2000. Long-run price effects were explicitly not measurable in advance. - Jeffrey P. Bezos, 2014 Letter to Shareholders (2014 reporting period; published 2015), Opening business-quality test and the Marketplace, Prime, FBA and AWS sections. https://ir.aboutamazon.com/files/doc_financials/annual/AMAZON-2014-Shareholder-Letter.pdf Evidence note: Retrospective account of several initiatives after they had shown traction. Management's business-quality assessments and causal explanations remain judgments. #### Reasoning rubric - Computes the annualized ending revenue and contribution run rates correctly and does not mislabel them as first-year cash. - States the horizon and variables the elasticity model can and cannot estimate. - Bounds the judgment with cohorts, inventory or cash limits and review triggers. - Avoids treating later Marketplace growth as proof of a missing counterfactual. - Revises for agent transparency and considers competitor response, segmentation and verified quality. ### Howard Marks · Buying while the system is in doubt Skill: Separate price from value, preserve survival, and pace a contrarian commitment #### The system may not hold · Historical reconstruction Period: September 19, 2008 Lehman Brothers has filed for bankruptcy after earlier failures and rescues. Funding markets are impaired, several major institutions appear vulnerable, and the possibility of a broader financial collapse cannot be dismissed. Prices of many credit claims have fallen sharply. Some declines may reflect default risk. Others may reflect margin calls, redemptions, and an absence of willing buyers. You manage a distressed-credit program with committed capital, but you do not know where the bottom lies. The information available does not support a confident macro forecast. This reconstruction stops before Oaktree's later public account of its purchases. Decision prompt: Choose a deployment posture for the next eight weeks. Give a rationale that separates fundamental impairment from forced selling, states how the program survives a further fall, and names evidence that would stop additional buying. Options: - Hold nearly all available capital until financial conditions stabilize, accepting that prices may recover before certainty arrives. - Deploy most available capital immediately because current prices already reflect a system-ending outcome. - Buy selected claims in planned increments, re-underwrite value after each material event, and preserve capacity for lower prices. - Buy selected claims while spending part of the expected return on explicit protection against a deeper system shock. Debrief: Marks wrote that the future was unknowable but that treating collapse as the sole premise produced no workable course. By October 15, he reported that Oaktree's purchase lists had been long on most days, sales almost absent, and available cash deployed while the firm averaged down. In 2020 he retrospectively reported an average pace of US$450 million per week over the final 15 weeks of 2008. That weekly figure implies US$6.75 billion across the period, but it does not reveal the daily rule, assets, client allocations, or later return. A paced answer fits the public record most closely, but another option can be defensible if it provides durable funding, prices protection honestly, and states a falsifiable value case. Hindsight about the recovery is not part of the rubric. Sources: - Howard Marks, Nobody Knows (2008-09-19), Does the Market Know?; The End of the Financial System; final bear-market-stage discussion. https://www.oaktreecapital.com/insights/memo/nobody-knows Evidence note: Contemporaneous memo after Lehman's bankruptcy. It records a decision premise and a search for bargains, but no security-level trade list. - Howard Marks, The Limits to Negativism (2008-10-15), Final discussion of third-stage pessimism and Oaktree purchases. https://www.oaktreecapital.com/insights/memo/the-limits-to-negativism Evidence note: Contemporaneous practitioner report. It describes purchase direction and averaging down, but not positions, allocations, or later investment results. - Howard Marks, Uncertainty (2020-05-11), All We Don't Know; In Praise of Doubt; confidence and position-sizing discussion. https://www.oaktreecapital.com/insights/memo/uncertainty Evidence note: The discussion of 2008 includes a retrospective practitioner report. The memo distinguishes evidence-based confidence from certainty. #### The bargain is tied to one fragile cash-flow source · Constructed 2026 scenario Period: April 6, 2026 This is a fictional case stated in US dollars. A closed-end credit fund has US$600 million of uncalled commitments and no redemption obligation before 2031. A forced liquidation by two levered funds has pushed a portfolio of senior data-center loans from US$96 to US$63 per US$100 of par value in ten trading days. Your base case gives a recovery value of US$82 to US$90 per US$100 of par within three years. Yet 58 percent of borrower cash flow comes from one power market, and a pending fictional regulatory decision on June 30, 2026 could raise annual energy costs by US$140 million. A US$45 recovery per US$100 of par is plausible if borrowers cannot pass through those costs. The fund can invest at most US$240 million in this theme under its concentration policy. Decision prompt: Choose today's commitment and write the next two decision dates. Quantify how much capacity you preserve, identify the evidence that would move the 45-cent case, and explain whether the current price reflects liquidity, impairment, or both. Options: - Commit $0 now and wait for the June 30 regulatory decision. - Commit $60 million now, reserve $180 million, and re-underwrite after the regulation and first-quarter borrower reporting. - Commit $120 million now, reserve $120 million, and require asset-level covenant and power-cost tests before any second purchase. - Commit the full $240 million theme limit now because the base-case recovery spread is large. Debrief: Several choices can be well reasoned. Waiting buys information but risks losing a liquidity-driven price. A $60 million or $120 million first step recognizes both the forced seller and the unresolved cash-flow impairment. Full deployment requires unusually strong evidence that the regulatory loss is already captured and that no later capital has greater option value. A good rationale does not call 63 cheap merely because the loans traded at 96. It rebuilds recovery from borrower cash flows, tests claim priority, and shows that the closed-end structure removes redemption risk without removing asset risk. Sources: - Howard Marks, Nobody Knows (2008-09-19), Does the Market Know?; The End of the Financial System; final bear-market-stage discussion. https://www.oaktreecapital.com/insights/memo/nobody-knows Evidence note: Contemporaneous memo after Lehman's bankruptcy. It records a decision premise and a search for bargains, but no security-level trade list. - Howard Marks, The Limits to Negativism (2008-10-15), Final discussion of third-stage pessimism and Oaktree purchases. https://www.oaktreecapital.com/insights/memo/the-limits-to-negativism Evidence note: Contemporaneous practitioner report. It describes purchase direction and averaging down, but not positions, allocations, or later investment results. - Howard Marks, Uncertainty (2020-05-11), All We Don't Know; In Praise of Doubt; confidence and position-sizing discussion. https://www.oaktreecapital.com/insights/memo/uncertainty Evidence note: The discussion of 2008 includes a retrospective practitioner report. The memo distinguishes evidence-based confidence from certainty. - Howard Marks, The Tide Goes Out (2008-03-18), Leveraging and Inflating; Mark-to-Market Accounting; Should Does Not Equal Will; Leverage and Risk. https://www.oaktreecapital.com/insights/memo/the-tide-goes-out Evidence note: Contemporaneous analysis written before the worst phase of the financial crisis. Examples and conclusions express the author's assessment and are not an independent performance study. #### Liquidity arrives before valuation does · Hypothetical future stress test Period: November 17, 2032 This is a fictional future case stated in US dollars. On October 6, 2032, the fund buys senior data-center loans at US$64 per US$100 of par. Six weeks later, a regulatory settlement removes the extreme cost case. An automated collateral exchange values the loans at US$78 per US$100 of par and offers four-year non-recourse financing equal to 60 percent of that value. The financing is marked daily. If the exchange value remains below US$70 for five trading days, the collateral is automatically sold and the loan is extinguished with no further claim on the fund. Market bids rise to US$76. Your revised unlevered recovery range is US$84 to US$91 over 30 months. The fund owns US$90 million of par. Equally senior water-infrastructure loans now trade at US$68 with an estimated recovery range of US$86 to US$94 and lower common-factor exposure. You have US$180 million of remaining investment capacity across both themes. Decision prompt: Revise your plan. Decide whether to add, hold, trim, finance, or redirect capital. Explain how the new financing changes path risk, how the price recovery changes prospective return, and whether the second opportunity improves the portfolio. Options: - Use the new financing and add to the original loans up to the theme limit. - Keep the original position unlevered and direct new capital to the water-infrastructure loans after underwriting. - Trim the original position near 76 and divide proceeds plus unused capacity between cash and the second theme. - Hold the original position and retain all remaining capacity because both recovery estimates remain uncertain. Debrief: The original purchase does not require another purchase. At US$76 per US$100 of par, the original loans offer less prospective return than they did at US$64. The financing avoids a claim beyond the collateral, but its daily valuation and automatic-sale rule can still remove the position during a temporary decline and surrender a later recovery. Its four-year term is longer than the 30-month recovery estimate, so refinancing is not the immediate issue. Diversifying can improve the set if the water-loan recovery work is genuinely independent. Trimming can be rational if the market now pays for much of the resolved uncertainty. Holding cash can also be defensible when both value ranges depend on weak evidence. The strong revision changes because price, impairment risk, financing path, and the alternative opportunity have changed. Sources: - Howard Marks, Nobody Knows (2008-09-19), Does the Market Know?; The End of the Financial System; final bear-market-stage discussion. https://www.oaktreecapital.com/insights/memo/nobody-knows Evidence note: Contemporaneous memo after Lehman's bankruptcy. It records a decision premise and a search for bargains, but no security-level trade list. - Howard Marks, The Limits to Negativism (2008-10-15), Final discussion of third-stage pessimism and Oaktree purchases. https://www.oaktreecapital.com/insights/memo/the-limits-to-negativism Evidence note: Contemporaneous practitioner report. It describes purchase direction and averaging down, but not positions, allocations, or later investment results. - Howard Marks, Uncertainty (2020-05-11), All We Don't Know; In Praise of Doubt; confidence and position-sizing discussion. https://www.oaktreecapital.com/insights/memo/uncertainty Evidence note: The discussion of 2008 includes a retrospective practitioner report. The memo distinguishes evidence-based confidence from certainty. - Howard Marks, The Tide Goes Out (2008-03-18), Leveraging and Inflating; Mark-to-Market Accounting; Should Does Not Equal Will; Leverage and Risk. https://www.oaktreecapital.com/insights/memo/the-tide-goes-out Evidence note: Contemporaneous analysis written before the worst phase of the financial crisis. Examples and conclusions express the author's assessment and are not an independent performance study. - Howard Marks, The Anatomy of a Rally (2020-06-18), Market chronology; psychology of the rally; final assessment questions. https://www.oaktreecapital.com/insights/memo/the-anatomy-of-a-rally Evidence note: Contemporaneous market assessment after the spring rally. The judgment is not evidence of later excess returns. - Howard Marks, Coming into Focus (2020-10-13), The Power of Interest Rates; crisis comparison; final portfolio-calibration discussion. https://www.oaktreecapital.com/insights/memo/coming-into-focus Evidence note: Retrospective report on Oaktree's pre-crisis posture and March buying, followed by a dated October assessment. It does not disclose portfolio allocations. #### Reasoning rubric - Separates a forced-sale price from an independently built recovery or value range. - Maps funding, concentration, and forced-action risks before discussing upside. - Uses staged commitment as a revisable choice and states what stops further buying. - Recalculates prospective return from each new price rather than defending the original purchase. - Treats the historical result as evidence about a process, not proof that panic buying always works. ### Howard Marks · Prepare before the shock, then reassess the rally Skill: Compare opposing scenarios, act without a macro forecast, and reset exposure as price changes #### A violent fall and an equally surprising bounce · Historical reconstruction Period: March 31, 2020 The S&P 500 fell 33.9 percent from February 19 to March 23, 2020, then recorded its strongest three-day rise since the 1930s. The pandemic path, depth of economic contraction, corporate defaults, and effectiveness of government support remain deeply uncertain. The Federal Reserve and Treasury have announced massive measures. On March 31, Marks judges that the bounce reflects too much optimism and that prices leave insufficient room for worsening news. Oaktree entered the episode more cautiously than usual while trying to remain invested, and some drawdown funds with committed capital have already bought selected March bargains. The exact allocations and security list are undisclosed. You must now set the posture after the bounce. Decision prompt: Set the posture for existing portfolios and available drawdown capital. Explain how you can buy selected distress while also believing the broad market bounce is too optimistic. Define the event that triggers a fresh calibration. Options: - Increase broad market exposure immediately because policy support has changed the distribution. - Keep existing cautious portfolios, buy selected distressed claims where price compensates for uncertainty, and preserve capacity for renewed declines. - Increase defensiveness across portfolios and make no new purchases until economic data stabilize. - Move substantially to cash and wait for a clear recovery path. Debrief: Marks's October memo later reported that Oaktree's earlier caution reduced remediation and enabled drawdown funds with available capital to buy bargains when availability peaked in March. It does not disclose allocations or say every strategy acted the same way. By the June 18 memo, after the S&P 500 closed June 17 at 3,113, about 8 percent below its February peak of 3,386, Marks judged listed-security odds unattractive. In October he described the bargain period as brief. The record supports selective action and later restraint rather than one permanent risk label. A different option may still be reasoned, but it must handle both current price and the cost of losing future choice. Sources: - Howard Marks, Which Way Now? (2020-03-31), The Positive Case; the negative case; The Government Programs; Summing Up. https://www.oaktreecapital.com/insights/memo/which-way-now Evidence note: Dated scenario analysis and author judgment during the pandemic shock. It is not an Oaktree transaction ledger. - Howard Marks, The Anatomy of a Rally (2020-06-18), Market chronology; psychology of the rally; final assessment questions. https://www.oaktreecapital.com/insights/memo/the-anatomy-of-a-rally Evidence note: Contemporaneous market assessment after the spring rally. The judgment is not evidence of later excess returns. - Howard Marks, Coming into Focus (2020-10-13), The Power of Interest Rates; crisis comparison; final portfolio-calibration discussion. https://www.oaktreecapital.com/insights/memo/coming-into-focus Evidence note: Retrospective report on Oaktree's pre-crisis posture and March buying, followed by a dated October assessment. It does not disclose portfolio allocations. #### Policy support lifts prices before earnings recover · Constructed 2026 scenario Period: May 18, 2026 This is a fictional case. A pension portfolio began 2026 with 38 percent public credit, 32 percent public equity, 20 percent government bonds, and 10 percent cash. A fictional cyber outage then interrupts settlement and billing across several industries. Public credit spreads widen from 280 to 690 basis points in three weeks. The pension invests 3 percentage points of cash in senior credits at an estimated 11 percent yield. A government guarantee restores settlement, spreads narrow to 390 basis points, and the new credits gain 14 percent in six weeks. Operating earnings for the affected issuers are still unknown. The pension is now two percentage points above its public-credit limit because other asset values moved. Decision prompt: Choose the next action. Explain whether the prior purchase thesis, today's prospective return, the policy change, and the portfolio limit point in the same direction. State what evidence you need from issuer reporting. Options: - Add another 2 percentage points because policy support confirms the recovery. - Hold every new credit despite the limit and wait for issuer reporting. - Trim enough to restore the public-credit limit while retaining claims with the strongest cash-flow coverage. - Sell all crisis purchases and lock in the six-week gain. Debrief: The gain is not evidence that the original thesis remains attractive at the new price. The government guarantee may reduce settlement risk while leaving operating impairment unresolved. Restoring the risk limit can preserve the portfolio's ability to act later and need not express a macro forecast. Holding can be defensible if the limit is a soft governance band and claim-level coverage is strong, but the rationale should make that exception explicit. Adding requires a new prospective-return case, not a reference to the 14 percent already earned. Sources: - Howard Marks, Which Way Now? (2020-03-31), The Positive Case; the negative case; The Government Programs; Summing Up. https://www.oaktreecapital.com/insights/memo/which-way-now Evidence note: Dated scenario analysis and author judgment during the pandemic shock. It is not an Oaktree transaction ledger. - Howard Marks, The Anatomy of a Rally (2020-06-18), Market chronology; psychology of the rally; final assessment questions. https://www.oaktreecapital.com/insights/memo/the-anatomy-of-a-rally Evidence note: Contemporaneous market assessment after the spring rally. The judgment is not evidence of later excess returns. - Howard Marks, Coming into Focus (2020-10-13), The Power of Interest Rates; crisis comparison; final portfolio-calibration discussion. https://www.oaktreecapital.com/insights/memo/coming-into-focus Evidence note: Retrospective report on Oaktree's pre-crisis posture and March buying, followed by a dated October assessment. It does not disclose portfolio allocations. #### The hedge removes the visible risk and adds a hidden one · Hypothetical future stress test Period: February 18, 2027 This is a fictional future case continuing the May 18, 2026 episode. The pension trims to its credit limit. Nine months later, on February 18, 2027, issuer cash flows have recovered to 92 percent of the pre-outage run rate, but a new clearing rule requires every credit fund to post daily tokenized collateral against price volatility. A bank offers a cheap collateral facility that covers 95 percent of expected calls. Its legal terms permit immediate withdrawal after a model change. Using the facility would allow the pension to regain the income sold during the trim. Without it, the pension must accept a projected annual portfolio return 0.7 percentage points below its actuarial target for the next three fictional years. Decision prompt: Revise the portfolio plan. Decide whether to use the facility, accept the lower projected return, reduce another risk elsewhere, or redesign the mandate. Explain the new forced-action path and how much confidence you place on the 0.7-point projection. Options: - Use the full collateral facility and restore the credit exposure. - Use the facility for only one-third of the proposed exposure and hold liquid reserves against withdrawal. - Decline the facility and accept the projected 0.7-point annual shortfall. - Decline the facility and ask trustees to revise benefits, contributions, or the actuarial target rather than hide the gap in financing risk. Debrief: The facility transforms an asset-allocation problem into a financing-path problem. Its low expected cost does not remove the asymmetric consequence of withdrawal during stress. Partial use can bound that exposure if reserves are sized to an explicit call path. Accepting a lower projection may be prudent, but the actuarial consequences need governance rather than silence. Redesign may be the most transparent answer when the return target cannot be met safely, though it carries stakeholder costs. The reasoning should discount the precision of a three-year 0.7-point forecast and show which obligations remain robust if it is wrong. Sources: - Howard Marks, Which Way Now? (2020-03-31), The Positive Case; the negative case; The Government Programs; Summing Up. https://www.oaktreecapital.com/insights/memo/which-way-now Evidence note: Dated scenario analysis and author judgment during the pandemic shock. It is not an Oaktree transaction ledger. - Howard Marks, The Anatomy of a Rally (2020-06-18), Market chronology; psychology of the rally; final assessment questions. https://www.oaktreecapital.com/insights/memo/the-anatomy-of-a-rally Evidence note: Contemporaneous market assessment after the spring rally. The judgment is not evidence of later excess returns. - Howard Marks, Coming into Focus (2020-10-13), The Power of Interest Rates; crisis comparison; final portfolio-calibration discussion. https://www.oaktreecapital.com/insights/memo/coming-into-focus Evidence note: Retrospective report on Oaktree's pre-crisis posture and March buying, followed by a dated October assessment. It does not disclose portfolio allocations. - Howard Marks, The Tide Goes Out (2008-03-18), Leveraging and Inflating; Mark-to-Market Accounting; Should Does Not Equal Will; Leverage and Risk. https://www.oaktreecapital.com/insights/memo/the-tide-goes-out Evidence note: Contemporaneous analysis written before the worst phase of the financial crisis. Examples and conclusions express the author's assessment and are not an independent performance study. - Howard Marks, Uncertainty (2020-05-11), All We Don't Know; In Praise of Doubt; confidence and position-sizing discussion. https://www.oaktreecapital.com/insights/memo/uncertainty Evidence note: The discussion of 2008 includes a retrospective practitioner report. The memo distinguishes evidence-based confidence from certainty. #### Reasoning rubric - Can hold a cautious broad view while acting on a specific price without contradiction. - Separates a policy-driven price recovery from recovery in issuer cash flows. - Revises exposure after price and portfolio weights change. - Treats financing terms as a source of forced-action risk. - Matches confidence and position size to the quality of the forecast evidence. ### Mark Leonard · When the old acquisition channel cannot absorb the cash Skill: Diagnose the deployment constraint before changing hurdles or entering a new domain #### Discipline creates an accumulation problem · Historical reconstruction Period: Before the February 2021 board decision Constellation has long kept high hurdles for small and midsized vertical-market software acquisitions. The discipline has left it unable to invest all cash available to shareholders. It has previously paid three special dividends and has paid a regular quarterly dividend for about a decade. Large VMS deals have absorbed less than 10 percent of the relevant cash, and only three have been completed during the company's 26-year history. Leonard reports that 40 to 70 large VMS businesses are sold annually. During the preceding five years CSI knew of about 80 percent of the large VMS businesses sold, yet brokers invited it into only 16 percent of the sale processes. The letter does not define whether that denominator covers all large VMS sale processes or only the roughly 80 percent CSI knew about. This reconstruction stops before the board's decision. Decision prompt: Recommend a capital policy to the board. Address special and regular dividends, the hurdle for existing small deals, access to large deals, and the risk of building competence outside VMS. State which constraint each action is meant to solve. Options: - Preserve all hurdles, continue regular and special dividends, and accept slower reinvestment. - Lower the hurdle across all acquisition sizes so operating groups can deploy nearly all available cash. - Keep traditional hurdles for small and midsized VMS deals, stop routine special dividends, and build separate teams and policies for large VMS and a bounded new domain. - Preserve existing dividends and fund a large-deal team only after a signed acquisition requires capital. Debrief: The February 15 letter records a segmented policy. The board stopped special dividends except in compelling circumstances and kept the regular quarterly dividend for then. Traditional hurdles remained for most small and midsized VMS acquisitions. Head office began building a dedicated large-deal team, considered lower hurdles for those deals, and started exploring a competence outside VMS. Leonard expected that completing one or two large VMS acquisitions per year would lower return on investors' capital while reducing the need to return cash. This is a decision and intended search, not evidence that later investments succeeded. A capital-return option remains defensible if its rationale judges the new-domain and large-deal risks to exceed their prospective returns. Sources: - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. - Mark Leonard, Constellation Software Inc. 2017 President's Letter (2018-04-20), Performance metrics; retained-capital policy; perpetual-owner objective; business-unit count. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-april-2018-final.pdf Evidence note: Management report. Statements about scalability and governance are Leonard's judgments, not independent causal findings. - Mark Leonard, Constellation Software Inc. 2021 President's Letter (2021-02-15), Complete letter, especially dividend policy, large VMS acquisitions, and head-office role. https://www.csisoftware.com/wp-content/uploads/2026/04/letter-to-shareholders-february-15-2021.pdf Evidence note: Records a board policy decision and management intentions as of publication. The reported 16 percent invitation figure does not have an unambiguous denominator in the letter. It does not establish the results of later investments. #### The constraint may be access rather than price · Constructed 2026 scenario Period: July 14, 2026 This is a fictional case. A permanent software owner generates US$420 million of free cash flow during the latest twelve months. Its operating groups can deploy about US$260 million annually in small acquisitions with modeled after-tax IRRs of 16 to 20 percent. The remaining US$160 million accumulates. A proposed large-deal policy uses a 12 percent hurdle. The company has been invited into only 9 of 72 known large sale processes during the latest 24 months. Building a dedicated team costs US$7 million annually and is expected to raise invitations, though no improvement is assured. The board can also return the excess cash or authorize a new healthcare-services search with a US$40 million first-year commitment cap. Decision prompt: Allocate the next twelve months of cash among existing acquisitions, access-building, large deals, a new domain, and shareholder returns. State the hurdle and evidence gate for each channel. Explain why the same hurdle should or should not apply to all of them. Options: - Deploy US$260 million in the existing channel and return the estimated US$160 million excess. - Fund the US$7 million large-deal team for two years, preserve the 12 percent large-deal hurdle, and return cash that remains unused. - Fund the large-deal team, use a separate 11 to 13 percent review band for large deals, cap the new-domain search at US$40 million, and retain the rest temporarily. - Lower all acquisition hurdles by three percentage points and require teams to deploy the full US$420 million. Debrief: The figures are fictional assumptions. The 9-of-72 invitation record points first to an access problem, so paying more across every channel does not directly solve the diagnosed constraint. A dedicated team can be treated as a measured sourcing investment, with invitation rate, qualified opportunities, and avoided bids tracked before any hurdle change. A separate large-deal review band can be reasonable because scale, competition, and ownership demands differ, but a band must not become permission to adjust forecasts until they fit. The healthcare search needs a written competence thesis and loss cap. Returning unused cash remains a real benchmark, not a failure of imagination. Sources: - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. - Mark Leonard, Constellation Software Inc. 2017 President's Letter (2018-04-20), Performance metrics; retained-capital policy; perpetual-owner objective; business-unit count. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-april-2018-final.pdf Evidence note: Management report. Statements about scalability and governance are Leonard's judgments, not independent causal findings. - Mark Leonard, Constellation Software Inc. 2021 President's Letter (2021-02-15), Complete letter, especially dividend policy, large VMS acquisitions, and head-office role. https://www.csisoftware.com/wp-content/uploads/2026/04/letter-to-shareholders-february-15-2021.pdf Evidence note: Records a board policy decision and management intentions as of publication. The reported 16 percent invitation figure does not have an unambiguous denominator in the letter. It does not establish the results of later investments. #### More access exposes the quality of the policy · Hypothetical future stress test Period: January 20, 2031 This is a fictional future case. A measured large-deal program begins on January 20, 2029. By January 20, 2031, the team has raised invitations from 12.5 percent to 54 percent of known processes and cost US$15 million in total. It closed one US$520 million acquisition on July 20, 2029 at an underwritten 12.4 percent after-tax IRR. Eighteen months later, that forecast has fallen to 9.6 percent because customer migrations cost more than modeled. The small-deal channel still forecasts 17 percent but can absorb only US$300 million of the latest US$510 million annual free cash flow. The healthcare search has spent US$28 million, completed no acquisition, and produced evidence that customer contracts are less durable than software maintenance agreements. A second large target is available at a modeled 12.2 percent IRR. Decision prompt: Revise the policy before bidding on the second large target. Decide what the first deal and healthcare evidence change, how much cash to retain or return, and whether the large-deal hurdle or model needs repair. Options: - Proceed under the existing 12 percent large-deal hurdle because one miss is insufficient to change the policy. - Pause new large bids, review the first deal's forecast errors, rebuild the migration-cost base rate, and return cash beyond a defined two-year opportunity reserve. - Raise the large-deal hurdle to 15 percent immediately and close the healthcare search. - Keep the 12 percent hurdle and redirect all undeployed cash to the small-deal groups by lowering their local approval limits. Debrief: More invitations show that the sourcing investment changed access. The first deal's revised 9.6 percent forecast raises a different question about underwriting and integration. A pause can protect the decision rule while the firm classifies the error and updates the relevant base rate. An immediate hurdle increase may be too coarse if one specific migration assumption caused the miss. Proceeding can be defensible if the second target is not exposed to that assumption and the rationale demonstrates the difference. The healthcare evidence weakens the original competence thesis, so continuation requires a narrower experiment or closure. Retaining cash should be tied to a defined opportunity reserve rather than an indefinite promise to deploy. Sources: - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. - Mark Leonard, Constellation Software Inc. 2017 President's Letter (2018-04-20), Performance metrics; retained-capital policy; perpetual-owner objective; business-unit count. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-april-2018-final.pdf Evidence note: Management report. Statements about scalability and governance are Leonard's judgments, not independent causal findings. - Mark Leonard, Constellation Software Inc. 2021 President's Letter (2021-02-15), Complete letter, especially dividend policy, large VMS acquisitions, and head-office role. https://www.csisoftware.com/wp-content/uploads/2026/04/letter-to-shareholders-february-15-2021.pdf Evidence note: Records a board policy decision and management intentions as of publication. The reported 16 percent invitation figure does not have an unambiguous denominator in the letter. It does not establish the results of later investments. - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. #### Reasoning rubric - Diagnoses sourcing access, market size, pricing, and underwriting as distinct constraints. - Uses different policies only where deal size or domain creates a relevant difference. - Treats returning capital as an explicit benchmark for forced deployment. - Uses bounded search costs and evidence gates for a new competence. - Updates the model after a miss without assuming one outcome proves or disproves the entire policy. ### Mark Leonard · The deal that missed by a few points Skill: Use a pre-committed hurdle to resist sunk effort and assign end-to-end ownership #### Months of work do not improve the forecast · Historical reconstruction Period: Before April 25, 2017, as reported in the 2016 president's letter Leonard is personally involved in a large acquisition process. As work accumulates, he feels more committed even though the news is not improving. The proposed investment falls just short of CSI's hurdle. The team tries to negotiate a structure that would add a few points to projected IRR but cannot secure it. Large transactions consume senior attention and can have portfolio-level consequences. CSI's exact hurdle, target, price, and forecast are not public. This reconstruction stops before the reported decision. Decision prompt: Decide whether to proceed, walk away, or redesign the proposal. Explain how you treat sunk diligence cost, forecast uncertainty, deal size, and responsibility for integration. State what evidence would justify an exception. Options: - Proceed because the forecast miss is small and the strategic opportunity may not return. - Walk away because the deal remains below the pre-committed hurdle after the structure fails to improve. - Proceed as a documented exception with less capital, a named owner, and a board-approved learning objective. - Submit a final lower price and stop work automatically if the seller rejects it. Debrief: Leonard reports that CSI did not invest after the transaction failed to meet the hurdle and the team could not negotiate the extra projected return. He used the experience to emphasize how time spent can create commitment without better evidence. He also concluded that a very large acquisition should have one highly experienced acquirer responsible for diligence, structure, negotiation, integration, and outcome. The record does not reveal whether a lower final bid was made or what the target later achieved. Walking away matches the reported outcome, but the reasoning is judged by use of the prior rule and evidence, not imitation. Sources: - Mark Leonard, Constellation Software Inc. 2014 President's Letter (2015-04-06), Invested capital; debt duration; ROIC; TSS contribution to maintenance revenue. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: Company-reported figures include management-defined non-IFRS measures explained in the letter's glossary. - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. #### A near miss with visible sunk cost · Constructed 2026 scenario Period: November 3, 2026 This is a fictional case. A decentralized software group has spent US$1.8 million and seven months evaluating a US$340 million enterprise-software target. Its hurdle for this size is a 12 percent after-tax unlevered acquisition IRR. The IRR of the current probability-weighted acquisition cash flows is 11.7 percent over a ten-year modeled period. Both the hurdle and model exclude acquisition financing. Forty percent of the projected value depends on two-year customer-cohort retention improving from 88 to 93 percent. The group's own comparable acquisitions have a median two-year customer-cohort retention of 90 percent and a 75th percentile of 92 percent. The seller rejects a lower price but offers an earn-out that raises the IRR of the same probability-weighted, ten-year, after-tax unlevered acquisition cash flows to 12.3 percent. The revised model includes the contingent payments in each retention outcome before weighting the cash flows. The underlying retention probabilities remain uncertain. The executive sponsor also controls integration. Decision prompt: Choose a response. Explain whether the earn-out changes economics or only moves risk, how the retention assumption compares with the base rate, and whether the sponsor's dual role creates an accountability or review problem. Options: - Accept the earn-out because the model now clears the hurdle. - Walk away because the revised return still depends on an assumption beyond the group's observed 75th percentile. - Pause for an independent retention review and proceed only if evidence supports the 93 percent case without relying on the sponsor. - Offer a smaller minority investment with an option to acquire control after two years of retention evidence. Debrief: The numbers are fictional assumptions. The earn-out can transfer part of the price risk to the seller, but it does not itself cause customers to stay. The comparable probability-weighted cash-flow models cross from 11.7 to 12.3 percent, but this should not end the inquiry when 40 percent of value rests above the observed retention distribution. An independent review addresses the sponsor's growing commitment and dual role. Walking is disciplined if the evidence remains weak. A minority structure can buy information but may add governance risk and may be unavailable. The US$1.8 million already spent is relevant to process improvement, not to the target's future cash flows. Sources: - Mark Leonard, Constellation Software Inc. 2014 President's Letter (2015-04-06), Invested capital; debt duration; ROIC; TSS contribution to maintenance revenue. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: Company-reported figures include management-defined non-IFRS measures explained in the letter's glossary. - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. #### The rejected deal returns with better evidence and worse structure · Hypothetical future stress test Period: August 8, 2029 This is a fictional future case. You paused in 2026 and the seller chose another buyer, but that financing later failed. The target returns on August 8, 2029 at a US$300 million enterprise value, consisting of US$190 million of equity value and US$110 million of assumed floating-rate debt. Audited data for two-year customer cohorts show 92.5 percent retention and lower support costs. The modeled after-tax unlevered operating IRR on the full US$300 million of capital is 13.4 percent, independent of how the purchase is financed. The debt matures in 14 months. Refinancing is available only with a parent guarantee and a covenant allowing lenders to restrict acquisitions after a 20 percent revenue decline. A local operating-group president can lead the purchase, but no one has yet accepted responsibility for both refinancing and post-close integration. Decision prompt: Revise the 2026 decision. Decide whether improved operating evidence outweighs the financing path and ownership gap. Specify who must own the full decision and what structure would make the deal acceptable. Options: - Acquire at US$300 million and provide the parent guarantee because the modeled return now clears the hurdle. - Acquire only if the debt is repaid at close with parent equity, then assess the lower-risk but larger equity commitment. - Delay commitment until one experienced leader accepts refinancing and integration ownership and negotiates removal of the restrictive covenant. - Walk away again because no forecast return compensates for the current financing and ownership structure. Debrief: The better cohort evidence directly addresses the earlier retention uncertainty, so repeating the old rejection without revision would ignore learning. The 13.4 percent after-tax unlevered operating IRR uses the full US$300 million enterprise value and does not change when debt is repaid with equity. Repayment changes the amount and risk of equity funding and removes a forced refinancing path, so the learner must separately calculate the equity return and portfolio exposure under each structure. Delaying for a named end-to-end owner is reasonable if the seller permits it. A parent guarantee can make the acquisition appear locally safe while transferring fragility to the whole group. The strong answer assigns one accountable leader and does not count the favorable operating update twice. Sources: - Mark Leonard, Constellation Software Inc. 2014 President's Letter (2015-04-06), Invested capital; debt duration; ROIC; TSS contribution to maintenance revenue. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: Company-reported figures include management-defined non-IFRS measures explained in the letter's glossary. - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. #### Reasoning rubric - Treats time and diligence already spent as sunk when valuing future cash flows. - Tests forecast assumptions against relevant internal distributions rather than the hurdle alone. - Distinguishes a changed deal structure from a genuine improvement in the underlying economics. - Assigns one experienced owner across diligence, financing, integration, and outcome for a large commitment. - Revises the decision when new evidence arrives while recognizing newly introduced financing risk. ### Mark Leonard · The profitable interruption Skill: Distinguish stronger economics from deferred investment #### A company put up for sale · Historical reconstruction Period: 2011 strategic review; evidence through the May 2, 2011 letter Constellation’s board is reviewing strategic alternatives, and Leonard has publicly said the company is likely to be sold. The process requires management attention and may provide liquidity for major shareholders. Its eventual outcome is unknown. The business depends on managers who make investments with long payback periods and whose compensation includes a continuing relationship with the company. You are advising the board about how to conduct the review while protecting the underlying business. This is a reconstruction from the 2010 reporting-year letter published in 2011; the options are course constructions, not a transcript of the board’s actual alternatives. Decision prompt: How should the board pursue a legitimate liquidity objective while keeping investment decisions credible? State what apparent improvement in the next results would require investigation. Options: - Pursue the sale quickly and let managers prioritize near-term financial performance. - Continue the review with explicit protection and separate monitoring for long-term investment commitments. - Reassess the whole-company sale and compare other routes to shareholder liquidity. Debrief: The 2011 reporting-year letter, published in May 2012, describes a process that slowed acquisitions and long-term initiatives while improving short-term profits. Leonard believed uncertainty had damaged prospects, and he revised his view that fundamentals alone would take care of the share price. This is his retrospective explanation, alongside other factors such as economic recovery and acquisition mix. It does not prove that every alternative would have been better. The strongest initial response anticipates the incentive mechanism without needing to know the later result, and gives the board a way to distinguish withheld investment from genuine efficiency. Sources: - Mark Leonard, Constellation Software 2010 annual president’s letter (2011-05-02), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/shareholder-letter-2010.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2011 annual president’s letter (2012-05-02), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/2011_presidents_letter.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. #### An attractive sale story · Constructed 2026 scenario Period: Constructed 2026 scenario; all company details below are fictional A privately owned software group is exploring a sale in 2026. Its owners want liquidity. During diligence, operating margin rises from 18% to 24%, while cash improves. The product teams have postponed two integration projects, customer-support hiring is frozen, and managers do not know whether a new owner will honor their long-term reward arrangements. Buyers describe the higher margin as the new normal. The group has no evidence yet that customer retention has deteriorated. These percentages and operating details are scenario assumptions, not Constellation results or current market observations. Decision prompt: What margin would you present as sustainable, and what action would you take before accepting a buyer’s valuation based on the improved figure? Options: - Accept the current margin as sustainable because retention has not yet declined. - Present a bridge separating efficiency, temporary reductions and deferred investment, then restore work with a sound long-term case. - Keep all spending frozen until the sale closes to avoid complicating diligence. Debrief: The available facts do not support treating the whole margin increase as recurring efficiency. Some costs have been deferred, and customer effects may arrive after the reporting period. A defensible response estimates the spending required to sustain the product and service promise, labels uncertainty and explains the compensation transition. Restoring every postponed project automatically would be equally weak: some may not earn their cost. Test the project economics and monitor leading evidence such as unresolved integrations and support load. The analogy concerns incentives and timing, not a forecast that this fictional company must suffer Constellation’s reported experience. Sources: - Mark Leonard, Constellation Software 2010 annual president’s letter (2011-05-02), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/shareholder-letter-2010.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2011 annual president’s letter (2012-05-02), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/2011_presidents_letter.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. #### The delayed bill · Hypothetical future stress test Period: Hypothetical 2028 follow-up; all developments are fictional The sale did not happen. Some postponed work has now been completed, but several experienced managers have left and two large customers demand costly remediation. New automation has also reduced the cost of one integration project, while making another planned feature less valuable. The owners ask management to reverse every decision made during the sale process. You must decide what to rebuild, what to discontinue and how to restore confidence without pretending that the earlier plan remains optimal. The case does not supply enough evidence to attribute every departure or customer problem to the abandoned sale. Decision prompt: Which commitments should be restored, which should be reconsidered, and how will you avoid confusing a changed environment with proof that every earlier decision was wrong? Options: - Restore the old plan in full to demonstrate that management regrets the interruption. - Reassess each commitment using current economics, repair credible promises and document the sources of uncertainty. - Ignore the earlier process and evaluate only this quarter’s financial performance. Debrief: The later evidence warrants review, but it does not make the old investment plan automatically correct. Separate harm from deferred work, changed technology, ordinary turnover and customer-specific issues. Repair commitments whose credibility matters, while explaining why a newly obsolete project should not be revived. Preserve the earlier decision record so the organization can distinguish what was knowable from what changed. A useful revision names both the mistake and the evidence that justifies the new choice, rather than presenting an all-purpose apology or attributing every problem to one event. Sources: - Mark Leonard, Constellation Software 2010 annual president’s letter (2011-05-02), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/shareholder-letter-2010.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2011 annual president’s letter (2012-05-02), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/2011_presidents_letter.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. #### Reasoning rubric - Reconstruct the decision without using the later outcome as proof. - Separate investor liquidity needs from the incentives of employees making long-term commitments. - Distinguish expense reductions, delayed investment and genuine operating improvement. - Specify evidence and a review trigger that could change the recommendation. ### Warren Buffett · Control changes the job Skill: Move from asset valuation to accountable operating control #### Dempster before the intervention · Historical reconstruction Period: April 1962, reconstructed from the January 18, 1963 partnership letter Buffett's partnership controls Dempster Mill, a manufacturer bought at a substantial discount to accounting asset values. Efforts to improve operations with the existing managers have not produced the required change. Inventory and other assets tie up capital while the operating business earns little. You advise the controlling owner before the appointment of a new president. The later sale result is unavailable to you. This is a reconstruction from Buffett's subsequent account, not a contemporaneous board transcript; the former managers' perspectives are absent from the supplied evidence. Decision prompt: What would justify replacing management, and what would you require a new operator to achieve before concluding that the investment was working? Options: - Keep the existing managers and continue discussing the plan. - Set an operating mandate, appoint a capable manager and align compensation with durable results. - Begin an orderly exit after comparing realizable asset value with the cost of continued operation. Debrief: Buffett's January 1963 letter describes appointing Harry Bottle in April 1962. It credits changes in inventory, overhead, marketing and facilities with releasing capital and reducing the break-even point. This supports the claim that execution mattered after the purchase; it does not establish the former managers' motives or every employee consequence. The strong initial answer makes operating milestones explicit and compares them with liquidation economics. A discounted balance sheet supplies a margin for error, but it does not make cash realization automatic. Retain uncertainty about recurring earning power even after cash improves. Sources: - Warren E. Buffett, Ground Rules and 1962 partnership review (January 18, 1963), Reproduced original, printed pages 32-41, especially Dempster on pages 37-39. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Contemporary account of Dempster before and after the April 1962 management change. #### A discounted distributor · Constructed 2026 scenario Period: Constructed 2026 scenario; all business details and figures are fictional You acquire control of a regional industrial distributor in 2026 because its stock and receivables appear valuable relative to the purchase price. After six months, sales are flat, obsolete stock continues growing and managers repeatedly agree to inventory targets without meeting them. A prospective operating leader proposes faster collections and a disposal program. Customers rely on some rarely ordered parts, so reducing every slow-moving item could damage service. You must distinguish useful stock from stranded capital and choose whether the existing team can execute. These are invented facts, not Dempster results. Decision prompt: Design the first operating mandate and compensation terms. What would prevent an apparent cash improvement from disguising damage to the business? Options: - Reward the new leader only for cash released this quarter. - Tie the mandate to cash conversion, service reliability and a documented review of obsolete stock. - Wait for sales growth to absorb excess inventory before changing management. Debrief: Cash released through one-time liquidation differs from improved replenishment and customer service. A defensible mandate distinguishes those sources, protects economically justified spare-parts availability and measures collections without encouraging uneconomic discounts. Compensation should not reward empty shelves while leaving replacement costs to the next period. Replacing a manager can be justified by persistent execution failures, but a new title alone changes nothing. Name the required actions, the time allowed and the evidence that would lead you to choose exit instead. This transfer concerns accountable control, not a prediction about what Buffett would buy today. Sources: - Warren E. Buffett, Ground Rules and 1962 partnership review (January 18, 1963), Reproduced original, printed pages 32-41, especially Dempster on pages 37-39. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Contemporary account of Dempster before and after the April 1962 management change. #### The asset thesis changes · Hypothetical future stress test Period: Hypothetical 2028 follow-up; all developments are fictional Inventory conversion improves, but a major supplier introduces direct fulfillment and customers begin replacing the equipment that requires your distributor's legacy parts. The new manager asks for capital to expand into an unfamiliar product line. The owner wants to preserve jobs and points to the successful cash release as proof that the operator deserves support. There is no reliable demand evidence for the proposed expansion. You must reassess the original asset case without treating the manager's earlier execution as a guarantee of a new market's attractiveness. Decision prompt: Would you reinvest, run down the legacy business or sell? Separate your judgment about the operator from your judgment about the opportunity. Options: - Approve the expansion because the operator has earned trust. - Ring-fence a limited demand test while comparing orderly runoff and sale values. - Liquidate immediately because the original industry assumptions changed. Debrief: Good execution does not make every adjacent investment sound. The changed customer and supplier economics require a new valuation, including runoff costs, service promises and employee effects. A limited test may preserve an option if its cost is justified, while an orderly exit may protect more value than indefinite expansion. Immediate liquidation also needs evidence about realizable prices and remaining customer obligations. Keep the original decision record: buying discounted assets, improving their conversion and funding a new line are distinct decisions with different uncertainties. Sources: - Warren E. Buffett, Ground Rules and 1962 partnership review (January 18, 1963), Reproduced original, printed pages 32-41, especially Dempster on pages 37-39. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Contemporary account of Dempster before and after the April 1962 management change. #### Reasoning rubric - Separate discounted asset value from the ability to release cash. - Use observable operating evidence before attributing motives to managers. - Compare replacement, continued coaching and exit, including costs borne by employees. - State a review trigger and revise it when the economics change. ### Jeff Bezos · Close the product, test the capability Skill: Learn from failure without using it to excuse continued losses #### After Fire Phone · Historical reconstruction Period: Fire Phone failure, examined through the 2018 reporting-year letter published in 2019 Amazon has developed a phone that has not succeeded with customers. The work has created technical knowledge and a team, but those resources do not by themselves establish demand for another device. You are asked to recommend what should happen to the product, people and technology. Do not use the later success of another product to make the answer obvious. The course reconstructs this choice from Bezos's later shareholder account; it does not claim access to Amazon's original internal options, forecasts or meeting records. Decision prompt: What should be closed, what might be retained, and what new evidence would justify spending on a different customer problem? Options: - Continue the phone until its development spending is recovered. - Close the failed proposition where justified and separately test transferable capabilities. - Abandon the product, team and technical work together because the launch failed. Debrief: In the 2018 letter, Bezos says lessons and people from Fire Phone contributed to Echo and Alexa. That is a retrospective account of retained capability, not proof that every failed product earns its cost through a later winner. The 2015 letter frames experimentation as a portfolio with bounded losses and asymmetric potential. A sound answer identifies the specific reusable knowledge, establishes a new customer job and demands new evidence. It also provides for remaining customers and employees. The lesson is selective reuse after an honest failure assessment, not automatic permission to keep spending. Sources: - Jeffrey P. Bezos, 2018 Letter to Shareholders (2018 reporting period; published 2019), Third-party sales, wandering, Fire Phone transfer, wage challenge and employee programs. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2018-Letter-to-Shareholders.pdf Evidence note: Retrospective comparison of first-party and third-party growth, plus management's account of retaining people and capability after the Fire Phone failure. - Jeffrey P. Bezos, 2015 Letter to Shareholders (2015 reporting period; published 2016), Failure, long-tailed business returns, and Type 1 versus Type 2 decisions. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2015-Letter-to-Shareholders.PDF Evidence note: Retrospective management account. The ten-percent and one-hundred-times illustration explains an asymmetric payoff shape, not a measured base rate for Amazon experiments. #### A device without demand · Constructed 2026 scenario Period: Constructed 2026 scenario; all company details are fictional A company launches a voice-enabled household device in 2026. Adoption is weak and retention declines after the initial trial. Its engineers believe the speech and personalization software could help technicians who need hands-free access to maintenance instructions. No technician has yet paid for it. Managers argue that closing the device would waste the development investment and embarrass the team. Existing buyers still need security updates and support. You must distinguish obligations to those customers from discretionary spending on the proposed new business. Decision prompt: Choose a plan for the product and team, and define a test that could show the industrial use is worth pursuing. Options: - Keep subsidizing the household device to avoid admitting failure. - Plan customer support, stop unjustified acquisition spending and run a bounded paid industrial pilot. - Rebrand the same product for industry and treat the engineering work as proof of fit. Debrief: The original investment is sunk, while support promises and avoidable future spending remain relevant. A useful pilot tests technicians' actual workflow, accuracy, willingness to pay and the integration burden. The code may transfer while the sales model, reliability requirements and economics do not. Protect enough team knowledge to run the test without assuming every role must transfer unchanged. Define the evidence needed to proceed and the cost of stopping. Emotional difficulty and embarrassment are facts of the constructed scenario, not claims about Bezos's private reaction to Fire Phone. Sources: - Jeffrey P. Bezos, 2018 Letter to Shareholders (2018 reporting period; published 2019), Third-party sales, wandering, Fire Phone transfer, wage challenge and employee programs. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2018-Letter-to-Shareholders.pdf Evidence note: Retrospective comparison of first-party and third-party growth, plus management's account of retaining people and capability after the Fire Phone failure. - Jeffrey P. Bezos, 2015 Letter to Shareholders (2015 reporting period; published 2016), Failure, long-tailed business returns, and Type 1 versus Type 2 decisions. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2015-Letter-to-Shareholders.PDF Evidence note: Retrospective management account. The ten-percent and one-hundred-times illustration explains an asymmetric payoff shape, not a measured base rate for Amazon experiments. #### A promising customer, a weaker moat · Hypothetical future stress test Period: Hypothetical 2028 follow-up; all developments are fictional A large maintenance company offers a paid pilot for the retained software, but a competitor now bundles similar voice features at no extra charge. Your potential customer needs specialized integrations and expects an extended support commitment. The team interprets the paid offer as validation and proposes a full product launch. You do not yet know whether other customers have the same needs or whether the service margin survives the requested customization. The original household device still has a small installed base with support obligations. Decision prompt: Does this new evidence justify a launch, a narrower business or an exit? Explain exactly which earlier assumption changed. Options: - Launch broadly because one paying customer validates the market. - Price and test the specialized job while measuring customization and support economics. - Exit automatically because a competitor offers the general feature for free. Debrief: A paid pilot is stronger demand evidence than internal enthusiasm, but it may validate a service engagement rather than a repeatable product. The bundled competitor weakens the generic feature's pricing power, while specialized integration may still create value. Separate those economics before scaling. A strong revision identifies the new evidence without erasing the earlier failure or assuming that all learning must become a large business. Include the remaining consumer support costs and any commitments made to the pilot customer in the decision. Sources: - Jeffrey P. Bezos, 2018 Letter to Shareholders (2018 reporting period; published 2019), Third-party sales, wandering, Fire Phone transfer, wage challenge and employee programs. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2018-Letter-to-Shareholders.pdf Evidence note: Retrospective comparison of first-party and third-party growth, plus management's account of retaining people and capability after the Fire Phone failure. - Jeffrey P. Bezos, 2015 Letter to Shareholders (2015 reporting period; published 2016), Failure, long-tailed business returns, and Type 1 versus Type 2 decisions. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2015-Letter-to-Shareholders.PDF Evidence note: Retrospective management account. The ten-percent and one-hundred-times illustration explains an asymmetric payoff shape, not a measured base rate for Amazon experiments. #### Reasoning rubric - Distinguish a failed customer proposition from a transferable capability. - Ignore sunk expenditure when comparing new alternatives. - Bound the next experiment with demand evidence and a stopping rule. - Count employee, customer and maintenance obligations alongside financial costs. ### Howard Marks · When the backdrop stops helping Skill: Update a regime assumption without pretending to forecast precisely #### A different investing environment · Historical reconstruction Period: December 13, 2022, at publication of Sea Change Marks argues that investors may be entering an environment different from the long period of declining or exceptionally low interest rates. Assets and strategies that benefited from cheap financing should no longer be evaluated as if that support were permanent. Credit may offer a more attractive starting point than it did under easier conditions. You advise an investment committee that has performed well using leverage and rising asset valuations. The persistence of the new regime is uncertain, and you cannot use subsequent rates or returns to resolve the decision. The committee and options are course constructions built from the memo’s conditions, not a record of an Oaktree meeting or portfolio. Decision prompt: What should the committee change now, and which changes would require stronger evidence about the future? Options: - Keep the allocation unchanged because the old approach produced strong returns. - Re-underwrite financing and expected returns, then adjust exposure to the opportunities available. - Replace the portfolio with one concentrated bet on persistently high rates. Debrief: Sea Change, dated December 2022, sets out Marks's argument for a changed backdrop and a potentially improved role for credit. Easy Money, dated January 2024, develops the analysis further; it is later evidence, not something the learner knew in 2022. Neither memo removes uncertainty about defaults, entry prices or future rates. The strongest response tests whether prior performance depended on falling financing costs, marks exposures at current economics and avoids treating one macro interpretation as certainty. A regime update can warrant changed underwriting without an all-or-nothing portfolio reversal. Sources: - Howard Marks, Sea Change (2022-12-13), Complete memo, sections “Sea Change #1,” “Sea Change #2,” and the closing outlook and 2-to-4-percent range. https://www.oaktreecapital.com/insights/memo/sea-change Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Easy Money (2024-01-09), Printed p. 1, ten years without reader response; p. 14, 3.0-to-3.5-percent guess over the next five to ten years. https://www.oaktreecapital.com/insights/memo/easy-money Evidence note: Read directly in the official Oaktree complete collection or official memo page. #### A strong record with weak refinancing · Constructed 2026 scenario Period: Constructed 2026 scenario; all figures and investment details are fictional An investment group has earned attractive historical returns buying businesses with floating-rate debt and selling them at higher valuation multiples. In this fictional 2026 case, an existing loan costs 5% and the available refinancing would cost 9%. Management's base case assumes the old financing cost returns soon. A new senior loan opportunity offers contractual income, but its borrower has volatile cash flow and weak covenants. The committee calls the loan safer simply because it is senior. Your task is to compare cash-flow resilience and price without relying on a precise rate forecast. Decision prompt: Which assumptions must change before you can compare the existing holdings with the new loan? Options: - Preserve the old assumptions until the central bank proves the new rate level permanent. - Model refinancing, cash coverage, downside recovery and terms under several plausible rate paths. - Switch into the senior loan because contractual payments eliminate uncertainty. Debrief: The 5% and 9% rates are invented scenario inputs, not a claim about actual 2026 markets or Marks's forecasts. They reveal sensitivity to refinancing and the danger of attributing every past gain to skill. Seniority affects loss distribution but cannot create cash that the borrower lacks. Compare prices, covenants, coverage, refinancing timing and recovery under adverse conditions. A defensible adjustment can reduce leverage or demand a better price without claiming to know future policy rates. Preserve the distinction between an attractive quoted yield and the return actually realized after losses. Sources: - Howard Marks, Sea Change (2022-12-13), Complete memo, sections “Sea Change #1,” “Sea Change #2,” and the closing outlook and 2-to-4-percent range. https://www.oaktreecapital.com/insights/memo/sea-change Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Easy Money (2024-01-09), Printed p. 1, ten years without reader response; p. 14, 3.0-to-3.5-percent guess over the next five to ten years. https://www.oaktreecapital.com/insights/memo/easy-money Evidence note: Read directly in the official Oaktree complete collection or official memo page. #### Rates fall, but losses rise · Hypothetical future stress test Period: Hypothetical 2028 follow-up; all developments are fictional Rates fall sharply in the constructed follow-up, but demand weakens and several borrowers breach covenants. The committee says the earlier regime analysis was wrong and wants to restore the old leverage policy. Some fixed-rate assets have appreciated, while other positions have suffered operating losses. The new credit portfolio's contractual yields were attractive, but realized results differ across borrowers. You must decide which part of the earlier thesis has been contradicted and whether cheaper financing is enough to justify returning to the former strategy. Decision prompt: What do you revise, and what remains valid despite the fall in rates? Options: - Restore the former leverage policy because rates fell. - Separate the rate forecast, credit underwriting and portfolio resilience before changing policy. - Keep every position unchanged to demonstrate consistency with the earlier thesis. Debrief: A rate forecast can be wrong while disciplined refinancing and loss analysis remain useful. Lower rates may ease debt service while weaker cash flows increase default risk. Attribute outcomes separately to duration, borrower selection, purchase price, leverage and economic conditions. A useful revision changes the assumptions contradicted by evidence, preserves safeguards that still earn their cost and admits forecast error where it occurred. Neither a profitable mark nor a loss by itself proves the entire method sound or unsound. This is a hypothetical stress test, not a forecast attributed to Marks. Sources: - Howard Marks, Sea Change (2022-12-13), Complete memo, sections “Sea Change #1,” “Sea Change #2,” and the closing outlook and 2-to-4-percent range. https://www.oaktreecapital.com/insights/memo/sea-change Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Easy Money (2024-01-09), Printed p. 1, ten years without reader response; p. 14, 3.0-to-3.5-percent guess over the next five to ten years. https://www.oaktreecapital.com/insights/memo/easy-money Evidence note: Read directly in the official Oaktree complete collection or official memo page. #### Reasoning rubric - Separate observed conditions from a forecast of their persistence. - Test how the old interest-rate environment contributed to apparent skill. - Compare opportunities using price, loss risk and financing terms. - Revise exposure proportionately when evidence changes, without all-or-nothing macro bets. ## Recall cards ### What does a shareholder letter establish most directly? What its author publicly said and reported at the stated time. It does not independently establish that a favored practice caused all subsequent success. Sources: - Warren E. Buffett, Berkshire Hathaway 2012 shareholder letter (2012 reporting year; published 2013), Dividends, including reinvestment, acquisitions, repurchases, and the conditions behind the payout argument.. https://www.berkshirehathaway.com/letters/2012ltr.pdf Evidence note: Written by the allocator. His assessments of acquisition value and alternative outcomes are his judgments, not independent causal measurements. - Jeffrey P. Bezos, Amazon 2016 shareholder letter (2016 reporting year; published 2017), Resist Proxies and High-Velocity Decision Making.. https://www.aboutamazon.com/news/company-news/2016-letter-to-shareholders Evidence note: Use the signed Bezos letter, not the hosting page’s staff byline. Distinguish his reported practices from proof that they caused Amazon’s results. - Howard Marks, Fewer Losers, or More Winners? (September 12, 2023), Putting It in Brief; Not Risk Avoidance; The Role of Risk Bearing; What About in Practice?. https://www.oaktreecapital.com/insights/memo/fewer-losers-or-more-winners Evidence note: A conceptual and retrospective memo. It distinguishes bond upside, strategies needing winners, and risk control. It does not establish that a reader can identify investment bargains. - Mark Leonard, Constellation Software 2014 president’s letter (2014 reporting year; published 2015), Pages 1–5, particularly intangible-asset condition, reliable capital, and business-unit decentralization.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: The URL reflects later file hosting, not a 2026 letter. Business figures in the original refer to their stated historical periods. Author explanations remain author explanations. ### Why decide before revealing the historical outcome? To separate the information available at the decision from hindsight and to assess the original reasoning fairly. Sources: - Warren E. Buffett, Berkshire Hathaway 2012 shareholder letter (2012 reporting year; published 2013), Dividends, including reinvestment, acquisitions, repurchases, and the conditions behind the payout argument.. https://www.berkshirehathaway.com/letters/2012ltr.pdf Evidence note: Written by the allocator. His assessments of acquisition value and alternative outcomes are his judgments, not independent causal measurements. - Jeffrey P. Bezos, Amazon 2016 shareholder letter (2016 reporting year; published 2017), Resist Proxies and High-Velocity Decision Making.. https://www.aboutamazon.com/news/company-news/2016-letter-to-shareholders Evidence note: Use the signed Bezos letter, not the hosting page’s staff byline. Distinguish his reported practices from proof that they caused Amazon’s results. - Howard Marks, Fewer Losers, or More Winners? (September 12, 2023), Putting It in Brief; Not Risk Avoidance; The Role of Risk Bearing; What About in Practice?. https://www.oaktreecapital.com/insights/memo/fewer-losers-or-more-winners Evidence note: A conceptual and retrospective memo. It distinguishes bond upside, strategies needing winners, and risk control. It does not establish that a reader can identify investment bargains. - Mark Leonard, Constellation Software 2014 president’s letter (2014 reporting year; published 2015), Pages 1–5, particularly intangible-asset condition, reliable capital, and business-unit decentralization.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: The URL reflects later file hosting, not a 2026 letter. Business figures in the original refer to their stated historical periods. Author explanations remain author explanations. ### What are the three judgments in reading a business thinker? Reconstruct the argument, evaluate its evidence in the case, and test whether its mechanism transfers to another setting. Sources: - Warren E. Buffett, Berkshire Hathaway 2012 shareholder letter (2012 reporting year; published 2013), Dividends, including reinvestment, acquisitions, repurchases, and the conditions behind the payout argument.. https://www.berkshirehathaway.com/letters/2012ltr.pdf Evidence note: Written by the allocator. His assessments of acquisition value and alternative outcomes are his judgments, not independent causal measurements. - Jeffrey P. Bezos, Amazon 2016 shareholder letter (2016 reporting year; published 2017), Resist Proxies and High-Velocity Decision Making.. https://www.aboutamazon.com/news/company-news/2016-letter-to-shareholders Evidence note: Use the signed Bezos letter, not the hosting page’s staff byline. Distinguish his reported practices from proof that they caused Amazon’s results. - Howard Marks, Fewer Losers, or More Winners? (September 12, 2023), Putting It in Brief; Not Risk Avoidance; The Role of Risk Bearing; What About in Practice?. https://www.oaktreecapital.com/insights/memo/fewer-losers-or-more-winners Evidence note: A conceptual and retrospective memo. It distinguishes bond upside, strategies needing winners, and risk control. It does not establish that a reader can identify investment bargains. - Mark Leonard, Constellation Software 2014 president’s letter (2014 reporting year; published 2015), Pages 1–5, particularly intangible-asset condition, reliable capital, and business-unit decentralization.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: The URL reflects later file hosting, not a 2026 letter. Business figures in the original refer to their stated historical periods. Author explanations remain author explanations. ### What makes a revision more useful than a changed answer? It identifies the failed assumption, the new evidence, and the decision procedure that must change while preserving the initial rationale. Sources: - Jeffrey P. Bezos, Amazon 2016 shareholder letter (2016 reporting year; published 2017), Resist Proxies and High-Velocity Decision Making.. https://www.aboutamazon.com/news/company-news/2016-letter-to-shareholders Evidence note: Use the signed Bezos letter, not the hosting page’s staff byline. Distinguish his reported practices from proof that they caused Amazon’s results. - Warren E. Buffett, Berkshire Hathaway 2012 shareholder letter (2012 reporting year; published 2013), Dividends, including reinvestment, acquisitions, repurchases, and the conditions behind the payout argument.. https://www.berkshirehathaway.com/letters/2012ltr.pdf Evidence note: Written by the allocator. His assessments of acquisition value and alternative outcomes are his judgments, not independent causal measurements. ### When is a decision record too vague? When it names goals without a feasible action, alternatives, owner, binding constraint, causal chain, or observable review condition. Sources: - Jeffrey P. Bezos, Amazon 2016 shareholder letter (2016 reporting year; published 2017), Resist Proxies and High-Velocity Decision Making.. https://www.aboutamazon.com/news/company-news/2016-letter-to-shareholders Evidence note: Use the signed Bezos letter, not the hosting page’s staff byline. Distinguish his reported practices from proof that they caused Amazon’s results. - Warren E. Buffett, Berkshire Hathaway 2012 shareholder letter (2012 reporting year; published 2013), Dividends, including reinvestment, acquisitions, repurchases, and the conditions behind the payout argument.. https://www.berkshirehathaway.com/letters/2012ltr.pdf Evidence note: Written by the allocator. His assessments of acquisition value and alternative outcomes are his judgments, not independent causal measurements. ### Why can a high historical return fail to justify new investment? The next commitment may have a different price, incremental return, sustaining cost, or competitive environment. Sources: - Warren E. Buffett, Berkshire Hathaway 2012 shareholder letter (2012 reporting year; published 2013), Dividends, including reinvestment, acquisitions, repurchases, and the conditions behind the payout argument.. https://www.berkshirehathaway.com/letters/2012ltr.pdf Evidence note: Written by the allocator. His assessments of acquisition value and alternative outcomes are his judgments, not independent causal measurements. - Mark Leonard, Constellation Software 2014 president’s letter (2014 reporting year; published 2015), Pages 1–5, particularly intangible-asset condition, reliable capital, and business-unit decentralization.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: The URL reflects later file hosting, not a 2026 letter. Business figures in the original refer to their stated historical periods. Author explanations remain author explanations. ### Why is an undiscounted cash surplus not an annual return? It aggregates receipts and outlays across a stated horizon without accounting for timing. An annual return must describe the rate over time. Sources: - Warren E. Buffett, Berkshire Hathaway 2012 shareholder letter (2012 reporting year; published 2013), Dividends, including reinvestment, acquisitions, repurchases, and the conditions behind the payout argument.. https://www.berkshirehathaway.com/letters/2012ltr.pdf Evidence note: Written by the allocator. His assessments of acquisition value and alternative outcomes are his judgments, not independent causal measurements. - Mark Leonard, Constellation Software 2014 president’s letter (2014 reporting year; published 2015), Pages 1–5, particularly intangible-asset condition, reliable capital, and business-unit decentralization.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: The URL reflects later file hosting, not a 2026 letter. Business figures in the original refer to their stated historical periods. Author explanations remain author explanations. ### What can correct net present value arithmetic fail to establish? Whether the cash-flow assumptions, discount rate, horizon, omitted commitments, and residual value are economically justified. Sources: - Warren E. Buffett, Berkshire Hathaway 2012 shareholder letter (2012 reporting year; published 2013), Dividends, including reinvestment, acquisitions, repurchases, and the conditions behind the payout argument.. https://www.berkshirehathaway.com/letters/2012ltr.pdf Evidence note: Written by the allocator. His assessments of acquisition value and alternative outcomes are his judgments, not independent causal measurements. - Mark Leonard, Constellation Software 2014 president’s letter (2014 reporting year; published 2015), Pages 1–5, particularly intangible-asset condition, reliable capital, and business-unit decentralization.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: The URL reflects later file hosting, not a 2026 letter. Business figures in the original refer to their stated historical periods. Author explanations remain author explanations. ### Why might a conventional lender and an experiment sponsor use different loss tests? The lender’s contractual upside is limited; an experiment may have open-ended upside. Both still need appropriate loss control and survival capacity. Sources: - Howard Marks, Fewer Losers, or More Winners? (September 12, 2023), Putting It in Brief; Not Risk Avoidance; The Role of Risk Bearing; What About in Practice?. https://www.oaktreecapital.com/insights/memo/fewer-losers-or-more-winners Evidence note: A conceptual and retrospective memo. It distinguishes bond upside, strategies needing winners, and risk control. It does not establish that a reader can identify investment bargains. - Jeffrey P. Bezos, Amazon 2016 shareholder letter (2016 reporting year; published 2017), Resist Proxies and High-Velocity Decision Making.. https://www.aboutamazon.com/news/company-news/2016-letter-to-shareholders Evidence note: Use the signed Bezos letter, not the hosting page’s staff byline. Distinguish his reported practices from proof that they caused Amazon’s results. ### Why do many experiments not necessarily diversify exposure? They can depend on the same customer behavior, supplier, technology, financing source, or distribution channel. Sources: - Howard Marks, Fewer Losers, or More Winners? (September 12, 2023), Putting It in Brief; Not Risk Avoidance; The Role of Risk Bearing; What About in Practice?. https://www.oaktreecapital.com/insights/memo/fewer-losers-or-more-winners Evidence note: A conceptual and retrospective memo. It distinguishes bond upside, strategies needing winners, and risk control. It does not establish that a reader can identify investment bargains. - Jeffrey P. Bezos, Amazon 2016 shareholder letter (2016 reporting year; published 2017), Resist Proxies and High-Velocity Decision Making.. https://www.aboutamazon.com/news/company-news/2016-letter-to-shareholders Evidence note: Use the signed Bezos letter, not the hosting page’s staff byline. Distinguish his reported practices from proof that they caused Amazon’s results. ### What four things make delegation reviewable? The decision owner, the boundary of authority, the required evidence, and the conditions for escalation. Sources: - Mark Leonard, Constellation Software 2014 president’s letter (2014 reporting year; published 2015), Pages 1–5, particularly intangible-asset condition, reliable capital, and business-unit decentralization.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: The URL reflects later file hosting, not a 2026 letter. Business figures in the original refer to their stated historical periods. Author explanations remain author explanations. - Jeffrey P. Bezos, Amazon 2016 shareholder letter (2016 reporting year; published 2017), Resist Proxies and High-Velocity Decision Making.. https://www.aboutamazon.com/news/company-news/2016-letter-to-shareholders Evidence note: Use the signed Bezos letter, not the hosting page’s staff byline. Distinguish his reported practices from proof that they caused Amazon’s results. ### Why does a higher-margin unit not automatically identify a better practice? Margin can reflect customer mix, inherited conditions, accounting differences, or deferred work rather than a transferable operating improvement. Sources: - Mark Leonard, Constellation Software 2014 president’s letter (2014 reporting year; published 2015), Pages 1–5, particularly intangible-asset condition, reliable capital, and business-unit decentralization.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: The URL reflects later file hosting, not a 2026 letter. Business figures in the original refer to their stated historical periods. Author explanations remain author explanations. - Jeffrey P. Bezos, Amazon 2016 shareholder letter (2016 reporting year; published 2017), Resist Proxies and High-Velocity Decision Making.. https://www.aboutamazon.com/news/company-news/2016-letter-to-shareholders Evidence note: Use the signed Bezos letter, not the hosting page’s staff byline. Distinguish his reported practices from proof that they caused Amazon’s results. ### What makes a modern adaptation substantive? It changes a mechanism, constraint, or information condition that could alter the choice, rather than merely changing the date or company name. Sources: - Jeffrey P. Bezos, Amazon 2016 shareholder letter (2016 reporting year; published 2017), Resist Proxies and High-Velocity Decision Making.. https://www.aboutamazon.com/news/company-news/2016-letter-to-shareholders Evidence note: Use the signed Bezos letter, not the hosting page’s staff byline. Distinguish his reported practices from proof that they caused Amazon’s results. - Mark Leonard, Constellation Software 2014 president’s letter (2014 reporting year; published 2015), Pages 1–5, particularly intangible-asset condition, reliable capital, and business-unit decentralization.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: The URL reflects later file hosting, not a 2026 letter. Business figures in the original refer to their stated historical periods. Author explanations remain author explanations. - Howard Marks, Fewer Losers, or More Winners? (September 12, 2023), Putting It in Brief; Not Risk Avoidance; The Role of Risk Bearing; What About in Practice?. https://www.oaktreecapital.com/insights/memo/fewer-losers-or-more-winners Evidence note: A conceptual and retrospective memo. It distinguishes bond upside, strategies needing winners, and risk control. It does not establish that a reader can identify investment bargains. ### Does cheaper software development automatically remove switching costs? No. Identify whether migration, training, downtime, integration, contracts, or replacement uncertainty has actually changed. Sources: - Jeffrey P. Bezos, Amazon 2016 shareholder letter (2016 reporting year; published 2017), Resist Proxies and High-Velocity Decision Making.. https://www.aboutamazon.com/news/company-news/2016-letter-to-shareholders Evidence note: Use the signed Bezos letter, not the hosting page’s staff byline. Distinguish his reported practices from proof that they caused Amazon’s results. - Mark Leonard, Constellation Software 2014 president’s letter (2014 reporting year; published 2015), Pages 1–5, particularly intangible-asset condition, reliable capital, and business-unit decentralization.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: The URL reflects later file hosting, not a 2026 letter. Business figures in the original refer to their stated historical periods. Author explanations remain author explanations. - Howard Marks, Fewer Losers, or More Winners? (September 12, 2023), Putting It in Brief; Not Risk Avoidance; The Role of Risk Bearing; What About in Practice?. https://www.oaktreecapital.com/insights/memo/fewer-losers-or-more-winners Evidence note: A conceptual and retrospective memo. It distinguishes bond upside, strategies needing winners, and risk control. It does not establish that a reader can identify investment bargains. ### What is the first financing test in the fictional capstone? Compare immediate cash commitments with discretionary cash after non-overlapping reserves and stress needs. Do not treat uncommitted financing as available. Sources: - Warren E. Buffett, Berkshire Hathaway 2012 shareholder letter (2012 reporting year; published 2013), Dividends, including reinvestment, acquisitions, repurchases, and the conditions behind the payout argument.. https://www.berkshirehathaway.com/letters/2012ltr.pdf Evidence note: Written by the allocator. His assessments of acquisition value and alternative outcomes are his judgments, not independent causal measurements. - Howard Marks, Fewer Losers, or More Winners? (September 12, 2023), Putting It in Brief; Not Risk Avoidance; The Role of Risk Bearing; What About in Practice?. https://www.oaktreecapital.com/insights/memo/fewer-losers-or-more-winners Evidence note: A conceptual and retrospective memo. It distinguishes bond upside, strategies needing winners, and risk control. It does not establish that a reader can identify investment bargains. - Mark Leonard, Constellation Software 2014 president’s letter (2014 reporting year; published 2015), Pages 1–5, particularly intangible-asset condition, reliable capital, and business-unit decentralization.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: The URL reflects later file hosting, not a 2026 letter. Business figures in the original refer to their stated historical periods. Author explanations remain author explanations. - Jeffrey P. Bezos, Amazon 2016 shareholder letter (2016 reporting year; published 2017), Resist Proxies and High-Velocity Decision Making.. https://www.aboutamazon.com/news/company-news/2016-letter-to-shareholders Evidence note: Use the signed Bezos letter, not the hosting page’s staff byline. Distinguish his reported practices from proof that they caused Amazon’s results. ### What does a high score on the guide rubric establish? A self-assessment against stated reasoning criteria. It does not establish an externally measured expertise percentile or predictive investing skill. Sources: - Warren E. Buffett, Berkshire Hathaway 2012 shareholder letter (2012 reporting year; published 2013), Dividends, including reinvestment, acquisitions, repurchases, and the conditions behind the payout argument.. https://www.berkshirehathaway.com/letters/2012ltr.pdf Evidence note: Written by the allocator. His assessments of acquisition value and alternative outcomes are his judgments, not independent causal measurements. - Howard Marks, Fewer Losers, or More Winners? (September 12, 2023), Putting It in Brief; Not Risk Avoidance; The Role of Risk Bearing; What About in Practice?. https://www.oaktreecapital.com/insights/memo/fewer-losers-or-more-winners Evidence note: A conceptual and retrospective memo. It distinguishes bond upside, strategies needing winners, and risk control. It does not establish that a reader can identify investment bargains. - Mark Leonard, Constellation Software 2014 president’s letter (2014 reporting year; published 2015), Pages 1–5, particularly intangible-asset condition, reliable capital, and business-unit decentralization.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: The URL reflects later file hosting, not a 2026 letter. Business figures in the original refer to their stated historical periods. Author explanations remain author explanations. - Jeffrey P. Bezos, Amazon 2016 shareholder letter (2016 reporting year; published 2017), Resist Proxies and High-Velocity Decision Making.. https://www.aboutamazon.com/news/company-news/2016-letter-to-shareholders Evidence note: Use the signed Bezos letter, not the hosting page’s staff byline. Distinguish his reported practices from proof that they caused Amazon’s results. ### What does Buffett's owner-earnings bridge subtract after adding back relevant non-cash charges? Average capital spending and working capital required to maintain long-term competitive position and unit volume. Sources: - Warren E. Buffett, 1986 Chairman's Letter (1986 reporting period; published 1987), Purchase-Price Accounting Adjustments and the owner-earnings discussion near the end of the letter. https://www.berkshirehathaway.com/letters/1986.html Evidence note: Explains Buffett's valuation adjustment using Scott Fetzer. Maintenance capital expenditure is explicitly an estimate, not an audited line item. ### Why is owner earnings not a mechanically precise GAAP figure? Maintenance capital and required working capital must be estimated, and the estimate depends on operating and competitive facts. Sources: - Warren E. Buffett, 1986 Chairman's Letter (1986 reporting period; published 1987), Purchase-Price Accounting Adjustments and the owner-earnings discussion near the end of the letter. https://www.berkshirehathaway.com/letters/1986.html Evidence note: Explains Buffett's valuation adjustment using Scott Fetzer. Maintenance capital expenditure is explicitly an estimate, not an audited line item. ### When can adding back all depreciation overstate distributable cash? When assets require equal or greater reinvestment over time to preserve capacity and competitive position. Sources: - Warren E. Buffett, 1986 Chairman's Letter (1986 reporting period; published 1987), Purchase-Price Accounting Adjustments and the owner-earnings discussion near the end of the letter. https://www.berkshirehathaway.com/letters/1986.html Evidence note: Explains Buffett's valuation adjustment using Scott Fetzer. Maintenance capital expenditure is explicitly an estimate, not an audited line item. ### What limited role does the section assign accounting in valuation? Accounting records are the starting language. Owners must evaluate the economic meaning of the figures. Sources: - Warren E. Buffett, 1986 Chairman's Letter (1986 reporting period; published 1987), Purchase-Price Accounting Adjustments and the owner-earnings discussion near the end of the letter. https://www.berkshirehathaway.com/letters/1986.html Evidence note: Explains Buffett's valuation adjustment using Scott Fetzer. Maintenance capital expenditure is explicitly an estimate, not an audited line item. ### Why did individually rational textile upgrades fail to improve industry returns? Competitors made similar upgrades, cost savings became lower market prices, and everyone had more capital committed at weak returns. Sources: - Warren E. Buffett, 1985 Chairman's Letter (March 4, 1986), Shutdown of Textile Business. https://www.berkshirehathaway.com/letters/1985.html Evidence note: Contemporary account written after the July 1985 closure decision. It reports Buffett's own diagnosis and hindsight about earlier delay. ### Which continuation condition did Buffett say proved wrong? The expectation that textiles would average modest cash returns relative to investment. Sources: - Warren E. Buffett, 1985 Chairman's Letter (March 4, 1986), Shutdown of Textile Business. https://www.berkshirehathaway.com/letters/1985.html Evidence note: Contemporary account written after the July 1985 closure decision. It reports Buffett's own diagnosis and hindsight about earlier delay. ### What did the textile machinery auction show about book and replacement values? Neither guaranteed realizable value when the assets could not earn adequate cash in their competitive setting. Sources: - Warren E. Buffett, 1985 Chairman's Letter (March 4, 1986), Shutdown of Textile Business. https://www.berkshirehathaway.com/letters/1985.html Evidence note: Contemporary account written after the July 1985 closure decision. It reports Buffett's own diagnosis and hindsight about earlier delay. ### Why is the textile case not a rule to close every low-return business immediately? Buffett weighed employment, candid management, cooperative labour and bounded cash support, but rejected an apparently endless loss claim. Sources: - Warren E. Buffett, 1985 Chairman's Letter (March 4, 1986), Shutdown of Textile Business. https://www.berkshirehathaway.com/letters/1985.html Evidence note: Contemporary account written after the July 1985 closure decision. It reports Buffett's own diagnosis and hindsight about earlier delay. ### What created economic goodwill in the See's example? An intangible customer franchise that supported returns far above ordinary rates on the tangible capital required. Sources: - Warren E. Buffett, 1983 Chairman's Letter (1983 reporting period; published 1984), Goodwill and its Amortization, especially the See's comparison. https://www.berkshirehathaway.com/letters/1983.html Evidence note: Contemporary discussion of accounting and economic goodwill using historical See's figures. The acquisition rationale is partly retrospective. ### Why can an asset-light franchise handle inflation better than an asset-heavy peer in Buffett's example? It needs fewer additional nominal dollars in tangible assets merely to preserve the same real unit volume and earnings. Sources: - Warren E. Buffett, 1983 Chairman's Letter (1983 reporting period; published 1984), Goodwill and its Amortization, especially the See's comparison. https://www.berkshirehathaway.com/letters/1983.html Evidence note: Contemporary discussion of accounting and economic goodwill using historical See's figures. The acquisition rationale is partly retrospective. ### Why is a famous brand not automatically See's-like? The brand must support customer choice or price and endure without absorbing most earnings in promotion, assets or renewal. Sources: - Warren E. Buffett, 1983 Chairman's Letter (1983 reporting period; published 1984), Goodwill and its Amortization, especially the See's comparison. https://www.berkshirehathaway.com/letters/1983.html Evidence note: Contemporary discussion of accounting and economic goodwill using historical See's figures. The acquisition rationale is partly retrospective. - Warren E. Buffett, 2007 Chairman's Letter (2007 reporting period; published 2008), Businesses, The Great, the Good and the Gruesome. https://www.berkshirehathaway.com/letters/2007ltr.pdf Evidence note: Retrospective comparison of See's, FlightSafety, capital-intensive growth, and Dexter. The Dexter share value is measured at the letter's February 2008 publication, not in the 2007 reporting year or in 1993. ### What made See's useful to Berkshire beyond its own earnings growth? It required relatively little incremental capital, so substantial cash could be distributed and invested in other businesses. Sources: - Warren E. Buffett, 2007 Chairman's Letter (2007 reporting period; published 2008), Businesses, The Great, the Good and the Gruesome. https://www.berkshirehathaway.com/letters/2007ltr.pdf Evidence note: Retrospective comparison of See's, FlightSafety, capital-intensive growth, and Dexter. The Dexter share value is measured at the letter's February 2008 publication, not in the 2007 reporting year or in 1993. - Warren E. Buffett, 2014 Chairman's Letter (2014 reporting period; published 2015), Berkshire, Past, Present and Future, especially the textile, See's and Dexter passages. https://www.berkshirehathaway.com/letters/2014ltr.pdf Evidence note: Fifty-year retrospective. It supplies later cumulative figures and Buffett's revised interpretation, not a contemporaneous record of the earlier decisions. The Dexter share value is measured at the letter's February 2015 publication. ### What is the economic cost of shares issued in an acquisition? The intrinsic value of the ownership claim surrendered, including its percentage of future business value. Sources: - Warren E. Buffett, 2014 Chairman's Letter (2014 reporting period; published 2015), Berkshire, Past, Present and Future, especially the textile, See's and Dexter passages. https://www.berkshirehathaway.com/letters/2014ltr.pdf Evidence note: Fifty-year retrospective. It supplies later cumulative figures and Buffett's revised interpretation, not a contemporaneous record of the earlier decisions. The Dexter share value is measured at the letter's February 2015 publication. ### Name two facts that supported the favorable 1993 Dexter case. Examples include 7.5 million pairs produced, 77 outlets, supplier awards, a golf-shoe position and strong results from Berkshire's earlier shoe acquisitions. Sources: - Warren E. Buffett, 1993 Chairman's Letter (1993 reporting period; published 1994), Dexter Shoe. https://www.berkshirehathaway.com/letters/1993.html Evidence note: Contemporary, favorable account after the November 7, 1993 merger. It is useful evidence of the case made then, not proof that all deliberations were disclosed. ### Which three Dexter errors did Buffett separate in 2001? Buying the company, paying with Berkshire stock, and delaying operational changes after the need became apparent. Sources: - Warren E. Buffett, 2001 Chairman's Letter (2001 reporting period; published 2002), Non-insurance operating highlights, shoe operations and Dexter. https://www.berkshirehathaway.com/2001ar/2001letter.html Evidence note: Retrospective self-assessment identifying the purchase, use of stock, and delayed operational response as three separate errors. ### Why are the Dexter values reported in the 2007 and 2014 letters not 1993 facts? They use Berkshire share values at the letters' 2008 and 2015 publication dates to measure opportunity cost. Those outcomes were unknown at acquisition. Sources: - Warren E. Buffett, 2007 Chairman's Letter (2007 reporting period; published 2008), Businesses, The Great, the Good and the Gruesome. https://www.berkshirehathaway.com/letters/2007ltr.pdf Evidence note: Retrospective comparison of See's, FlightSafety, capital-intensive growth, and Dexter. The Dexter share value is measured at the letter's February 2008 publication, not in the 2007 reporting year or in 1993. - Warren E. Buffett, 2014 Chairman's Letter (2014 reporting period; published 2015), Berkshire, Past, Present and Future, especially the textile, See's and Dexter passages. https://www.berkshirehathaway.com/letters/2014ltr.pdf Evidence note: Fifty-year retrospective. It supplies later cumulative figures and Buffett's revised interpretation, not a contemporaneous record of the earlier decisions. The Dexter share value is measured at the letter's February 2015 publication. ### Why test reinvestment at the margin rather than rely only on historical return on capital? The decision concerns the next dollar, whose projects, timing, competitive response and return may differ from the old asset base. Sources: - Warren E. Buffett, 2007 Chairman's Letter (2007 reporting period; published 2008), Businesses, The Great, the Good and the Gruesome. https://www.berkshirehathaway.com/letters/2007ltr.pdf Evidence note: Retrospective comparison of See's, FlightSafety, capital-intensive growth, and Dexter. The Dexter share value is measured at the letter's February 2008 publication, not in the 2007 reporting year or in 1993. ### When can growth make a business worse for owners? When each increment demands substantial capital and produces an inadequate durable cash return. Sources: - Warren E. Buffett, 2007 Chairman's Letter (2007 reporting period; published 2008), Businesses, The Great, the Good and the Gruesome. https://www.berkshirehathaway.com/letters/2007ltr.pdf Evidence note: Retrospective comparison of See's, FlightSafety, capital-intensive growth, and Dexter. The Dexter share value is measured at the letter's February 2008 publication, not in the 2007 reporting year or in 1993. ### Which two disciplines kept Amazon's 1997 long-term policy from being only a call for growth? Program-level analytical review and cost-conscious capital management, including stopping weak programs. Sources: - Jeffrey P. Bezos, 1997 Letter to Shareholders (1997 reporting period; published 1998), It's All About the Long Term, Obsess Over Customers, Infrastructure, and Goals for 1998. https://www.aboutamazon.com/news/company-news/amazons-original-1997-letter-to-shareholders Evidence note: Contemporary statement of management policy and 1997 results. Management's claims about market leadership and future economics are its judgments at the time. ### What customer measure reported for the fourth quarter of 1997 supported the case that growth included repeat use? Orders from repeat customers exceeded 58 percent, up from more than 46 percent in the fourth quarter of 1996. Sources: - Jeffrey P. Bezos, 1997 Letter to Shareholders (1997 reporting period; published 1998), It's All About the Long Term, Obsess Over Customers, Infrastructure, and Goals for 1998. https://www.aboutamazon.com/news/company-news/amazons-original-1997-letter-to-shareholders Evidence note: Contemporary statement of management policy and 1997 results. Management's claims about market leadership and future economics are its judgments at the time. ### What must a long-term policy name in addition to its ambition? A causal hypothesis, measures, review dates, stop or scale conditions, and a liquidity boundary. Sources: - Jeffrey P. Bezos, 1997 Letter to Shareholders (1997 reporting period; published 1998), It's All About the Long Term, Obsess Over Customers, Infrastructure, and Goals for 1998. https://www.aboutamazon.com/news/company-news/amazons-original-1997-letter-to-shareholders Evidence note: Contemporary statement of management policy and 1997 results. Management's claims about market leadership and future economics are its judgments at the time. ### Why do Amazon's 1998 results not prove the 1997 policy caused them? They are a later company outcome without a counterfactual and do not isolate returns from each investment program. Sources: - Jeffrey P. Bezos, 1998 Letter to Shareholders (1998 reporting period; published 1999), A Recap of 1998, Our Customers, and Goals for 1999. https://s2.q4cdn.com/299287126/files/doc_financials/annual/Shareholderletter98.pdf Evidence note: Contemporary follow-up reporting customer, revenue, repeat-order, cash-flow, inventory and infrastructure figures. The outcome does not isolate the effect of the 1997 policy. ### State the operating loop described in Amazon's 2001 letter. Lower cost can fund lower prices, customer value can increase demand, volume can spread fixed costs, and the resulting efficiency can fund further value. Sources: - Jeffrey P. Bezos, 2001 Letter to Shareholders (2001 reporting period; published 2002), Opening discussion of price, cost reduction, growth, fixed costs and cash flow. https://ir.aboutamazon.com/files/doc_financials/annual/2001_shareholderLetter.pdf Evidence note: Management's contemporary explanation of a reinforcing operating cycle after reporting its first profitable quarter on the letter's pro forma measures. ### What should every arrow in a reinforcing-cycle map carry? A measure, an expected lag and a failure condition. Sources: - Jeffrey P. Bezos, 2001 Letter to Shareholders (2001 reporting period; published 2002), Opening discussion of price, cost reduction, growth, fixed costs and cash flow. https://ir.aboutamazon.com/files/doc_financials/annual/2001_shareholderLetter.pdf Evidence note: Management's contemporary explanation of a reinforcing operating cycle after reporting its first profitable quarter on the letter's pro forma measures. - Jeffrey P. Bezos, 2002 Letter to Shareholders (2002 reporting period; published 2003), Opening discussion of customer experience, price, satisfaction, the 100-book comparison, and free cash flow. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2002_shareholderLetter.pdf Evidence note: Contemporary company evidence. The 100-book basket, satisfaction score and management's free-cash-flow definition have specific scopes and do not prove a universal price rule. ### Why can the 2002 letter's 100-book comparison not establish Amazon's general price position? It covered the other retailer's published 2002 bestseller basket in named locations. The letter gives no exact visit date, and the comparison does not cover every product or total delivered cost. Sources: - Jeffrey P. Bezos, 2002 Letter to Shareholders (2002 reporting period; published 2003), Opening discussion of customer experience, price, satisfaction, the 100-book comparison, and free cash flow. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2002_shareholderLetter.pdf Evidence note: Contemporary company evidence. The 100-book basket, satisfaction score and management's free-cash-flow definition have specific scopes and do not prove a universal price rule. ### Give one balancing force that can weaken a growth flywheel. Examples include congestion, working-capital strain, returns, seller dissatisfaction, service defects or bureaucracy. Sources: - Jeffrey P. Bezos, 2001 Letter to Shareholders (2001 reporting period; published 2002), Opening discussion of price, cost reduction, growth, fixed costs and cash flow. https://ir.aboutamazon.com/files/doc_financials/annual/2001_shareholderLetter.pdf Evidence note: Management's contemporary explanation of a reinforcing operating cycle after reporting its first profitable quarter on the letter's pro forma measures. - Jeffrey P. Bezos, 2002 Letter to Shareholders (2002 reporting period; published 2003), Opening discussion of customer experience, price, satisfaction, the 100-book comparison, and free cash flow. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2002_shareholderLetter.pdf Evidence note: Contemporary company evidence. The 100-book basket, satisfaction score and management's free-cash-flow definition have specific scopes and do not prove a universal price rule. ### In the 2004 letter's fictional four-year transportation-machine example, what was cumulative free cash flow? Negative $530 million in United States dollars under the four-year assumptions. Sources: - Jeffrey P. Bezos, 2004 Letter to Shareholders (April 2005), Our Most Important Financial Measure, Free Cash Flow Per Share, including the transportation-machine example. https://ir.aboutamazon.com/files/doc_financials/annual/2004_Annual_report.pdf Evidence note: Uses a deliberately simplified fictional machine to separate earnings growth from cash generation, followed by Amazon's 2004 company figures. ### Why did EBITDA mislead in the 2004 letter's fictional four-year transportation example? It omitted the $1.28 billion of machines required during the four years to create the growing trip capacity. Sources: - Jeffrey P. Bezos, 2004 Letter to Shareholders (April 2005), Our Most Important Financial Measure, Free Cash Flow Per Share, including the transportation-machine example. https://ir.aboutamazon.com/files/doc_financials/annual/2004_Annual_report.pdf Evidence note: Uses a deliberately simplified fictional machine to separate earnings growth from cash generation, followed by Amazon's 2004 company figures. ### Why add per share to a free-cash-flow objective? Issuing shares or stock awards can increase total company cash while reducing each existing owner's claim. Sources: - Jeffrey P. Bezos, 2004 Letter to Shareholders (April 2005), Our Most Important Financial Measure, Free Cash Flow Per Share, including the transportation-machine example. https://ir.aboutamazon.com/files/doc_financials/annual/2004_Annual_report.pdf Evidence note: Uses a deliberately simplified fictional machine to separate earnings growth from cash generation, followed by Amazon's 2004 company figures. ### Name two reasons current free cash flow may misstate durable owner economics. Examples include supplier-payment timing, leases, dilution, acquisitions, deferred maintenance or investment ahead of demand. Sources: - Jeffrey P. Bezos, 2004 Letter to Shareholders (April 2005), Our Most Important Financial Measure, Free Cash Flow Per Share, including the transportation-machine example. https://ir.aboutamazon.com/files/doc_financials/annual/2004_Annual_report.pdf Evidence note: Uses a deliberately simplified fictional machine to separate earnings growth from cash generation, followed by Amazon's 2004 company figures. ### What made fulfilment location more math-based in Amazon's account? Repeated operating data allowed alternatives to be modelled using demand, product, vendor, transport and customer-location evidence. Sources: - Jeffrey P. Bezos, 2005 Letter to Shareholders (2005 reporting period; published 2006), Opening discussion of math-based and judgment-based decisions, price reductions, and the single detail page. https://ir.aboutamazon.com/files/doc_financials/annual/AMZN2005AnnualReport.pdf Evidence note: Contemporary account of Amazon's decision policy with retrospective results for the Marketplace choice made in 2000. Long-run price effects were explicitly not measurable in advance. ### Why did Amazon say its price model could not settle the long-term decision? It estimated weekly or quarterly elasticity but not the five-to-ten-year effect of repeated price reductions on customers and cash flow. Sources: - Jeffrey P. Bezos, 2005 Letter to Shareholders (2005 reporting period; published 2006), Opening discussion of math-based and judgment-based decisions, price reductions, and the single detail page. https://ir.aboutamazon.com/files/doc_financials/annual/AMZN2005AnnualReport.pdf Evidence note: Contemporary account of Amazon's decision policy with retrospective results for the Marketplace choice made in 2000. Long-run price effects were explicitly not measurable in advance. ### What keeps judgment from becoming an untestable story? Named model limits, comparable evidence, bounded downside, staged commitment, leading indicators and review triggers. Sources: - Jeffrey P. Bezos, 2005 Letter to Shareholders (2005 reporting period; published 2006), Opening discussion of math-based and judgment-based decisions, price reductions, and the single detail page. https://ir.aboutamazon.com/files/doc_financials/annual/AMZN2005AnnualReport.pdf Evidence note: Contemporary account of Amazon's decision policy with retrospective results for the Marketplace choice made in 2000. Long-run price effects were explicitly not measurable in advance. ### What remained unknown in the 2005 letter even though Marketplace and Amazon retail had both grown? The counterfactual outcome without the single detail page and the causal contribution of that choice to each channel's growth. Sources: - Jeffrey P. Bezos, 2005 Letter to Shareholders (2005 reporting period; published 2006), Opening discussion of math-based and judgment-based decisions, price reductions, and the single detail page. https://ir.aboutamazon.com/files/doc_financials/annual/AMZN2005AnnualReport.pdf Evidence note: Contemporary account of Amazon's decision policy with retrospective results for the Marketplace choice made in 2000. Long-run price effects were explicitly not measurable in advance. ### Why can an input goal be more actionable than a revenue command? A team can directly change selection, defects, availability, delivery or cost, while revenue is an output affected by several causes. Sources: - Jeffrey P. Bezos, 2009 Letter to Shareholders (April 2010), Opening results and the 452 goals for 2010. https://s2.q4cdn.com/299287126/files/doc_financials/annual/AMZN_Shareholder-Letter-2009-(final).pdf Evidence note: Reports the content of one annual planning set. It shows Amazon's chosen management emphasis, not proof that financial outputs can be ignored. ### What does the 2009 letter's analysis of 452 goals for the 2010 planning set establish and not establish? It establishes Amazon's stated emphasis in one planning set. It does not prove that financial constraints were absent or that the approach caused results. Sources: - Jeffrey P. Bezos, 2009 Letter to Shareholders (April 2010), Opening results and the 452 goals for 2010. https://s2.q4cdn.com/299287126/files/doc_financials/annual/AMZN_Shareholder-Letter-2009-(final).pdf Evidence note: Reports the content of one annual planning set. It shows Amazon's chosen management emphasis, not proof that financial outputs can be ignored. ### What facts determine whether a decision is truly reversible? Capital recovery, contracts, switching costs, data migration, customer harm, legal commitments and reputation, not the label assigned by a team. Sources: - Jeffrey P. Bezos, 2015 Letter to Shareholders (2015 reporting period; published 2016), Failure, long-tailed business returns, and Type 1 versus Type 2 decisions. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2015-Letter-to-Shareholders.PDF Evidence note: Retrospective management account. The ten-percent and one-hundred-times illustration explains an asymmetric payoff shape, not a measured base rate for Amazon experiments. ### Why is the 2015 letter's hypothetical ten-percent, hundred-times example not a base rate? Bezos used it to illustrate asymmetric payoffs. The letter did not report that probability or payoff as Amazon's measured experiment distribution. Sources: - Jeffrey P. Bezos, 2015 Letter to Shareholders (2015 reporting period; published 2016), Failure, long-tailed business returns, and Type 1 versus Type 2 decisions. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2015-Letter-to-Shareholders.PDF Evidence note: Retrospective management account. The ten-percent and one-hundred-times illustration explains an asymmetric payoff shape, not a measured base rate for Amazon experiments. ### Why can a forced-sale price differ from a patient owner's value estimate? The forced seller is optimizing for immediate liquidity, while the patient owner is valuing future cash flows. Funding pressure can determine the transaction price before fundamentals are resolved. Sources: - Howard Marks, The Tide Goes Out (2008-03-18), Leveraging and Inflating; Mark-to-Market Accounting; Should Does Not Equal Will; Leverage and Risk. https://www.oaktreecapital.com/insights/memo/the-tide-goes-out Evidence note: Contemporaneous analysis written before the worst phase of the financial crisis. Examples and conclusions express the author's assessment and are not an independent performance study. ### What two estimates should an investor keep separate during a dislocation? A fundamental value range under stated operating assumptions and a path-dependent range of market prices that liquidity, leverage, and forced selling may produce. Sources: - Howard Marks, The Tide Goes Out (2008-03-18), Leveraging and Inflating; Mark-to-Market Accounting; Should Does Not Equal Will; Leverage and Risk. https://www.oaktreecapital.com/insights/memo/the-tide-goes-out Evidence note: Contemporaneous analysis written before the worst phase of the financial crisis. Examples and conclusions express the author's assessment and are not an independent performance study. ### What evidence would falsify a liquidity-dislocation thesis? New evidence of weaker cash flows, impaired collateral, worse claim priority, covenant failure, or lower recoveries sufficient to justify the price decline. Sources: - Howard Marks, The Tide Goes Out (2008-03-18), Leveraging and Inflating; Mark-to-Market Accounting; Should Does Not Equal Will; Leverage and Risk. https://www.oaktreecapital.com/insights/memo/the-tide-goes-out Evidence note: Contemporaneous analysis written before the worst phase of the financial crisis. Examples and conclusions express the author's assessment and are not an independent performance study. - Howard Marks, Nobody Knows (2008-09-19), Does the Market Know?; The End of the Financial System; final bear-market-stage discussion. https://www.oaktreecapital.com/insights/memo/nobody-knows Evidence note: Contemporaneous memo after Lehman's bankruptcy. It records a decision premise and a search for bargains, but no security-level trade list. ### Why is buying solely because an asset has fallen a weak argument? The prior high is not proof of value. The fall may reflect permanent impairment, so value must be rebuilt independently from cash flows, structure, and recoveries. Sources: - Howard Marks, The Limits to Negativism (2008-10-15), Final discussion of third-stage pessimism and Oaktree purchases. https://www.oaktreecapital.com/insights/memo/the-limits-to-negativism Evidence note: Contemporaneous practitioner report. It describes purchase direction and averaging down, but not positions, allocations, or later investment results. ### What extra downside does leverage add beyond magnifying losses? It can force liquidation or end the strategy through margin calls, covenants, or withdrawn funding before the asset's terminal value is known. Sources: - Howard Marks, The Tide Goes Out (2008-03-18), Leveraging and Inflating; Mark-to-Market Accounting; Should Does Not Equal Will; Leverage and Risk. https://www.oaktreecapital.com/insights/memo/the-tide-goes-out Evidence note: Contemporaneous analysis written before the worst phase of the financial crisis. Examples and conclusions express the author's assessment and are not an independent performance study. ### Why were highly rated assets insufficient to make Carlyle Capital safe in Marks's example? The figures imply about 32.4 times assets to equity, or about 31.4 times debt to equity. A roughly 3 percent asset decline could consume equity while lenders demanded collateral or seized assets. Sources: - Howard Marks, The Tide Goes Out (2008-03-18), Leveraging and Inflating; Mark-to-Market Accounting; Should Does Not Equal Will; Leverage and Risk. https://www.oaktreecapital.com/insights/memo/the-tide-goes-out Evidence note: Contemporaneous analysis written before the worst phase of the financial crisis. Examples and conclusions express the author's assessment and are not an independent performance study. ### What is the most consequential category of adverse event in a survival analysis? An event that forces action, impairs obligations, or removes the ability to invest later, rather than one that merely lowers a reported return. Sources: - Howard Marks, The Tide Goes Out (2008-03-18), Leveraging and Inflating; Mark-to-Market Accounting; Should Does Not Equal Will; Leverage and Risk. https://www.oaktreecapital.com/insights/memo/the-tide-goes-out Evidence note: Contemporaneous analysis written before the worst phase of the financial crisis. Examples and conclusions express the author's assessment and are not an independent performance study. - Howard Marks, Coming into Focus (2020-10-13), The Power of Interest Rates; crisis comparison; final portfolio-calibration discussion. https://www.oaktreecapital.com/insights/memo/coming-into-focus Evidence note: Retrospective report on Oaktree's pre-crisis posture and March buying, followed by a dated October assessment. It does not disclose portfolio allocations. ### When does dry powder have strategic value? When durable financing and liquid capacity let an investor act after prices improve without having been forced to repair earlier positions. Sources: - Howard Marks, Coming into Focus (2020-10-13), The Power of Interest Rates; crisis comparison; final portfolio-calibration discussion. https://www.oaktreecapital.com/insights/memo/coming-into-focus Evidence note: Retrospective report on Oaktree's pre-crisis posture and March buying, followed by a dated October assessment. It does not disclose portfolio allocations. ### Why is disagreement with the crowd insufficient for contrarian investing? The investor still needs a causal reason the crowd's price is wrong, an independent value estimate, funding to survive being early, and falsification evidence. Sources: - Howard Marks, Nobody Knows (2008-09-19), Does the Market Know?; The End of the Financial System; final bear-market-stage discussion. https://www.oaktreecapital.com/insights/memo/nobody-knows Evidence note: Contemporaneous memo after Lehman's bankruptcy. It records a decision premise and a search for bargains, but no security-level trade list. - Howard Marks, The Limits to Negativism (2008-10-15), Final discussion of third-stage pessimism and Oaktree purchases. https://www.oaktreecapital.com/insights/memo/the-limits-to-negativism Evidence note: Contemporaneous practitioner report. It describes purchase direction and averaging down, but not positions, allocations, or later investment results. ### What practical premise let Marks continue analyzing investments in September 2008? He judged that assuming the system would continue was the only viable basis for action because end-of-system hedges were limited and costly if ordinary life continued. Sources: - Howard Marks, Nobody Knows (2008-09-19), Does the Market Know?; The End of the Financial System; final bear-market-stage discussion. https://www.oaktreecapital.com/insights/memo/nobody-knows Evidence note: Contemporaneous memo after Lehman's bankruptcy. It records a decision premise and a search for bargains, but no security-level trade list. ### What did Oaktree publicly report doing by October 15, 2008? Marks reported long purchase lists on most recent days, almost no sales, substantial use of available cash, and averaging down. Sources: - Howard Marks, The Limits to Negativism (2008-10-15), Final discussion of third-stage pessimism and Oaktree purchases. https://www.oaktreecapital.com/insights/memo/the-limits-to-negativism Evidence note: Contemporaneous practitioner report. It describes purchase direction and averaging down, but not positions, allocations, or later investment results. ### What does the reported $450 million weekly pace in late 2008 prove and not prove? It is Marks's 2020 retrospective account of average deployment over 15 weeks. It does not disclose assets, client allocations, daily rules, or independently audited results. Sources: - Howard Marks, Uncertainty (2020-05-11), All We Don't Know; In Praise of Doubt; confidence and position-sizing discussion. https://www.oaktreecapital.com/insights/memo/uncertainty Evidence note: The discussion of 2008 includes a retrospective practitioner report. The memo distinguishes evidence-based confidence from certainty. ### What second judgment should accompany every forecast? An assessment of how likely the forecast is to be dependable given the domain, evidence, precedent, and interaction of uncertain variables. Sources: - Howard Marks, Uncertainty (2020-05-11), All We Don't Know; In Praise of Doubt; confidence and position-sizing discussion. https://www.oaktreecapital.com/insights/memo/uncertainty Evidence note: The discussion of 2008 includes a retrospective practitioner report. The memo distinguishes evidence-based confidence from certainty. ### Why may a consensus macro forecast offer little investment advantage even when it is right? Common expectations are usually already reflected in asset prices, so being broadly correct may not produce a mispricing. Sources: - Howard Marks, Uncertainty (2020-05-11), All We Don't Know; In Praise of Doubt; confidence and position-sizing discussion. https://www.oaktreecapital.com/insights/memo/uncertainty Evidence note: The discussion of 2008 includes a retrospective practitioner report. The memo distinguishes evidence-based confidence from certainty. ### How does intellectual humility constrain exposure rather than merely language? It lowers reliance on fragile claims, widens tested scenarios, preserves revision capacity, and sizes the position to the weakest consequential evidence. Sources: - Howard Marks, Uncertainty (2020-05-11), All We Don't Know; In Praise of Doubt; confidence and position-sizing discussion. https://www.oaktreecapital.com/insights/memo/uncertainty Evidence note: The discussion of 2008 includes a retrospective practitioner report. The memo distinguishes evidence-based confidence from certainty. ### What was the central decision posture in Marks's March 31, 2020 memo? He laid out opposing paths, judged prices insufficiently protective against worse news, and emphasized readiness to exploit further declines. Sources: - Howard Marks, Which Way Now? (2020-03-31), The Positive Case; the negative case; The Government Programs; Summing Up. https://www.oaktreecapital.com/insights/memo/which-way-now Evidence note: Dated scenario analysis and author judgment during the pandemic shock. It is not an Oaktree transaction ledger. ### What operational benefit did Marks attribute to Oaktree's pre-2020 caution? He said it reduced portfolio remediation and left some drawdown funds with capital to buy when bargains peaked in March. Sources: - Howard Marks, Coming into Focus (2020-10-13), The Power of Interest Rates; crisis comparison; final portfolio-calibration discussion. https://www.oaktreecapital.com/insights/memo/coming-into-focus Evidence note: Retrospective report on Oaktree's pre-crisis posture and March buying, followed by a dated October assessment. It does not disclose portfolio allocations. ### Why should an investor reassess after a rapid recovery even if the economic story remains uncertain? Prospective return and downside are functions of today's price. A prior bargain can become fully priced before the macro uncertainty resolves. Sources: - Howard Marks, The Anatomy of a Rally (2020-06-18), Market chronology; psychology of the rally; final assessment questions. https://www.oaktreecapital.com/insights/memo/the-anatomy-of-a-rally Evidence note: Contemporaneous market assessment after the spring rally. The judgment is not evidence of later excess returns. - Howard Marks, Coming into Focus (2020-10-13), The Power of Interest Rates; crisis comparison; final portfolio-calibration discussion. https://www.oaktreecapital.com/insights/memo/coming-into-focus Evidence note: Retrospective report on Oaktree's pre-crisis posture and March buying, followed by a dated October assessment. It does not disclose portfolio allocations. ### What bottleneck was distributed acquisition work meant to relieve at CSI? Cash and small acquisition opportunities were growing faster than a central head-office team could source, assess, and own every decision. Sources: - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. ### What is the difference between distributed judgment and distributed permission? Distributed judgment gives local managers bounded authority, comparable evidence, challenge, ownership, and review. Permission alone lacks those controls. Sources: - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. ### Why can a permanent holding period justify more acquisition effort on small deals? One-time search and transaction costs can be spread across a long ownership period if the underlying return remains attractive. Sources: - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. ### Why should decision authority and capital limits sit at different organizational levels? Local managers may have better specific information, while a higher level may be better able to absorb, compare, and control the cost of a large error. Sources: - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. ### What did CSI's internal vintage analysis suggest about changed hurdle rates? Leonard reported that weighted expected acquisition IRRs tended to move toward the hurdle then in force, including after reductions. Sources: - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. ### Why might lowering a hurdle affect more than marginal approvals? It can alter sourcing, bidding, forecast assumptions, negotiation pressure, and what the entire team treats as acceptable. Sources: - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. ### How did CSI respond to the large transaction described in its April 25, 2017 letter that narrowly missed its hurdle? It tried to improve the structure by a few projected IRR points, failed to do so, and did not invest. Sources: - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. ### Why can the same borderline forecast justify different choices for small and very large deals? The cost of error and value of learning differ. A small bounded experiment can teach, while a large mistake can overwhelm the portfolio and needs stronger control. Sources: - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. ### Why did CSI try to review acquisitions roughly one year after investment? That timing could reveal leading operating errors while the original assumptions, records, and people were still available. Sources: - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. ### Why should a post-acquisition review separate process from outcome? A reasonable decision can be unlucky and a weak decision can be lucky. The firm needs to assess original evidence and update base rates from actual results separately. Sources: - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. ### What makes an exceptional forecast assumption useful even before the outcome is known? Its place in the tail of historical results becomes visible, requiring the sponsor to explain why the case should differ from the base rate. Sources: - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. ### What costs did CSI's autonomy design deliberately risk? Duplicated functions, missed economies of scale, inconsistent practices, and possible difficulty coordinating products across units. Sources: - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. ### What evidence limit did Leonard state about human-scale business units? He said the idea was not universally accepted inside CSI and lacked compelling data, even though several groups had experimented with it. Sources: - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. ### What should an autonomy charter contain? Local decision rights, a few parent-monitored outcomes, magnitude-based escalation triggers, and measures of both duplicated cost and local benefits. Sources: - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. ### What changed in CSI's special-dividend policy in February 2021? The board stopped special dividends except in compelling circumstances, while retaining the regular quarterly dividend for the time being. Sources: - Mark Leonard, Constellation Software Inc. 2021 President's Letter (2021-02-15), Complete letter, especially dividend policy, large VMS acquisitions, and head-office role. https://www.csisoftware.com/wp-content/uploads/2026/04/letter-to-shareholders-february-15-2021.pdf Evidence note: Records a board policy decision and management intentions as of publication. The reported 16 percent invitation figure does not have an unambiguous denominator in the letter. It does not establish the results of later investments. ### What access problem did Leonard report in the large VMS market? CSI knew of roughly 80 percent of the large VMS businesses sold over five years but had been invited into only 16 percent of the sale processes. The letter does not define whether that denominator covers all large VMS sale processes or only the roughly 80 percent CSI knew about. Sources: - Mark Leonard, Constellation Software Inc. 2021 President's Letter (2021-02-15), Complete letter, especially dividend policy, large VMS acquisitions, and head-office role. https://www.csisoftware.com/wp-content/uploads/2026/04/letter-to-shareholders-february-15-2021.pdf Evidence note: Records a board policy decision and management intentions as of publication. The reported 16 percent invitation figure does not have an unambiguous denominator in the letter. It does not establish the results of later investments. ### What tradeoff did Leonard expect if CSI completed one or two large VMS acquisitions each year? More cash could be deployed, but return on investors' capital was expected to decline. Sources: - Mark Leonard, Constellation Software Inc. 2021 President's Letter (2021-02-15), Complete letter, especially dividend policy, large VMS acquisitions, and head-office role. https://www.csisoftware.com/wp-content/uploads/2026/04/letter-to-shareholders-february-15-2021.pdf Evidence note: Records a board policy decision and management intentions as of publication. The reported 16 percent invitation figure does not have an unambiguous denominator in the letter. It does not establish the results of later investments. ### What diagnosis should precede a lower hurdle when cash accumulates? Determine whether the constraint is weak sourcing, poor process access, limited market size, high prices, or a real lack of attractive opportunities. Sources: - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. - Mark Leonard, Constellation Software Inc. 2017 President's Letter (2018-04-20), Performance metrics; retained-capital policy; perpetual-owner objective; business-unit count. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-april-2018-final.pdf Evidence note: Management report. Statements about scalability and governance are Leonard's judgments, not independent causal findings. - Mark Leonard, Constellation Software Inc. 2021 President's Letter (2021-02-15), Complete letter, especially dividend policy, large VMS acquisitions, and head-office role. https://www.csisoftware.com/wp-content/uploads/2026/04/letter-to-shareholders-february-15-2021.pdf Evidence note: Records a board policy decision and management intentions as of publication. The reported 16 percent invitation figure does not have an unambiguous denominator in the letter. It does not establish the results of later investments. ### What complicates the story that Buffett recognized GEICO once and simply held it? He sold his early holding, returned during GEICO's 1976 crisis, and Berkshire acquired the whole company only in 1996. Sources: - Warren E. Buffett, 1995 Chairman's Letter (1995 reporting year; published 1996), GEICO section, including Buffett's retrospective account of 1951-1952, the 1976 return and the 1995 agreement to acquire the remaining shares. https://www.berkshirehathaway.com/letters/1995.html Evidence note: Official original. This is later self-reporting about the 1951 episode, not contemporaneous evidence of what the unread 1951 article said in full. - Warren E. Buffett, Buffett Partnership letters compilation (1957-1970), Complete 152-page compilation, especially the 1957 and 1958 letters, January 18, 1963 Ground Rules and Dempster discussion, January 25 and October 9, 1967 letters, January 22 and May 29, 1969 letters, and February 25, 1970 letter. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: A third-party-hosted compilation of reproduced original letters. It is not an official Berkshire archive. Datelines and signatures were checked in the documents. All 29 located documents dated 1957 through 1970 were read in full; the earlier 1956 original referenced in the 1957 letter was not present. ### Why did Buffett warn that the early partnership might lag in a strong bull market? Illiquid undervalued securities and workouts could remain dormant while popular stocks rose quickly. Sources: - Warren E. Buffett, Second Annual Letter to Limited Partners (1957 reporting year; circulated 1958), Reproduced original, printed pages 1-3. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. The reproduced original identifies itself as the second annual letter and includes a fragment of the unavailable prior letter. - Warren E. Buffett, 1958 partnership letter (1958 reporting year; circulated 1959), Reproduced original, printed pages 4-6. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Includes the complete Commonwealth Trust account and the stated expectation for performance in a rising market. - Warren E. Buffett, Buffett Partnership letters compilation (1957-1970), Complete 152-page compilation, especially the 1957 and 1958 letters, January 18, 1963 Ground Rules and Dempster discussion, January 25 and October 9, 1967 letters, January 22 and May 29, 1969 letters, and February 25, 1970 letter. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: A third-party-hosted compilation of reproduced original letters. It is not an official Berkshire archive. Datelines and signatures were checked in the documents. All 29 located documents dated 1957 through 1970 were read in full; the earlier 1956 original referenced in the 1957 letter was not present. ### Why could selling Commonwealth below estimated full value still be rational? Another opportunity offered a better use of the same capital, so the relevant comparison was opportunity cost rather than maximum possible sale price. Sources: - Warren E. Buffett, 1958 partnership letter (1958 reporting year; circulated 1959), Reproduced original, printed pages 4-6. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Includes the complete Commonwealth Trust account and the stated expectation for performance in a rising market. - Warren E. Buffett, Buffett Partnership letters compilation (1957-1970), Complete 152-page compilation, especially the 1957 and 1958 letters, January 18, 1963 Ground Rules and Dempster discussion, January 25 and October 9, 1967 letters, January 22 and May 29, 1969 letters, and February 25, 1970 letter. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: A third-party-hosted compilation of reproduced original letters. It is not an official Berkshire archive. Datelines and signatures were checked in the documents. All 29 located documents dated 1957 through 1970 were read in full; the earlier 1956 original referenced in the 1957 letter was not present. ### What new responsibility appeared when a cheap security became a control investment? Buffett had to convert assets, select managers and bear operating consequences rather than wait for market repricing. Sources: - Warren E. Buffett, Ground Rules and 1962 partnership review (January 18, 1963), Reproduced original, printed pages 32-41, especially Dempster on pages 37-39. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Contemporary account of Dempster before and after the April 1962 management change. - Warren E. Buffett, 1963 partnership review (January 18, 1964), Reproduced original, printed pages 51-62, especially the Dempster outcome on pages 59-61. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full in the local original compilation, printed pages 51-62. The source records the expanding capital base, performance comparisons and Dempster follow-through. - Warren E. Buffett, Buffett Partnership letters compilation (1957-1970), Complete 152-page compilation, especially the 1957 and 1958 letters, January 18, 1963 Ground Rules and Dempster discussion, January 25 and October 9, 1967 letters, January 22 and May 29, 1969 letters, and February 25, 1970 letter. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: A third-party-hosted compilation of reproduced original letters. It is not an official Berkshire archive. Datelines and signatures were checked in the documents. All 29 located documents dated 1957 through 1970 were read in full; the earlier 1956 original referenced in the 1957 letter was not present. ### Why was delay costly at Dempster even with discounted asset value? Operating losses and stagnant inventory could consume the asset protection while management failed to act. Sources: - Warren E. Buffett, Ground Rules and 1962 partnership review (January 18, 1963), Reproduced original, printed pages 32-41, especially Dempster on pages 37-39. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Contemporary account of Dempster before and after the April 1962 management change. - Warren E. Buffett, Buffett Partnership letters compilation (1957-1970), Complete 152-page compilation, especially the 1957 and 1958 letters, January 18, 1963 Ground Rules and Dempster discussion, January 25 and October 9, 1967 letters, January 22 and May 29, 1969 letters, and February 25, 1970 letter. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: A third-party-hosted compilation of reproduced original letters. It is not an official Berkshire archive. Datelines and signatures were checked in the documents. All 29 located documents dated 1957 through 1970 were read in full; the earlier 1956 original referenced in the 1957 letter was not present. ### What did the partnership Ground Rules control? They aligned expectations about withdrawals, comparison periods, market forecasts, loss risk and Buffett's own financial alignment. Sources: - Warren E. Buffett, Ground Rules and 1962 partnership review (January 18, 1963), Reproduced original, printed pages 32-41, especially Dempster on pages 37-39. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Contemporary account of Dempster before and after the April 1962 management change. - Warren E. Buffett, Buffett Partnership letters compilation (1957-1970), Complete 152-page compilation, especially the 1957 and 1958 letters, January 18, 1963 Ground Rules and Dempster discussion, January 25 and October 9, 1967 letters, January 22 and May 29, 1969 letters, and February 25, 1970 letter. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: A third-party-hosted compilation of reproduced original letters. It is not an official Berkshire archive. Datelines and signatures were checked in the documents. All 29 located documents dated 1957 through 1970 were read in full; the earlier 1956 original referenced in the 1957 letter was not present. ### How did partnership success weaken the method that produced it? A much larger capital base could no longer use many small bargains, while obvious quantitative opportunities had become scarcer. Sources: - Warren E. Buffett, Buffett Partnership letters compilation (1957-1970), Complete 152-page compilation, especially the 1957 and 1958 letters, January 18, 1963 Ground Rules and Dempster discussion, January 25 and October 9, 1967 letters, January 22 and May 29, 1969 letters, and February 25, 1970 letter. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: A third-party-hosted compilation of reproduced original letters. It is not an official Berkshire archive. Datelines and signatures were checked in the documents. All 29 located documents dated 1957 through 1970 were read in full; the earlier 1956 original referenced in the 1957 letter was not present. ### Which four pressures drove Buffett's October 1967 goal revision? Scarcer bargains, shorter-term professional speculation, a larger capital base and his desire for a less compulsive pace. Sources: - Warren E. Buffett, Partnership objective-revision letter (October 9, 1967), Reproduced original, printed pages 111-114. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Buffett explains why scale, scarce bargains and a changed market led him to reduce the partnership objective. - Warren E. Buffett, Buffett Partnership letters compilation (1957-1970), Complete 152-page compilation, especially the 1957 and 1958 letters, January 18, 1963 Ground Rules and Dempster discussion, January 25 and October 9, 1967 letters, January 22 and May 29, 1969 letters, and February 25, 1970 letter. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: A third-party-hosted compilation of reproduced original letters. It is not an official Berkshire archive. Datelines and signatures were checked in the documents. All 29 located documents dated 1957 through 1970 were read in full; the earlier 1956 original referenced in the 1957 letter was not present. ### Why did the exceptional 1968 result not invalidate the decision to close the partnership? One concentrated success did not restore the repeatable supply and scale of suitable ideas required by the mandate. Sources: - Warren E. Buffett, Buffett Partnership letters compilation (1957-1970), Complete 152-page compilation, especially the 1957 and 1958 letters, January 18, 1963 Ground Rules and Dempster discussion, January 25 and October 9, 1967 letters, January 22 and May 29, 1969 letters, and February 25, 1970 letter. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: A third-party-hosted compilation of reproduced original letters. It is not an official Berkshire archive. Datelines and signatures were checked in the documents. All 29 located documents dated 1957 through 1970 were read in full; the earlier 1956 original referenced in the 1957 letter was not present. ### What did Berkshire's 1967 report imply about textiles and insurance at the time? Insurance diversified earnings and used capital better, but management still expected relative profitability to reverse in some years. Sources: - Warren E. Buffett; signed by Berkshire officers as noted, Early Berkshire Hathaway annual reports (1965-1976 reporting years), Independent archive of hosted original-report facsimiles. Exact report locators used here: 1965 report, printed pages 1-2, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1965-Berkshire-AR.pdf; 1966 report, printed pages 1, 3-6, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1966-Berkshire-AR.pdf; 1967 report, printed pages 1, 3-4, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1967-Berkshire-AR.pdf; 1975 report, printed pages 1 and 7-8, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1975-Berkshire-AR.pdf; 1976 report, printed pages 1-2, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1976-Berkshire-AR.pdf. Berkshire’s official individual-letter archive begins at 1977.. https://theoraclesclassroom.com/archives/ Evidence note: These are independently hosted reproductions of the original reports, checked against the local original PDFs. The 1965-1969 reports are signed by Malcolm G. Chace Jr. and/or Kenneth V. Chace, while the local originals state that Buffett wrote the letter. The 1970-1976 reports are signed by Buffett. Berkshire’s official page supplies individual letters only from 1977 and a broader 1965-2024 compilation. ### Which evidence shows that the textile exit was not an immediate response to Munger's advice? Berkshire continued equipment spending, forecasts of recovery, repair attempts and consideration of another acquisition through the 1976 reporting year. Sources: - Warren E. Buffett; signed by Berkshire officers as noted, Early Berkshire Hathaway annual reports (1965-1976 reporting years), Independent archive of hosted original-report facsimiles. Exact report locators used here: 1965 report, printed pages 1-2, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1965-Berkshire-AR.pdf; 1966 report, printed pages 1, 3-6, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1966-Berkshire-AR.pdf; 1967 report, printed pages 1, 3-4, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1967-Berkshire-AR.pdf; 1975 report, printed pages 1 and 7-8, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1975-Berkshire-AR.pdf; 1976 report, printed pages 1-2, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1976-Berkshire-AR.pdf. Berkshire’s official individual-letter archive begins at 1977.. https://theoraclesclassroom.com/archives/ Evidence note: These are independently hosted reproductions of the original reports, checked against the local original PDFs. The 1965-1969 reports are signed by Malcolm G. Chace Jr. and/or Kenneth V. Chace, while the local originals state that Buffett wrote the letter. The 1970-1976 reports are signed by Buffett. Berkshire’s official page supplies individual letters only from 1977 and a broader 1965-2024 compilation. ### How did Buffett revise behavior after the 1974 insurance error? He restricted volume, increased liquidity and waited for prices that matched the risk instead of pursuing growth. Sources: - Warren E. Buffett, Berkshire Hathaway 1974 annual letter (1974 reporting year; signed March 31, 1975), Insurance underwriting, Home and Auto, liquidity and Blue Chip Stamps. https://www.berkshirehathaway.com/letters/letters.html Evidence note: Contemporary report during an insurance downturn. The full text was available locally; the public URL is Berkshire's archive page because it does not expose an individual 1974 file. ### What did See's change about the comparison applied to textile capital? It showed that a business could grow earnings with little additional capital and release cash for other uses. Sources: - Warren E. Buffett, 1983 Chairman's Letter (1983 reporting period; published 1984), Goodwill and its Amortization, especially the See's comparison. https://www.berkshirehathaway.com/letters/1983.html Evidence note: Contemporary discussion of accounting and economic goodwill using historical See's figures. The acquisition rationale is partly retrospective. ### How could individually attractive textile machines still destroy value? Competitors could make the same investments, pass savings into lower prices and leave every owner with more capital earning poor returns. Sources: - Warren E. Buffett, 1985 Chairman's Letter (March 4, 1986), Shutdown of Textile Business. https://www.berkshirehathaway.com/letters/1985.html Evidence note: Contemporary account written after the July 1985 closure decision. It reports Buffett's own diagnosis and hindsight about earlier delay. ### When did Buffett's persistence boundary finally break in textiles? When continued support looked capable of becoming an endless capital claim without a durable improvement in industry economics. Sources: - Warren E. Buffett, 1985 Chairman's Letter (March 4, 1986), Shutdown of Textile Business. https://www.berkshirehathaway.com/letters/1985.html Evidence note: Contemporary account written after the July 1985 closure decision. It reports Buffett's own diagnosis and hindsight about earlier delay. ### What replaced asset labels in Buffett's developing account of risk? The probability and consequence of permanent loss, including claims, inflation, leverage, liquidity needs and required reinvestment. Sources: - Warren E. Buffett, Berkshire Hathaway 1974 annual letter (1974 reporting year; signed March 31, 1975), Insurance underwriting, Home and Auto, liquidity and Blue Chip Stamps. https://www.berkshirehathaway.com/letters/letters.html Evidence note: Contemporary report during an insurance downturn. The full text was available locally; the public URL is Berkshire's archive page because it does not expose an individual 1974 file. - Warren E. Buffett, How Inflation Swindles the Equity Investor (May 1, 1977), The sticky equity coupon, retained earnings and capital required merely to maintain physical output. https://fortune.com/article/buffett-how-inflation-swindles-the-equity-investor-fortune-classics-1977/ Evidence note: Original publisher page. The article develops an argument that was contemporaneous with Berkshire's capital-intensive textile experience. ### What limited Buffett's 1984 evidence against pure chance? He presented pre-identified value investors with audited records, but the group was selected through a shared intellectual lineage rather than a randomized test. Sources: - Warren E. Buffett, The Superinvestors of Graham-and-Doddsville (May 17, 1984), Accessible article text and performance tables. https://business.columbia.edu/insights/chazen-global-insights/superinvestors-graham-and-doddsville Evidence note: Columbia Business School's publication of Buffett's argument against a pure-chance explanation for the records of value investors he had identified in advance. ### Why can an insurance mistake persist after management recognizes it? Existing contracts and underestimated reserves can produce claims and corrections many years after the original pricing decision. Sources: - Warren E. Buffett, 1989 Chairman's Letter (1989 reporting year; signed March 2, 1990), Mistakes of the First Twenty-five Years and the institutional imperative. https://www.berkshirehathaway.com/letters/1989.html Evidence note: Retrospective classification of commission, omission and institutional errors after twenty-five years of Berkshire control. ### What did Buffett add beyond legal compliance at Salomon? He required conduct that could survive informed public scrutiny and made reputation an operating constraint. Sources: - Warren E. Buffett, Salomon Brothers securities-trading investigation testimony (September 4-5, 1991), Opening apology, cooperation commitments, compliance standard and response to lawmakers. https://www.youtube.com/watch?v=MtaeGt3KwuA Evidence note: Public recording of Buffett's testimony before a House subcommittee. The hearing was reviewed in part; the course does not claim a complete transcript read. ### What does the Salomon rescue fail to prove? It does not prove that Berkshire or Salomon detected misconduct early or that decentralized trust prevents governance failures. Sources: - Warren E. Buffett, 1991 Chairman's Letter (1991 reporting year; signed February 28, 1992), Salomon, omission errors, USAir and fixed-income securities. https://www.berkshirehathaway.com/letters/1991.html Evidence note: Buffett's own report after taking the interim chair at Salomon. It does not replace the hearing record or independent reporting on the scandal. - Warren E. Buffett, Salomon Brothers securities-trading investigation testimony (September 4-5, 1991), Opening apology, cooperation commitments, compliance standard and response to lawmakers. https://www.youtube.com/watch?v=MtaeGt3KwuA Evidence note: Public recording of Buffett's testimony before a House subcommittee. The hearing was reviewed in part; the course does not claim a complete transcript read. ### Which later admission qualifies the claim that Buffett always acted promptly on personnel problems? In 2025 he said he and Munger had failed several times to act when a once-excellent CEO developed serious cognitive impairment. Sources: - Warren E. Buffett, Thanksgiving Message to Fellow Shareholders (November 10, 2025), Complete message, including succession, philanthropy, delayed personnel action and final reflections. https://berkshirehathaway.com/news/nov1025.pdf Evidence note: Latest Buffett-authored material located on Berkshire's official site as of September 9, 2026. Read in full. ### Which three Dexter decisions did Buffett separate by the 2001 reporting year? Buying the company, paying with Berkshire stock and delaying operational changes. Sources: - Warren E. Buffett, 2001 Chairman's Letter (2001 reporting period; published 2002), Non-insurance operating highlights, shoe operations and Dexter. https://www.berkshirehathaway.com/2001ar/2001letter.html Evidence note: Retrospective self-assessment identifying the purchase, use of stock, and delayed operational response as three separate errors. ### What changed between the 1998 and 2002 accounts of General Re? The first praised the business; the later account identified weak reserving, dangerous aggregation and problems Buffett said he should have detected before the merger. Sources: - Warren E. Buffett, 1998 Chairman's Letter (1998 reporting year; signed March 1, 1999), General Re merger, share issuance and intrinsic value versus book value. https://www.berkshirehathaway.com/letters/1998pdf.pdf Evidence note: Contemporary favorable account of the General Re acquisition. Later letters identify underwriting, reserving and acquisition-currency errors that were not visible here. - Warren E. Buffett, 2002 Chairman's Letter (2002 reporting year; signed February 21, 2003), General Re reserving, aggregation risk and derivatives. https://www.berkshirehathaway.com/letters/2002pdf.pdf Evidence note: Follow-up after the 2001 letter. It documents another year of reserve corrections and Buffett's account of what he failed to detect before the General Re merger. ### Why must acquisition currency be judged separately from the acquired company? Issuing undervalued buyer shares can transfer a growing portion of every future winner even if the acquired business recovers. Sources: - Warren E. Buffett, 2016 Chairman's Letter (2016 reporting year; signed February 25, 2017), Dexter and General Re acquisition currency, BNSF share issuance and future returns. https://www.berkshirehathaway.com/letters/2016ltr.pdf Evidence note: Later comparison of three stock-funded acquisitions, showing why the instrument used to pay can be a separate decision from the asset purchased. ### What pressure made the 1999 competence boundary difficult to maintain? Berkshire visibly lagged a rapidly rising market while technology businesses attracted capital and public confidence. Sources: - Warren E. Buffett; edited by Carol Loomis, Mr. Buffett on the Stock Market (November 22, 1999), Interest rates, corporate profits and the two seventeen-year market periods. https://fortune.com/article/warren-buffett-on-stock-market/ Evidence note: Original publisher page. It records Buffett's late-1999 valuation argument and should not be rewritten as a precise forecast of the market's turning date. - Warren E. Buffett, 2000 Chairman's Letter (2000 reporting year; signed February 28, 2001), Speculation, technology businesses and limits of competence. https://www.berkshirehathaway.com/2000ar/2000letter.html Evidence note: Contemporary explanation of why Berkshire avoided most technology investments. It also admits earlier failures to understand several supposedly familiar industries. ### What is omitted from a victory story about avoiding the technology bubble? The lasting technology winners Berkshire also missed must be counted alongside the failures it avoided. Sources: - Warren E. Buffett, 2000 Chairman's Letter (2000 reporting year; signed February 28, 2001), Speculation, technology businesses and limits of competence. https://www.berkshirehathaway.com/2000ar/2000letter.html Evidence note: Contemporary explanation of why Berkshire avoided most technology investments. It also admits earlier failures to understand several supposedly familiar industries. ### How does chosen patience differ from forced patience? Chosen patience preserves a supported thesis with ample financing; forced patience can result from contracts, poor diligence or an expensive exit. Sources: - Warren E. Buffett, 2002 Chairman's Letter (2002 reporting year; signed February 21, 2003), General Re reserving, aggregation risk and derivatives. https://www.berkshirehathaway.com/letters/2002pdf.pdf Evidence note: Follow-up after the 2001 letter. It documents another year of reserve corrections and Buffett's account of what he failed to detect before the General Re merger. ### What converted Buffett's 2008 confidence into deployable action? Cash, recurring operating earnings, limited parent obligations and insurance exposure controlled before the panic. Sources: - Warren E. Buffett, 2008 Chairman's Letter (2008 reporting year; signed February 27, 2009), Financial panic, government action, liquidity and investments made during disarray. https://www.berkshirehathaway.com/letters/2008ltr.pdf Evidence note: Contemporary report after Berkshire's worst annual decline in per-share book value to that date. - Warren E. Buffett, 2009 Chairman's Letter (2009 reporting year; signed February 26, 2010), BNSF acquisition, stock consideration and liquidity, printed pages 15-17; signed February 26, 2010. https://www.berkshirehathaway.com/letters/2009ltr.pdf Evidence note: Contemporary account of the cash Berkshire deployed in 2008-2009 and the tradeoff created by using Berkshire shares for part of BNSF. ### Which 2008 error prevents a blanket claim that Buffett got the crisis right? He admitted buying ConocoPhillips near the peak in oil and gas prices, showing that the broad posture did not make every investment correct. Sources: - Warren E. Buffett, 2008 Chairman's Letter (2008 reporting year; signed February 27, 2009), Financial panic, government action, liquidity and investments made during disarray. https://www.berkshirehathaway.com/letters/2008ltr.pdf Evidence note: Contemporary report after Berkshire's worst annual decline in per-share book value to that date. ### Why did BNSF revise the simple lesson that capital-intensive businesses are unattractive? Essential demand and regulated returns could make decades of heavy reinvestment productive, unlike copyable textile investment. Sources: - Warren E. Buffett, 2009 Chairman's Letter (2009 reporting year; signed February 26, 2010), BNSF acquisition, stock consideration and liquidity, printed pages 15-17; signed February 26, 2010. https://www.berkshirehathaway.com/letters/2009ltr.pdf Evidence note: Contemporary account of the cash Berkshire deployed in 2008-2009 and the tradeoff created by using Berkshire shares for part of BNSF. ### What exactly did Buffett say was wrong about Precision Castparts? He remained positive about returns on tangible assets but admitted that he overestimated normalized earnings and therefore paid too much. Sources: - Warren E. Buffett, 2020 Chairman's Letter (2020 reporting year; signed February 27, 2021), Precision Castparts $11 billion write-down and Buffett responsibility statement, printed pages 3-4. https://www.berkshirehathaway.com/letters/2020ltr.pdf Evidence note: Buffett attributes the Precision Castparts overpayment to his own optimism rather than deception by the seller. ### Did the pandemic create the Precision Castparts acquisition error? No. It exposed the thin margin around the normalized-earnings estimate made when Berkshire bought the company in 2016. Sources: - Warren E. Buffett, 2015 Chairman's Letter (2015 reporting year; signed February 27, 2016), Precision Castparts acquisition, printed pages 4-5; BHE regulatory assumptions, printed pages 12-13. https://www.berkshirehathaway.com/letters/2015ltr.pdf Evidence note: Contemporary enthusiastic account of Precision Castparts before the later impairment. - Warren E. Buffett, 2020 Chairman's Letter (2020 reporting year; signed February 27, 2021), Precision Castparts $11 billion write-down and Buffett responsibility statement, printed pages 3-4. https://www.berkshirehathaway.com/letters/2020ltr.pdf Evidence note: Buffett attributes the Precision Castparts overpayment to his own optimism rather than deception by the seller. ### What organizational revision did the Apple position represent? Berkshire could obtain a growing share of an exceptional business without controlling it, while preserving more flexibility than a wholly owned acquisition. Sources: - Warren E. Buffett, 2020 Chairman's Letter (2020 reporting year; signed February 27, 2021), Precision Castparts $11 billion write-down and Buffett responsibility statement, printed pages 3-4. https://www.berkshirehathaway.com/letters/2020ltr.pdf Evidence note: Buffett attributes the Precision Castparts overpayment to his own optimism rather than deception by the seller. ### What did the 2014 Tesco exit add to Buffett’s development record? It showed that recognizing deterioration and holding a liquid security did not guarantee prompt action. Buffett had begun selling in 2013, but his slower exit ended with a $444 million after-tax loss in 2014. Sources: - Warren E. Buffett, 2014 Chairman's Letter (2014 reporting year; signed February 28, 2015), Tesco exit and delayed response, printed pages 17-18; Buffett retrospective, printed pages 24-37. https://www.berkshirehathaway.com/letters/2014ltr.pdf Evidence note: The Tesco discussion is a contemporary admission about the 2012-2014 holding and exit. The fifty-year review later in the same document is retrospective. ### What does Buffett's dozen-decisions account reveal and conceal? It reveals winner asymmetry and long holding periods while concealing repairs, controls, financing, errors and organizational work that kept Berkshire alive. Sources: - Warren E. Buffett, 2022 Chairman's Letter (2022 reporting year; signed February 25, 2023), A dozen good decisions, luck, Coca-Cola and American Express, and Berkshire's 1967 insurance pivot. https://www.berkshirehathaway.com/letters/2022ltr.pdf Evidence note: A late retrospective that deliberately compresses a long record. The course tests it against contemporary documents rather than accepting the compression as the whole story. ### Why should Munger's architect role be compared with the 1965-1976 reports? The tribute states the eventual direction, while the contemporary reports show continued textile hope, investment and delayed implementation. Sources: - Warren E. Buffett; signed by Berkshire officers as noted, Early Berkshire Hathaway annual reports (1965-1976 reporting years), Independent archive of hosted original-report facsimiles. Exact report locators used here: 1965 report, printed pages 1-2, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1965-Berkshire-AR.pdf; 1966 report, printed pages 1, 3-6, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1966-Berkshire-AR.pdf; 1967 report, printed pages 1, 3-4, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1967-Berkshire-AR.pdf; 1975 report, printed pages 1 and 7-8, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1975-Berkshire-AR.pdf; 1976 report, printed pages 1-2, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1976-Berkshire-AR.pdf. Berkshire’s official individual-letter archive begins at 1977.. https://theoraclesclassroom.com/archives/ Evidence note: These are independently hosted reproductions of the original reports, checked against the local original PDFs. The 1965-1969 reports are signed by Malcolm G. Chace Jr. and/or Kenneth V. Chace, while the local originals state that Buffett wrote the letter. The 1970-1976 reports are signed by Buffett. Berkshire’s official page supplies individual letters only from 1977 and a broader 1965-2024 compilation. - Warren E. Buffett, 2023 Chairman's Letter (2023 reporting year; signed February 24, 2024), BHE regulatory and wildfire reassessment, printed pages 13-14; Charlie Munger tribute, printed pages 5-6. https://www.berkshirehathaway.com/letters/2023ltr.pdf Evidence note: First annual letter after Munger's death. Buffett's account of their relationship is retrospective and affectionate; the contemporaneous letters show the revisions unfolding more gradually. ### Who wrote Berkshire's letter for the 2025 reporting year? Greg Abel. Buffett's latest located authored material is his November 10, 2025 Thanksgiving message. Sources: - Warren E. Buffett, Thanksgiving Message to Fellow Shareholders (November 10, 2025), Complete message, including succession, philanthropy, delayed personnel action and final reflections. https://berkshirehathaway.com/news/nov1025.pdf Evidence note: Latest Buffett-authored material located on Berkshire's official site as of September 9, 2026. Read in full. - Greg Abel, 2025 Berkshire Hathaway annual letter (2025 reporting year; published February 28, 2026), Opening, authorship, culture and stewardship. https://www.berkshirehathaway.com/letters/2025ltr.pdf Evidence note: This is Greg Abel's first annual letter as Berkshire CEO. It is successor evidence, not Buffett-authored material. ### Why should Bezos's regret-minimization story be separated from the 1994 decision record? It is a later interpretation told after Amazon succeeded. The event record separately shows a risky move, MacKenzie Bezos's support, parental and investor capital, an intentionally provisional plan, and operational corrections supplied by coworkers. Sources: - Jeffrey P. Bezos, Jeff Bezos interview (May 4, 2001), Full Academy of Achievement interview transcript. https://achievement.org/achiever/jeffrey-p-bezos/ Evidence note: Early retrospective account with specific credit to MacKenzie Bezos, Shel Kaphan, Bezos's parents, investors and early coworkers. Motivational claims remain his account. - Jeffrey P. Bezos, 2010 Baccalaureate Remarks (May 30, 2010), Complete Princeton transcript. https://www.princeton.edu/news/2010/05/30/2010-baccalaureate-remarks Evidence note: Later moral retelling of childhood and the decision to start Amazon. Use as retrospective framing rather than contemporaneous evidence of 1994 motives. ### What does the packing-table episode teach about founder development? Demand exposed a bad physical process. Bezos proposed knee pads, while a coworker proposed tables and roughly doubled productivity. Development meant accepting a better answer from the work, not defending the founder's first idea. Sources: - Jeffrey P. Bezos, Jeff Bezos interview (May 4, 2001), Full Academy of Achievement interview transcript. https://achievement.org/achiever/jeffrey-p-bezos/ Evidence note: Early retrospective account with specific credit to MacKenzie Bezos, Shel Kaphan, Bezos's parents, investors and early coworkers. Motivational claims remain his account. ### Which 1999 fact best signals capability debt rather than simple growth? Distribution capacity expanded from roughly 300,000 to more than 5 million square feet in less than 12 months while Amazon added many categories and geographies. Coordination commitments were growing as quickly as demand. Sources: - Jeffrey P. Bezos, 1999 Letter to Shareholders (1999 reporting period; published 2000), A Recap of 1999, Goals for 2000, distribution capacity and platform expansion. https://s2.q4cdn.com/299287126/files/doc_financials/annual/Shareholderletter99.pdf Evidence note: Contemporary management account of hypergrowth, category expansion and infrastructure burden. Later outcomes show that several named initiatives failed. ### Why are Auctions and zShops important to the 1999 platform story? They show that the contemporary portfolio contained weak launches as well as the seed of Marketplace. Later success should not make every early initiative look inevitable or equally well designed. Sources: - Jeffrey P. Bezos, 1999 Letter to Shareholders (1999 reporting period; published 2000), A Recap of 1999, Goals for 2000, distribution capacity and platform expansion. https://s2.q4cdn.com/299287126/files/doc_financials/annual/Shareholderletter99.pdf Evidence note: Contemporary management account of hypergrowth, category expansion and infrastructure burden. Later outcomes show that several named initiatives failed. - Jeffrey P. Bezos, 2014 Letter to Shareholders (2014 reporting period; published 2015), Opening business-quality test and the Marketplace, Prime, FBA and AWS sections. https://ir.aboutamazon.com/files/doc_financials/annual/AMAZON-2014-Shareholder-Letter.pdf Evidence note: Retrospective account of several initiatives after they had shown traction. Management's business-quality assessments and causal explanations remain judgments. ### What exactly did Bezos revise after the dot-com collapse? He retained the online-customer thesis but rejected the assumption that single-category e-commerce companies could quickly reach scale and profitability. Amazon stopped treating continued funding as automatic. Sources: - Jeffrey P. Bezos, 2000 Letter to Shareholders (2000 reporting period; published 2001), Opening share-price discussion, platform investments and Goal for 2001. https://s2.q4cdn.com/299287126/files/doc_financials/annual/00ar_letter.pdf Evidence note: Contemporary admission that the land-rush metaphor had understated the time and difficulty required for single-category companies to reach scale. ### Why is capital part of the persistence explanation for 2000 and 2001? Amazon entered the pressure with about $1.1 billion in cash and marketable securities after earlier financing. Cost reduction and operating improvement mattered, but resolve without a financial bridge would not have produced the same option to persist. Sources: - Jeffrey P. Bezos, 2000 Letter to Shareholders (2000 reporting period; published 2001), Opening share-price discussion, platform investments and Goal for 2001. https://s2.q4cdn.com/299287126/files/doc_financials/annual/00ar_letter.pdf Evidence note: Contemporary admission that the land-rush metaphor had understated the time and difficulty required for single-category companies to reach scale. - Jeffrey P. Bezos, 2001 Letter to Shareholders (2001 reporting period; published 2002), Opening discussion of price, cost reduction, growth, fixed costs and cash flow. https://ir.aboutamazon.com/files/doc_financials/annual/2001_shareholderLetter.pdf Evidence note: Management's contemporary explanation of a reinforcing operating cycle after reporting its first profitable quarter on the letter's pro forma measures. ### What is the difference between preserving a decision criterion and preserving a conclusion? A criterion such as differentiation can remain useful while the answer changes with facts. Bezos rejected physical stores in 2006, then accepted differentiated physical models through Whole Foods and Amazon Go by 2017. Sources: - Jeffrey P. Bezos, 2006 Letter to Shareholders (2006 reporting period; published 2007), Planting Seeds section and discussion of physical stores, FBA, AWS and new categories. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2006.PDF Evidence note: Contemporary new-business test. The rejection of physical stores records a position that later changed when management saw differentiation. - Jeffrey P. Bezos, 2017 Letter to Shareholders (2017 reporting period; published 2018), High standards, operating standards admission, six-page memos and Whole Foods. https://s2.q4cdn.com/299287126/files/doc_financials/annual/Amazon_Shareholder_Letter.pdf Evidence note: Bezos states that colleagues taught him operating standards he initially lacked. The Whole Foods acquisition also revises his 2006 physical-store position. ### What prevents a Bezos-style decision framework from becoming a winner-only retrospective? Record the thesis, expected cash path, burden, loss limit and review trigger before the result. Then preserve failures and abandoned work in the same denominator as Marketplace, Prime and AWS. Sources: - Jeffrey P. Bezos, 2004 Letter to Shareholders (April 2005), Our Most Important Financial Measure, Free Cash Flow Per Share, including the transportation-machine example. https://ir.aboutamazon.com/files/doc_financials/annual/2004_Annual_report.pdf Evidence note: Uses a deliberately simplified fictional machine to separate earnings growth from cash generation, followed by Amazon's 2004 company figures. - Jeffrey P. Bezos, 2005 Letter to Shareholders (2005 reporting period; published 2006), Opening discussion of math-based and judgment-based decisions, price reductions, and the single detail page. https://ir.aboutamazon.com/files/doc_financials/annual/AMZN2005AnnualReport.pdf Evidence note: Contemporary account of Amazon's decision policy with retrospective results for the Marketplace choice made in 2000. Long-run price effects were explicitly not measurable in advance. - Jeffrey P. Bezos, 2006 Letter to Shareholders (2006 reporting period; published 2007), Planting Seeds section and discussion of physical stores, FBA, AWS and new categories. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2006.PDF Evidence note: Contemporary new-business test. The rejection of physical stores records a position that later changed when management saw differentiation. ### What did Kindle's 5.5-hour sellout reveal after more than three years of development? Demand exceeded launch inventory, so supply and demand forecasting became immediate constraints. The letter does not isolate product appeal from the amount initially stocked. Long preparation did not eliminate market uncertainty. Sources: - Jeffrey P. Bezos, 2007 Letter to Shareholders (2007 reporting period; published 2008), Kindle development, launch demand and supply response. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2007letter.pdf Evidence note: Contemporary account of a product developed for more than three years and then constrained by demand that exhausted launch inventory in 5.5 hours. ### What evidence complicates a pure skill explanation for AWS? Amazon built reusable infrastructure from internal constraints and invested for years, but Bezos later called the long period without a comparable competitor exceptional luck. Market timing belongs in the causal account. Sources: - Jeffrey P. Bezos, 2010 Letter to Shareholders (2010 reporting period; published 2011), Service architecture, technology examples and connection to free cash flow. https://s2.q4cdn.com/299287126/files/doc_financials/annual/117006_ltr_ltr2.pdf Evidence note: Management account of technical capabilities developed under operating scale. It links engineering work to cash flow without isolating causal contribution. - Jeffrey P. Bezos and David M. Rubenstein, Interview at the Economic Club of Washington (September 13, 2018), Complete edited transcript. https://www.economicclub.org/sites/default/files/transcripts/Jeff_Bezos_Edited_Transcript.pdf Evidence note: Later retrospective account that distributes credit among mentors, colleagues and family, and identifies exceptional luck in AWS's competitive lead. ### How can a platform that removes gatekeepers become one? By concentrating customers, data, fulfilment and dispute control. Participants may gain access and sales while losing practical alternatives, which gives the platform power over fees, rankings and suspension. Sources: - Jeffrey P. Bezos, 2011 Letter to Shareholders (2011 reporting period; published 2012), Self-service platforms, AWS, Marketplace, KDP and FBA. https://s2.q4cdn.com/299287126/files/doc_financials/annual/letter.PDF Evidence note: Management's case that Amazon platforms let outside participants bypass older gatekeepers. Later institutional evidence tests Amazon's own gatekeeper role. - U.S. House Judiciary Subcommittee on Antitrust, Commercial and Administrative Law, Investigation of Competition in Digital Markets: Majority Staff Report and Recommendations (October 6, 2020), Amazon overview and selected marketplace, seller-data, fee, private-label, acquisition and logistics passages. https://www.govinfo.gov/content/pkg/GOVPUB-Y4_J89_1-PURL-gpo145949/pdf/GOVPUB-Y4_J89_1-PURL-gpo145949.pdf Evidence note: Read in part. The majority staff report draws on documents and market participants but is not a judicial finding and was not read outside the Amazon-relevant portions. ### What evidence is missing from seller success anecdotes? They need the distribution of seller outcomes, survival, fee burden, dependence, appeal results and multi-homing. A few large successes demonstrate possibility but not the typical bargain or bargaining power. Sources: - Jeffrey P. Bezos, 2011 Letter to Shareholders (2011 reporting period; published 2012), Self-service platforms, AWS, Marketplace, KDP and FBA. https://s2.q4cdn.com/299287126/files/doc_financials/annual/letter.PDF Evidence note: Management's case that Amazon platforms let outside participants bypass older gatekeepers. Later institutional evidence tests Amazon's own gatekeeper role. - U.S. House Judiciary Subcommittee on Antitrust, Commercial and Administrative Law, Investigation of Competition in Digital Markets: Majority Staff Report and Recommendations (October 6, 2020), Amazon overview and selected marketplace, seller-data, fee, private-label, acquisition and logistics passages. https://www.govinfo.gov/content/pkg/GOVPUB-Y4_J89_1-PURL-gpo145949/pdf/GOVPUB-Y4_J89_1-PURL-gpo145949.pdf Evidence note: Read in part. The majority staff report draws on documents and market participants but is not a judicial finding and was not read outside the Amazon-relevant portions. ### Why can a reversible product decision still impose irreversible costs? The launch can be reversed while chronic overwork, health damage, family disruption, talent loss or a defensive culture persist. Decision classification must include the execution burden as well as the commercial commitment. Sources: - Jeffrey P. Bezos, 2015 Letter to Shareholders (2015 reporting period; published 2016), Failure, long-tailed business returns, and Type 1 versus Type 2 decisions. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2015-Letter-to-Shareholders.PDF Evidence note: Retrospective management account. The ten-percent and one-hundred-times illustration explains an asymmetric payoff shape, not a measured base rate for Amazon experiments. - Jodi Kantor and David Streitfeld, Inside Amazon: Wrestling Big Ideas in a Bruising Workplace (August 15, 2015), Complete reported article, including Amazon's responses. https://www.nytimes.com/2015/08/16/technology/inside-amazon-wrestling-big-ideas-in-a-bruising-workplace.html Evidence note: Reporting based on more than 100 current and former employees. Accounts vary, but the article documents costs and contest around the culture Bezos praised. ### What would distinguish productive pressure from destructive pressure at Amazon? Track burden duration, consent, recovery, error, turnover, health and whether candid information improves. Product output and speed alone cannot show that the work system is sustainable or decision-enhancing. Sources: - Jeffrey P. Bezos, Amazon 2016 shareholder letter (2016 reporting year; published 2017), Resist Proxies and High-Velocity Decision Making.. https://www.aboutamazon.com/news/company-news/2016-letter-to-shareholders Evidence note: Use the signed Bezos letter, not the hosting page’s staff byline. Distinguish his reported practices from proof that they caused Amazon’s results. - Jodi Kantor and David Streitfeld, Inside Amazon: Wrestling Big Ideas in a Bruising Workplace (August 15, 2015), Complete reported article, including Amazon's responses. https://www.nytimes.com/2015/08/16/technology/inside-amazon-wrestling-big-ideas-in-a-bruising-workplace.html Evidence note: Reporting based on more than 100 current and former employees. Accounts vary, but the article documents costs and contest around the culture Bezos praised. ### Which admission most directly contradicts a lone-founder account of Amazon's operating excellence? Bezos says his early standards for operating processes were low and colleagues taught him what high standards looked like. That makes distributed teaching part of the causal history. Sources: - Jeffrey P. Bezos, 2017 Letter to Shareholders (2017 reporting period; published 2018), High standards, operating standards admission, six-page memos and Whole Foods. https://s2.q4cdn.com/299287126/files/doc_financials/annual/Amazon_Shareholder_Letter.pdf Evidence note: Bezos states that colleagues taught him operating standards he initially lacked. The Whole Foods acquisition also revises his 2006 physical-store position. ### When does capability retention turn a product failure into useful learning? After the failed product is actually closed, the company must identify people or technology that serve a new tested need. Bezos says Fire Phone teams and knowledge moved into Echo and Alexa. That does not recover every dollar lost, but it shows a specific transfer. Sources: - Jeffrey P. Bezos, 2018 Letter to Shareholders (2018 reporting period; published 2019), Third-party sales, wandering, Fire Phone transfer, wage challenge and employee programs. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2018-Letter-to-Shareholders.pdf Evidence note: Retrospective comparison of first-party and third-party growth, plus management's account of retaining people and capability after the Fire Phone failure. ### What does Amazon’s 2018 US minimum-wage increase to $15 per hour establish, and what does it leave unresolved? It establishes a defined pay increase for a large employee group. It does not measure pace, injury, schedule control, turnover, surveillance or worker voice, so those outcomes require separate evidence. Sources: - Jeffrey P. Bezos, 2018 Letter to Shareholders (2018 reporting period; published 2019), Third-party sales, wandering, Fire Phone transfer, wage challenge and employee programs. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2018-Letter-to-Shareholders.pdf Evidence note: Retrospective comparison of first-party and third-party growth, plus management's account of retaining people and capability after the Fire Phone failure. - Jodi Kantor and David Streitfeld, Inside Amazon: Wrestling Big Ideas in a Bruising Workplace (August 15, 2015), Complete reported article, including Amazon's responses. https://www.nytimes.com/2015/08/16/technology/inside-amazon-wrestling-big-ideas-in-a-bruising-workplace.html Evidence note: Reporting based on more than 100 current and former employees. Accounts vary, but the article documents costs and contest around the culture Bezos praised. ### Why must Amazon's reported pandemic actions carry a period label? The 2019 letter was published during rapidly changing conditions in April 2020. Its process changes and spending estimate describe an early response, not protection across every facility or the full pandemic. Sources: - Jeffrey P. Bezos, 2019 Letter to Shareholders (2019 reporting period; published April 2020), COVID-19 response, testing work, temporary pay, jobs and climate commitments. https://s2.q4cdn.com/299287126/files/doc_financials/2020/ar/2019-Shareholder-Letter.pdf Evidence note: Management's early-pandemic account, written while operating conditions were changing quickly. It does not independently assess worker experience or adequacy of protection. ### Why does Amazon's scale change the governance test for long-term investment? A startup's investment chiefly risks its capital and survival. A dominant platform can also alter seller access, rival viability and market rules. Long horizons then require conflict, data and competition safeguards. Sources: - Jeffrey P. Bezos, Written Testimony before the House Judiciary Subcommittee (July 29, 2020), Complete written testimony. https://docs.house.gov/meetings/JU/JU05/20200729/110883/HHRG-116-JU05-Wstate-BezosJ-20200729.pdf Evidence note: Founder narrative delivered under antitrust scrutiny. It supplies specific claims but is advocacy, not an independent assessment of Amazon's market power. - U.S. House Judiciary Subcommittee on Antitrust, Commercial and Administrative Law, Investigation of Competition in Digital Markets: Majority Staff Report and Recommendations (October 6, 2020), Amazon overview and selected marketplace, seller-data, fee, private-label, acquisition and logistics passages. https://www.govinfo.gov/content/pkg/GOVPUB-Y4_J89_1-PURL-gpo145949/pdf/GOVPUB-Y4_J89_1-PURL-gpo145949.pdf Evidence note: Read in part. The majority staff report draws on documents and market participants but is not a judicial finding and was not read outside the Amazon-relevant portions. ### Why do large seller counts and sales totals fail to settle the Marketplace welfare question? They omit distribution, fees, survival, alternatives and dependence. The testimony reports access and sales, while the House report describes bargaining and data concerns. The assessment needs both benefit and power measures. Sources: - Jeffrey P. Bezos, Written Testimony before the House Judiciary Subcommittee (July 29, 2020), Complete written testimony. https://docs.house.gov/meetings/JU/JU05/20200729/110883/HHRG-116-JU05-Wstate-BezosJ-20200729.pdf Evidence note: Founder narrative delivered under antitrust scrutiny. It supplies specific claims but is advocacy, not an independent assessment of Amazon's market power. - U.S. House Judiciary Subcommittee on Antitrust, Commercial and Administrative Law, Investigation of Competition in Digital Markets: Majority Staff Report and Recommendations (October 6, 2020), Amazon overview and selected marketplace, seller-data, fee, private-label, acquisition and logistics passages. https://www.govinfo.gov/content/pkg/GOVPUB-Y4_J89_1-PURL-gpo145949/pdf/GOVPUB-Y4_J89_1-PURL-gpo145949.pdf Evidence note: Read in part. The majority staff report draws on documents and market participants but is not a judicial finding and was not read outside the Amazon-relevant portions. ### What is the evidence status of Bezos's best-employer and safest-workplace commitment? It is recognition plus a promised response. Safety spending, staff and rotation plans are inputs. Injury, turnover, worker-voice and independent findings after the commitment are needed to judge outcomes. Sources: - Jeffrey P. Bezos, 2020 Letter to Shareholders (2020 reporting period; published April 2021), Stakeholder value estimates, Bessemer discussion, safety commitments and final CEO reflection. https://s2.q4cdn.com/299287126/files/doc_financials/2021/ar/Amazon-2020-Shareholder-Letter-and-1997-Shareholder-Letter.pdf Evidence note: Bezos's final CEO letter. It responds to worker criticism and makes new employee and safety commitments while acknowledging uncertainty in some stakeholder-value estimates. - National Labor Relations Board, Amazon.com Services LLC, Case 10-RC-269250 (2021), Official case page, initial tally and election-result records. https://www.nlrb.gov/case/10-RC-269250 Evidence note: Read in part. The official case and tally pages were reviewed, but the complete docket and later regional decision were not read in full. ### Why should results in the 2021 through 2024 letters not be attributed to Bezos? Andy Jassy signed the current-period letters. An appended reprint of Bezos's 1997 letter does not change that authorship. Later outcomes are succession evidence and should not automatically be attributed to Bezos. Sources: - Andy Jassy, 2021 Letter to Shareholders (2021 reporting period; published 2022), Author byline, closing signature and appended 1997-letter label. https://www.aboutamazon.com/news/company-news/2021-letter-to-shareholders Evidence note: The current-period letter is authored by Andy Jassy. The appended 1997 letter is separately authored by Jeffrey P. Bezos. Only authorship and the reprint boundary were checked for this source; no full-reading claim is made. - Andy Jassy, 2022 Letter to Shareholders (2022 reporting period; published 2023), Author byline, closing signature and appended 1997-letter label. https://www.aboutamazon.com/news/company-news/amazon-ceo-andy-jassy-2022-letter-to-shareholders Evidence note: The current-period letter is authored by Andy Jassy. The appended 1997 letter is separately authored by Jeffrey P. Bezos. Only authorship and the reprint boundary were checked for this source; no full-reading claim is made. - Andy Jassy, 2023 Letter to Shareholders (2023 reporting period; published 2024), Author byline, closing signature and appended 1997-letter label. https://www.aboutamazon.com/news/company-news/amazon-ceo-andy-jassy-2023-letter-to-shareholders Evidence note: The current-period letter is authored by Andy Jassy. The appended 1997 letter is separately authored by Jeffrey P. Bezos. Only authorship and the reprint boundary were checked for this source; no full-reading claim is made. - Andy Jassy, 2024 Letter to Shareholders (2024 reporting period; published 2025), Author byline, closing signature and appended 1997-letter label. https://www.aboutamazon.com/news/company-news/amazon-ceo-andy-jassy-2024-letter-to-shareholders Evidence note: The current-period letter is authored by Andy Jassy. The appended 1997 letter is separately authored by Jeffrey P. Bezos. Only authorship and the reprint boundary were checked for this source; no full-reading claim is made. ### What career event did Marks later call his lucky break? The 1978 reassignment from equities to convertibles and high yield bonds. It placed him in a stigmatized, inefficient market suited to his conservative temperament. Sources: - Howard Marks, Getting Lucky (2014-01-16), Complete memo, the Wharton, Chicago, 1978 high-yield reassignment, and decision-quality passages. https://www.oaktreecapital.com/docs/default-source/memos/2014-01-16-getting-lucky.pdf?sfvrsn=c4b70f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Who proposed the distressed debt fund, and why does that matter? Bruce Karsh proposed it. The fact qualifies a lone-genius story and shows Marks’s development depending on a partner with a more opportunistic idea. Sources: - Howard Marks, Bruce Karsh, and Sheldon Stone, 30 Years of Oaktree (2025-04-30), Official eight-page transcript, pp. 1-3 on formation and early funds, pp. 4-6 on partnership and operating choices, and pp. 7-8 on succession. https://www.oaktreecapital.com/insights/insight-podcast/education/30-years-of-oaktree-with-howard-marks-bruce-karsh-and-sheldon-stone Evidence note: Read in full. The founders reconstruct their partnership, early funds, culture, mistakes avoided and succession aims. ### What two observations prompted the first memo? A pension fund compounded a fourth-percentile long record without spectacular years, while a value manager excused a disastrous year as the cost of aiming for the top. Sources: - Howard Marks, The Route to Performance (1990-10-12), Printed pp. 1-2, the fourteen-year pension-plan record and the concluding avoidance-of-losers passage. https://www.oaktreecapital.com/docs/default-source/memos/1990-10-12-the-route-to-performance.pdf?sfvrsn=33bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Why must a useful non-consensus forecast clear more than one hurdle? It must differ from the price-setting consensus, be correct, matter on an investable horizon and improve a decision after costs. Sources: - Howard Marks, The Value of Predictions, or Where'd All This Rain Come From? (1993-02-15), Complete memo, the seven-step forecasting chain and the closing tests for whether a forecast is actionable. https://www.oaktreecapital.com/docs/default-source/memos/1993-02-15-the-value-of-predictions-or-where-39-d-all-this-rain-come-from.pdf?sfvrsn=6fbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### What did Oaktree give up to protect its edge? It turned away mandates and kept some funds smaller, sacrificing fees and growth when too much capital could reduce prospective returns. Sources: - Howard Marks, Oaktree at Ten (2005-04-11), Printed pp. 1-5, sections “Priorities,” “Clients,” “Performance,” “Growth,” and “Plans for the Future”. https://www.oaktreecapital.com/docs/default-source/memos/2005-04-11-oaktree-at-ten.pdf?sfvrsn=5fbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Why were Oaktree’s principles available at formation? The five founders had already worked together for about nine years on average. They wrote down practices they had tested rather than inventing a creed for a new brand. Sources: - Howard Marks, Bruce Karsh, and Sheldon Stone, 30 Years of Oaktree (2025-04-30), Official eight-page transcript, pp. 1-3 on formation and early funds, pp. 4-6 on partnership and operating choices, and pp. 7-8 on succession. https://www.oaktreecapital.com/insights/insight-podcast/education/30-years-of-oaktree-with-howard-marks-bruce-karsh-and-sheldon-stone Evidence note: Read in full. The founders reconstruct their partnership, early funds, culture, mistakes avoided and succession aims. ### What is the cost of being early even when valuation is eventually right? The market can continue against the position, creating underperformance, client pressure and solvency risk before the thesis is vindicated. Sources: - Howard Marks, Are You An Investor or a Speculator? (1997-09-03), Complete memo, passages defining investment versus speculation and applying the distinction to technology stocks. https://www.oaktreecapital.com/docs/default-source/memos/are-you-an-investor-or-a-speculator.pdf?sfvrsn=1e37cf65_2 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, bubble.com (2000-01-02), Complete memo, the opening bubble diagnosis and the numbered discussion of what happened, why, and what follows. https://www.oaktreecapital.com/docs/default-source/memos/2000-01-02-bubble.pdf?sfvrsn=37bc0f65_5 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### When did “second-level thinking” become an explicit concept? Marks says he named and developed it in 2009 when Columbia asked for a sample chapter, even though related habits were present earlier. Sources: - Howard Marks, It’s not Easy (2015-09-09), Printed pp. 1-3, section “Second-Level Thinking” and the account of drafting the 2009 sample chapter. https://www.oaktreecapital.com/docs/default-source/memos/2015-09-09-its-not-easy.pdf?sfvrsn=47bb0f65_7 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### What did Marks admit on July 30, 2007? When he wrote two weeks earlier, he did not expect the credit cycle to turn before month-end. The catalyst and timing had surprised him. Sources: - Howard Marks, It’s All Good . . . Really? (2007-07-30), Complete memo, opening admission that the credit turn arrived sooner than expected and the revised assessment that follows. https://www.oaktreecapital.com/docs/default-source/memos/2007-07-30-its-all-good-really.pdf?sfvrsn=97bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### How did Oaktree prepare without knowing the catalyst? The 2023 retrospective says it raised an $11 billion reserve fund between January 2007 and March 2008. The 2007 originals document the declining standards and uncertainty about timing; they do not supply that amount and period. Sources: - Howard Marks, The Race to the Bottom (2007-02-14), Complete memo, opening covenant and leverage evidence and closing “race to the bottom” diagnosis. https://www.oaktreecapital.com/docs/default-source/memos/2007-02-14-the-race-to-the-bottom.pdf?sfvrsn=9bbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Taking the Temperature (2023-07-10), Printed p. 1, five-memo selection rule; pp. 3-4, reserve-fund deployment and $6 billion/$7.5 billion figures; pp. 8-11, method and limits. https://www.oaktreecapital.com/insights/memo/taking-the-temperature Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### What assumption allowed Oaktree to buy after Lehman? If the system ended, portfolio choices would matter little. If it survived, buying viable claims cheaply was the job. The assumption enabled action without claiming certainty. Sources: - Howard Marks, Nobody Knows (2008-09-19), Does the Market Know?; The End of the Financial System; final bear-market-stage discussion. https://www.oaktreecapital.com/insights/memo/nobody-knows Evidence note: Contemporaneous memo after Lehman's bankruptcy. It records a decision premise and a search for bargains, but no security-level trade list. ### What did The Long View admit? Marks saw the short credit cycle but failed to appreciate the larger structural picture and initially treated 2003-07 as a familiar cycle that was merely more extreme. Sources: - Howard Marks, The Long View (2009-01-09), Complete memo, “The Short View” and “The Long View” passages distinguishing the foreseen credit cycle from the missed systemic mechanism. https://www.oaktreecapital.com/docs/default-source/memos/2009-01-09-the-long-view.pdf?sfvrsn=c3bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Did Oaktree avoid the leverage problem entirely in 2008? No. Marks disclosed that four evergreen funds used leverage and suffered when loan and equity prices collapsed. He said Oaktree was not perfect and apologized if he appeared holier than thou. Sources: - Howard Marks, Volatility + Leverage = Dynamite (2008-12-17), Printed pp. 10-12, “Are You Tall Enough to Use Leverage?” and Oaktree’s disclosure of four affected evergreen funds. https://www.oaktreecapital.com/docs/default-source/memos/2008-12-17-volatility-leverage-dynamite.pdf?sfvrsn=c7bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### What was Marks’s preferred working title for the cycle book? Listening to the Cycle. The publisher preferred Mastering the Market Cycle because it sounded more saleable. Sources: - Howard Marks, Conversation at Panmure House (2022-06-23), Complete interview transcript, discussion of cycle causality and the “Listening to the Cycle” working title. https://www.oaktreecapital.com/insights/memo/conversation-at-panmure-house Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### What idea arrived during drafting of Mastering the Market Cycle? Marks shifted from cycles as recurring ups and downs to causal sequences of excesses and corrections. Sources: - Howard Marks, Conversation at Panmure House (2022-06-23), Complete interview transcript, discussion of cycle causality and the “Listening to the Cycle” working title. https://www.oaktreecapital.com/insights/memo/conversation-at-panmure-house Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### What early pandemic guess did Marks later events overturn? On March 3, 2020, he guessed coronavirus would resemble another seasonal disease and would not fundamentally alter life or business. By March 19, his update described far more severe consequences. Sources: - Howard Marks, Nobody Knows II (2020-03-03), Printed pp. 1-7, facts/inferences/guesses framework; p. 6, the seasonal-disease guess; p. 7, partial buying response. https://www.oaktreecapital.com/docs/default-source/memos/nobody-knows-ii.pdf?sfvrsn=108eb165_8 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Latest Update (2020-03-19), Printed pp. 1-6, worsening health and economic update and concluding case for incremental buying rather than calling the bottom. https://www.oaktreecapital.com/docs/default-source/memos/weekly.pdf?sfvrsn=cbf3b065_8 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### How did Marks observe confirmation bias in himself? He found negative pandemic evidence easy to absorb and positive evidence easy to counter, because the negative case fit his wary disposition. Sources: - Howard Marks, Uncertainty (2020-05-11), All We Don't Know; In Praise of Doubt; confidence and position-sizing discussion. https://www.oaktreecapital.com/insights/memo/uncertainty Evidence note: The discussion of 2008 includes a retrospective practitioner report. The memo distinguishes evidence-based confidence from certainty. ### Why did 2020 produce less distressed opportunity than prior crises? The Fed and Treasury restored liquidity and confidence rapidly, preventing the prolonged credit freeze, forced selling and investor pain that had created earlier bargains. Sources: - Howard Marks, Coming into Focus (2020-10-13), The Power of Interest Rates; crisis comparison; final portfolio-calibration discussion. https://www.oaktreecapital.com/insights/memo/coming-into-focus Evidence note: Retrospective report on Oaktree's pre-crisis posture and March buying, followed by a dated October assessment. It does not disclose portfolio allocations. ### Which selling habit did Marks come to see as potentially life-altering error? Scaling out of a rare compounder merely because its price had risen or its multiple looked high. Sources: - Howard Marks, Something of Value (2021-01-11), Complete memo and appendix, sections on Marks’s personal value-investing journey and the dialogue with Andrew Marks. https://www.oaktreecapital.com/insights/memo/something-of-value Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Did Marks become bullish on cryptocurrency as he revised his view? No. In September 2017 he conceded he had framed Bitcoin incorrectly while still declining to invest. By 2021 he said his skepticism had not worked to date and he lacked enough knowledge for a firm view. Sources: - Howard Marks, Yet Again? (2017-09-07), Printed pp. 4-8, revised Bitcoin framing after conversations; pp. 9-11, response to criticism that the July memo lacked prescriptions. https://www.oaktreecapital.com/docs/default-source/memos/yet-again.pdf?sfvrsn=3767f765_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, Something of Value (2021-01-11), Complete memo and appendix, sections on Marks’s personal value-investing journey and the dialogue with Andrew Marks. https://www.oaktreecapital.com/insights/memo/something-of-value Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Why is the five-call set not a complete scorecard? Marks selected the memos partly because the calls turned out to be right, and the set omits the much larger denominator of public memos. Sources: - Howard Marks, Taking the Temperature (2023-07-10), Printed p. 1, five-memo selection rule; pp. 3-4, reserve-fund deployment and $6 billion/$7.5 billion figures; pp. 8-11, method and limits. https://www.oaktreecapital.com/insights/memo/taking-the-temperature Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### What is the fair claim supported by the five calls? Marks made a few useful posture decisions at extremes. The record does not show dependable routine market timing or complete foresight of crisis mechanisms. Sources: - Howard Marks, Taking the Temperature (2023-07-10), Printed p. 1, five-memo selection rule; pp. 3-4, reserve-fund deployment and $6 billion/$7.5 billion figures; pp. 8-11, method and limits. https://www.oaktreecapital.com/insights/memo/taking-the-temperature Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, The Long View (2009-01-09), Complete memo, “The Short View” and “The Long View” passages distinguishing the foreseen credit cycle from the missed systemic mechanism. https://www.oaktreecapital.com/docs/default-source/memos/2009-01-09-the-long-view.pdf?sfvrsn=c3bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### What did Marks admit about his caution during the long post-crisis recovery? In July 2017 he said caution since 2011 had not been right and that being six years early might never deserve to be called right. Sources: - Howard Marks, There They Go Again... Again (2017-07-26), Printed p. 1, admission that caution since 2011 had not been right; pp. 19-22, limits and conditional response. https://www.oaktreecapital.com/docs/default-source/memos/there-they-go-again-again.pdf?sfvrsn=56d4f265_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### What does Oaktree’s 2017 use of subscription lines reveal about pressure and revision? Oaktree adopted the financing after client requests and wider market use, then began an internal guidelines process after Marks raised concerns about performance presentation and liquidity risk. The memo records reconsideration, but not the final rules or their results. Sources: - Howard Marks, Lines in the Sand (2017-04-18), Printed p. 1, client pressure and Oaktree’s adoption of subscription lines; p. 7, the internal guidelines process and its unresolved status. https://www.oaktreecapital.com/docs/default-source/memos/lines-in-the-sand.pdf?sfvrsn=bf5dfa65_2 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### What part of Sea Change is least dependent on forecasting? The forty-year decline in rates and the 2022 shift from zero rates to rapid tightening are historical observations. Sources: - Howard Marks, Sea Change (2022-12-13), Complete memo, sections “Sea Change #1,” “Sea Change #2,” and the closing outlook and 2-to-4-percent range. https://www.oaktreecapital.com/insights/memo/sea-change Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### What numerical forecast did Marks label a guess in 2024? He guessed the federal funds rate might average 3.0 to 3.5 percent over the next five to ten years, starting in 2024. Sources: - Howard Marks, Easy Money (2024-01-09), Printed p. 1, ten years without reader response; p. 14, 3.0-to-3.5-percent guess over the next five to ten years. https://www.oaktreecapital.com/insights/memo/easy-money Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Does risk control mean avoiding every default? No. Marks says Oaktree had defaults in high yield nearly every year, but fewer than the market and fewer than the spread compensated it for. Sources: - Howard Marks, Fewer Losers, or More Winners? (September 12, 2023), Putting It in Brief; Not Risk Avoidance; The Role of Risk Bearing; What About in Practice?. https://www.oaktreecapital.com/insights/memo/fewer-losers-or-more-winners Evidence note: A conceptual and retrospective memo. It distinguishes bond upside, strategies needing winners, and risk control. It does not establish that a reader can identify investment bargains. ### What did Marks learn when he checked a familiar warning with Warren Buffett? Buffett told him he had not made the statement Marks repeatedly attributed to him. Marks published the correction. The admission documents the sourcing error, not the false quotation. Sources: - Howard Marks, Mr. Market Miscalculates (2024-08-22), Printed p. 5, correction after Buffett denied making the repeatedly attributed profit-growth warning. https://www.oaktreecapital.com/insights/memo/mr-market-miscalculates Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### How did Marks’s view of computers and investing change from 2018 to 2026? In 2018 he already accepted that computers could beat most investors, while provisionally reserving qualitative and long-horizon judgment for the best humans. After using Claude in 2026, he moved the boundary further and accepted that AI could perform whole knowledge-work tasks, while retaining doubts about novel situations, reliability, judgment and risk bearing. Sources: - Howard Marks, Investing Without People (2018-06-18), Printed p. 1, nonexpert caveat; pp. 12 and 14-17, Andrew’s challenges, limits of active management, provisional human advantages, and the corrected Einstein attribution. https://www.oaktreecapital.com/docs/default-source/memos/investing-without-people.pdf?sfvrsn=7a5ec465_8 Evidence note: Read directly in the official Oaktree complete collection or official memo page. - Howard Marks, AI Hurtles Ahead (2026-02-26), Official HTML, opening Claude tutorial and sections on investment implications, AI limitations, and risk posture. https://www.oaktreecapital.com/insights/memo/ai-hurtles-ahead Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Why should the 2026 private-credit thesis be tested carefully? Oaktree may benefit when competitors retreat, and the memo uses firm supplied exposure and performance claims. The incentive does not refute the analysis, but it matters. Sources: - Howard Marks, What’s Going on in Private Credit? (2026-04-09), Official HTML, sections “Direct Lending and Software,” “What Does the Market Know?,” and “What’s a Manager to Do?”; Oaktree disclosure gives 20-percent and 15-percent exposures. https://www.oaktreecapital.com/insights/memo/whats-going-on-in-private-credit Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### What cost of his own defensive stance did Marks acknowledge in 2001? He said his caution probably reduced profits in good years, that he disliked being cast as a bear, and that he and Oaktree might be less psychologically flexible than an idealized cycle timer. Sources: - Howard Marks, What Lies Ahead? (2001-10-04), Printed pp. 6-7, the “A Bear’s Eye View” response and closing discussion of caution, lost upside, and investor agility. https://www.oaktreecapital.com/docs/default-source/memos/2001-10-04-what-lies-ahead.pdf?sfvrsn=bbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Why is Leonard’s permanent-owner philosophy a development story rather than a founding assumption? His later accounts describe an early business sale he came to regret, financing whose dilution he reconsidered, and delegation forced by his own capacity limit. These are retrospective accounts and do not reveal every contemporary alternative. Sources: - Mark Leonard, Constellation Software Inc. 2014 President's Letter (2015-04-06), Invested capital; debt duration; ROIC; TSS contribution to maintenance revenue. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: Company-reported figures include management-defined non-IFRS measures explained in the letter's glossary. - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. - Mark Leonard and named Constellation respondents, November 5, 2018 · License revenue volatility, CSI’s biggest failure, Business systems (2018-11-05), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-nov-5-2018-final-1.pdf Evidence note: Full Q&A read. Leonard names excess early equity and a business sale as failures. Jamal explains license volatility. ### What could the same early financing teach correctly and incorrectly? It could correctly teach that unnecessary equity has a lasting cost, while incorrectly hardening into a refusal to issue equity under any circumstances. Leonard later questioned his own sensitivity to dilution. Sources: - Mark Leonard, Constellation Software Inc. 2014 President's Letter (2015-04-06), Invested capital; debt duration; ROIC; TSS contribution to maintenance revenue. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: Company-reported figures include management-defined non-IFRS measures explained in the letter's glossary. - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. - Mark Leonard and named Constellation respondents, November 5, 2018 · License revenue volatility, CSI’s biggest failure, Business systems (2018-11-05), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-nov-5-2018-final-1.pdf Evidence note: Full Q&A read. Leonard names excess early equity and a business sale as failures. Jamal explains license volatility. ### What changed between Leonard’s first and second quarterly letters of 2007? He moved from expecting initiative investment to self-correct toward concern that culling had overshot and too few new initiatives were emerging. He was willing to accept lower current margins for worthwhile growth. Sources: - Mark Leonard, Constellation Software 2007 Q1 president’s letter (2007-05-08), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q1_2007_shareholders_report.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2007 Q2 president’s letter (2007 Q2 reporting period; undated letter), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q2_2007_shareholders_report.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2007 Q3 president’s letter (2007 Q3 reporting period; undated letter), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q3_2007_shareholders_report.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. ### Why does a public growth target create an acquisition risk? It can make closing a deal valuable to the manager even when the deal’s economics are weak. The investment should still make sense without satisfying the target. Sources: - Mark Leonard, Constellation Software 2007 Q1 president’s letter (2007-05-08), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q1_2007_shareholders_report.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2007 Q2 president’s letter (2007 Q2 reporting period; undated letter), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q2_2007_shareholders_report.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2007 Q3 president’s letter (2007 Q3 reporting period; undated letter), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q3_2007_shareholders_report.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. ### Why is MAJES useful for studying Leonard without hindsight? The early letters disclose unresolved contracts and a gap between reported adjusted earnings and cash. They allow a decision under uncertainty but do not establish the completed investment return. Sources: - Mark Leonard, Constellation Software 2008 Q2 president’s letter (2008-08-07), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q2_2008_shareholdersreport.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2008 Q3 president’s letter (2008-11-06), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidentsletter50ed1eb201c049b287838ec5e943432b.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2008 Q4 president’s letter (2009-03-04), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q4_2008presidentsletter.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2009 Q1 president’s letter (2009-05-06), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q1_2009_presidentletter.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. ### Why can a recession thesis fail to produce acquisition opportunities? Owners may wait, lenders may avoid forced sales, and the desired businesses may not become available at financeable prices. A macro forecast is not a transaction pipeline. Sources: - Mark Leonard, Constellation Software 2008 Q2 president’s letter (2008-08-07), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q2_2008_shareholdersreport.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2008 Q3 president’s letter (2008-11-06), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidentsletter50ed1eb201c049b287838ec5e943432b.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2008 Q4 president’s letter (2009-03-04), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q4_2008presidentsletter.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2009 Q1 president’s letter (2009-05-06), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q1_2009_presidentletter.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. ### What must be true before amortization can be treated as economically unimportant? The underlying earning assets must retain their economic value. Maintenance health and customer retention are checks, not a guarantee that every acquired asset is sound. Sources: - Mark Leonard, Constellation Software 2007 Q4 president’s letter (2008-03-05), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q4_2007_shareholders_report.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2008 Q1 president’s letter (2008-05-07), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q1_2008_shareholdersreportf49d4032f35645668472bbd0e4d94b0b.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2009 annual president’s letter (2010-03-25), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/shareholder-letter-2009.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2010 annual president’s letter (2011-05-02), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/shareholder-letter-2010.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. ### Why should Constellation’s historical ROIC not be copied as a universal benchmark? The company defines its own adjusted income and invested-capital history. Its business mix, financing, growth and intangible-asset assumptions matter to the interpretation. Sources: - Mark Leonard, Constellation Software 2007 Q4 president’s letter (2008-03-05), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q4_2007_shareholders_report.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2008 Q1 president’s letter (2008-05-07), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q1_2008_shareholdersreportf49d4032f35645668472bbd0e4d94b0b.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2009 annual president’s letter (2010-03-25), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/shareholder-letter-2009.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2010 annual president’s letter (2011-05-02), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/shareholder-letter-2010.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. ### What belief did Leonard explicitly revise after the strategic review? He reconsidered the view that operating fundamentals alone would take care of the share price. He saw valuation and shareholder stability as capable of affecting the business itself. Sources: - Mark Leonard, Constellation Software 2010 annual president’s letter (2011-05-02), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/shareholder-letter-2010.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2011 annual president’s letter (2012-05-02), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/2011_presidents_letter.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. ### How can strong current profits signal weakened long-term incentives? Managers may defer initiatives, acquisitions or staffing when future rewards and ownership are uncertain. Lower investment can boost current profits while reducing future value. Sources: - Mark Leonard, Constellation Software 2010 annual president’s letter (2011-05-02), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/shareholder-letter-2010.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software 2011 annual president’s letter (2012-05-02), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/2011_presidents_letter.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. ### What surprised Leonard in the analyst-model exercise? A substantial part of modeled value depended on future acquisition activity. Existing assets and an ongoing ability to deploy capital were distinct sources of expected value. Sources: - Mark Leonard, Constellation Software 2012 annual president’s letter (2013-05-01), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-final.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software Inc. 2013 President's Letter (2014-04-30), Maintenance revenue; valuation sensitivity; TSS-scale acquisition; capital-funding discussion. https://www.csisoftware.com/wp-content/uploads/2026/04/presidentletter_2013.pdf Evidence note: Management report using company models and assumptions. Model outputs are not independently verified forecasts. - Mark Leonard, Constellation Software Inc. 2014 President's Letter (2015-04-06), Invested capital; debt duration; ROIC; TSS contribution to maintenance revenue. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: Company-reported figures include management-defined non-IFRS measures explained in the letter's glossary. ### Why can revisiting an earlier bad financing idea be rational? The instrument, cost, opportunity set and deployment capacity may have changed. The old error supplies a constraint to test, not a prohibition against every related future action. Sources: - Mark Leonard, Constellation Software 2012 annual president’s letter (2013-05-01), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-final.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Mark Leonard, Constellation Software Inc. 2013 President's Letter (2014-04-30), Maintenance revenue; valuation sensitivity; TSS-scale acquisition; capital-funding discussion. https://www.csisoftware.com/wp-content/uploads/2026/04/presidentletter_2013.pdf Evidence note: Management report using company models and assumptions. Model outputs are not independently verified forecasts. - Mark Leonard, Constellation Software Inc. 2014 President's Letter (2015-04-06), Invested capital; debt duration; ROIC; TSS contribution to maintenance revenue. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: Company-reported figures include management-defined non-IFRS measures explained in the letter's glossary. ### What makes Leonard’s decision to stop taking compensation context-dependent? He had significant ownership, accumulated wealth and experienced colleagues. The broader lesson is to redesign dependence on founder effort, not tell every founder to work for free. Sources: - Mark Leonard, Constellation Software Inc. 2014 President's Letter (2015-04-06), Invested capital; debt duration; ROIC; TSS contribution to maintenance revenue. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: Company-reported figures include management-defined non-IFRS measures explained in the letter's glossary. - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. ### Why might capping an incentive plan damage information quality? It can encourage moving revenue or profit between periods and weaken trust in long-term rewards. A superficially cheaper plan can make the reported information less useful. Sources: - Mark Leonard, Constellation Software Inc. 2014 President's Letter (2015-04-06), Invested capital; debt duration; ROIC; TSS contribution to maintenance revenue. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: Company-reported figures include management-defined non-IFRS measures explained in the letter's glossary. - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. ### What was the limitation in Leonard’s statistical case for small business units? The apparent advantage was weak and sensitive to removing a few outliers. He retained the preference as managerial judgment, not as a conclusively demonstrated causal effect. Sources: - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. - Mark Leonard, Constellation Software Inc. 2017 President's Letter (2018-04-20), Performance metrics; retained-capital policy; perpetual-owner objective; business-unit count. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-april-2018-final.pdf Evidence note: Management report. Statements about scalability and governance are Leonard's judgments, not independent causal findings. ### Why does acquisition experience not eliminate the need for review? Time spent on a target can increase commitment without improving the economics. Leonard described feeling that pressure himself; the review has to work when the experienced decision-maker is tempted. Sources: - Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. - Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. - Mark Leonard, Constellation Software Inc. 2017 President's Letter (2018-04-20), Performance metrics; retained-capital policy; perpetual-owner objective; business-unit count. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-april-2018-final.pdf Evidence note: Management report. Statements about scalability and governance are Leonard's judgments, not independent causal findings. ### How did the August 2018 Q&A qualify Leonard’s conglomerate study? He acknowledged that it had not formally studied failed conglomerates. The successful cases could suggest hypotheses but could not establish what distinguished success from failure. Sources: - Mark Leonard, Constellation Software Inc. 2017 President's Letter (2018-04-20), Performance metrics; retained-capital policy; perpetual-owner objective; business-unit count. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-april-2018-final.pdf Evidence note: Management report. Statements about scalability and governance are Leonard's judgments, not independent causal findings. - Mark Leonard and named Constellation respondents, August 3, 2018 · HPC’s, Blackboard, Customer Acquisition Economics, Margin Trajectory, Organic Growth Profile, Tax Rate, SaaS vs On-Premise (2018-08-03), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-august-2018-final.pdf Evidence note: Full Q&A read. Leonard admits selection bias in the conglomerate study; other responses cover margins, growth, taxes and SaaS. - Mark Leonard and named Constellation respondents, July 25, 2018 · Employee Retention, Consensus, IFTODH, Bias, Incentive Compensation (2018-07-25), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-july-25-2018-final.pdf Evidence note: Full Q&A read. Leonard discusses skeptics, formulaic incentives, downside borne by homebuilding managers and uncertainty in supposedly private facts. - Mark Leonard and named Constellation respondents, October 4, 2018 · Employee attraction, retention, motivation and engagement (2018-10-04), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-oct-4-2018-final.pdf Evidence note: Full Q&A read. Leonard questions engagement measurement and discusses opportunity, empathy and supervisor quality. - Mark Leonard and named Constellation respondents, October 9, 2018 · Moat, SaaS, TAM, corporate culture (2018-10-09), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-oct-9-2018-final-1.pdf Evidence note: Full Q&A read. Leonard discusses SaaS value capture and plural cultures. The club-software example belongs to Barry Symons. ### Why is board tenure not the same thing as expertise? Long service can produce contextual learning or entrenched assumptions. A credible tenure policy needs evidence of contribution, challenge and continuing competence. Sources: - Mark Leonard, Constellation Software Inc. 2017 President's Letter (2018-04-20), Performance metrics; retained-capital policy; perpetual-owner objective; business-unit count. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-april-2018-final.pdf Evidence note: Management report. Statements about scalability and governance are Leonard's judgments, not independent causal findings. - Mark Leonard and named Constellation respondents, August 3, 2018 · HPC’s, Blackboard, Customer Acquisition Economics, Margin Trajectory, Organic Growth Profile, Tax Rate, SaaS vs On-Premise (2018-08-03), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-august-2018-final.pdf Evidence note: Full Q&A read. Leonard admits selection bias in the conglomerate study; other responses cover margins, growth, taxes and SaaS. - Mark Leonard and named Constellation respondents, July 25, 2018 · Employee Retention, Consensus, IFTODH, Bias, Incentive Compensation (2018-07-25), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-july-25-2018-final.pdf Evidence note: Full Q&A read. Leonard discusses skeptics, formulaic incentives, downside borne by homebuilding managers and uncertainty in supposedly private facts. - Mark Leonard and named Constellation respondents, October 4, 2018 · Employee attraction, retention, motivation and engagement (2018-10-04), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-oct-4-2018-final.pdf Evidence note: Full Q&A read. Leonard questions engagement measurement and discusses opportunity, empathy and supervisor quality. - Mark Leonard and named Constellation respondents, October 9, 2018 · Moat, SaaS, TAM, corporate culture (2018-10-09), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-oct-9-2018-final-1.pdf Evidence note: Full Q&A read. Leonard discusses SaaS value capture and plural cultures. The club-software example belongs to Barry Symons. ### What is the important boundary in Leonard’s 2021 reversal? The traditional small- and medium-sized VMS acquisition hurdles remained. The shift concerned large transactions, broader capital deployment and distributions, not a blanket abandonment of standards. Sources: - Mark Leonard and named Constellation respondents, September 19, 2018 · Additional sectors for investment, Buybacks, ROIC, Value Investing (2018-09-19), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-september-2018-final.pdf Evidence note: Full Q&A read. Leonard qualifies his buyback position and distinguishes beliefs from certainty; Jamal supplies the invested-capital reconciliation. - Mark Leonard and named Constellation respondents, February 20, 2019 · Special Dividend (2019-02-20), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q-a-february-2019-divrelated-final.pdf Evidence note: Full Q&A read. Leonard defends the 2019 special dividend, magnetic hurdles and employee sharing of excess returns. - Mark Leonard, Constellation Software Inc. 2021 President's Letter (2021-02-15), Complete letter, especially dividend policy, large VMS acquisitions, and head-office role. https://www.csisoftware.com/wp-content/uploads/2026/04/letter-to-shareholders-february-15-2021.pdf Evidence note: Records a board policy decision and management intentions as of publication. The reported 16 percent invitation figure does not have an unambiguous denominator in the letter. It does not establish the results of later investments. - Jamal Baksh with Tegus, May 26, 2022 · CFO Interview (2022-04-06; posted May 26, 2022), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/April-6-2022-Tegus-interview-with-CFO.pdf Evidence note: Full seven-page company-reviewed interview read. Baksh is the speaker throughout; this is not Leonard’s testimony. It explains VMS Ventures, delegated acquisitions and limits to outside-sector investment. ### What does Baksh’s 2022 interview add to the 2021 policy announcement? It describes the difficulty of finding large opportunities outside VMS and cautions against assuming them in models. It is a separate executive’s account and keeps an intention from being mistaken for an achieved capability. Sources: - Mark Leonard and named Constellation respondents, September 19, 2018 · Additional sectors for investment, Buybacks, ROIC, Value Investing (2018-09-19), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-september-2018-final.pdf Evidence note: Full Q&A read. Leonard qualifies his buyback position and distinguishes beliefs from certainty; Jamal supplies the invested-capital reconciliation. - Mark Leonard and named Constellation respondents, February 20, 2019 · Special Dividend (2019-02-20), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q-a-february-2019-divrelated-final.pdf Evidence note: Full Q&A read. Leonard defends the 2019 special dividend, magnetic hurdles and employee sharing of excess returns. - Mark Leonard, Constellation Software Inc. 2021 President's Letter (2021-02-15), Complete letter, especially dividend policy, large VMS acquisitions, and head-office role. https://www.csisoftware.com/wp-content/uploads/2026/04/letter-to-shareholders-february-15-2021.pdf Evidence note: Records a board policy decision and management intentions as of publication. The reported 16 percent invitation figure does not have an unambiguous denominator in the letter. It does not establish the results of later investments. - Jamal Baksh with Tegus, May 26, 2022 · CFO Interview (2022-04-06; posted May 26, 2022), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/April-6-2022-Tegus-interview-with-CFO.pdf Evidence note: Full seven-page company-reviewed interview read. Baksh is the speaker throughout; this is not Leonard’s testimony. It explains VMS Ventures, delegated acquisitions and limits to outside-sector investment. ### Why does an ownership option’s exercise date matter to operating behavior? An approaching date can shorten the participant’s effective horizon and encourage short-term value drivers. Leonard’s 2021 Topicus answer presented a longer horizon as a way to improve alignment. Sources: - Mark Leonard, Constellation Software 2008 Q1 president’s letter (2008-05-07), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q1_2008_shareholdersreportf49d4032f35645668472bbd0e4d94b0b.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Constellation Software; named respondents vary, September 17, 2021 · Members Agreement, Acquisitions, NCI and Dividends (2021-09-17), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-september-2021-final.pdf Evidence note: Full Q&A read. Leonard discusses Topicus option horizons and judging managers through customer and employee records. Jamal answers accounting and tax questions. - Jamal Baksh with Tegus, May 26, 2022 · CFO Interview (2022-04-06; posted May 26, 2022), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/April-6-2022-Tegus-interview-with-CFO.pdf Evidence note: Full seven-page company-reviewed interview read. Baksh is the speaker throughout; this is not Leonard’s testimony. It explains VMS Ventures, delegated acquisitions and limits to outside-sector investment. - Constellation Software, Mark Leonard resignation and Mark Miller appointment (2025-09-25), Announcement body and separately attributed statements.. https://www.csisoftware.com/constellation-software-inc.-announces-the-resignation-of-mark-leonard-and-appointment-of-mark-miller-as-president-of-constellation-software/ Evidence note: Company announcement establishes the leadership change and stated health reason. Expressions of confidence are expectations, not proof of succession outcomes. - Constellation Software, Mark Leonard decision not to stand for board re-election (2026-03-27), Announcement body and separately attributed statements.. https://www.csisoftware.com/constellation-software-inc-announces-mark-leonards-decision-to-not-stand-for-re-election-to-board-of-directors/ Evidence note: Company announcement gives the end-of-term plan and continuing advisory role focused on PEMS. The company’s retrospective praise is not an independent assessment. ### What do succession announcements prove and leave unproved? They establish announced appointments, departures and stated plans. Expressions of confidence do not establish that capabilities will persist or that future performance will match the founder’s era. Sources: - Mark Leonard, Constellation Software 2008 Q1 president’s letter (2008-05-07), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q1_2008_shareholdersreportf49d4032f35645668472bbd0e4d94b0b.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. - Constellation Software; named respondents vary, September 17, 2021 · Members Agreement, Acquisitions, NCI and Dividends (2021-09-17), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-september-2021-final.pdf Evidence note: Full Q&A read. Leonard discusses Topicus option horizons and judging managers through customer and employee records. Jamal answers accounting and tax questions. - Jamal Baksh with Tegus, May 26, 2022 · CFO Interview (2022-04-06; posted May 26, 2022), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/April-6-2022-Tegus-interview-with-CFO.pdf Evidence note: Full seven-page company-reviewed interview read. Baksh is the speaker throughout; this is not Leonard’s testimony. It explains VMS Ventures, delegated acquisitions and limits to outside-sector investment. - Constellation Software, Mark Leonard resignation and Mark Miller appointment (2025-09-25), Announcement body and separately attributed statements.. https://www.csisoftware.com/constellation-software-inc.-announces-the-resignation-of-mark-leonard-and-appointment-of-mark-miller-as-president-of-constellation-software/ Evidence note: Company announcement establishes the leadership change and stated health reason. Expressions of confidence are expectations, not proof of succession outcomes. - Constellation Software, Mark Leonard decision not to stand for board re-election (2026-03-27), Announcement body and separately attributed statements.. https://www.csisoftware.com/constellation-software-inc-announces-mark-leonards-decision-to-not-stand-for-re-election-to-board-of-directors/ Evidence note: Company announcement gives the end-of-term plan and continuing advisory role focused on PEMS. The company’s retrospective praise is not an independent assessment. ### What separates a development dossier from a list of a thinker’s principles? A dossier records a dated belief, the pressure that challenged it, the response, consequences, contrary evidence and a transfer to changed conditions. It distinguishes contemporary evidence from later recollection. Sources: - Warren E. Buffett, 2014 Chairman's Letter (2014 reporting period; published 2015), Berkshire, Past, Present and Future, especially the textile, See's and Dexter passages. https://www.berkshirehathaway.com/letters/2014ltr.pdf Evidence note: Fifty-year retrospective. It supplies later cumulative figures and Buffett's revised interpretation, not a contemporaneous record of the earlier decisions. The Dexter share value is measured at the letter's February 2015 publication. - Jeffrey P. Bezos, 2015 Letter to Shareholders (2015 reporting period; published 2016), Failure, long-tailed business returns, and Type 1 versus Type 2 decisions. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2015-Letter-to-Shareholders.PDF Evidence note: Retrospective management account. The ten-percent and one-hundred-times illustration explains an asymmetric payoff shape, not a measured base rate for Amazon experiments. - Howard Marks, Fewer Losers, or More Winners? (September 12, 2023), Putting It in Brief; Not Risk Avoidance; The Role of Risk Bearing; What About in Practice?. https://www.oaktreecapital.com/insights/memo/fewer-losers-or-more-winners Evidence note: A conceptual and retrospective memo. It distinguishes bond upside, strategies needing winners, and risk control. It does not establish that a reader can identify investment bargains. - Mark Leonard, Constellation Software Inc. 2017 President's Letter (2018-04-20), Performance metrics; retained-capital policy; perpetual-owner objective; business-unit count. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-april-2018-final.pdf Evidence note: Management report. Statements about scalability and governance are Leonard's judgments, not independent causal findings. - Mark Leonard, Constellation Software Inc. 2021 President's Letter (2021-02-15), Complete letter, especially dividend policy, large VMS acquisitions, and head-office role. https://www.csisoftware.com/wp-content/uploads/2026/04/letter-to-shareholders-february-15-2021.pdf Evidence note: Records a board policy decision and management intentions as of publication. The reported 16 percent invitation figure does not have an unambiguous denominator in the letter. It does not establish the results of later investments. ### What makes a revision evidence of learning rather than hindsight? It preserves the original answer, identifies genuinely new evidence, explains the resulting change and keeps unresolved uncertainty visible. It does not merely choose the historical winner. Sources: - Warren E. Buffett, 2014 Chairman's Letter (2014 reporting period; published 2015), Berkshire, Past, Present and Future, especially the textile, See's and Dexter passages. https://www.berkshirehathaway.com/letters/2014ltr.pdf Evidence note: Fifty-year retrospective. It supplies later cumulative figures and Buffett's revised interpretation, not a contemporaneous record of the earlier decisions. The Dexter share value is measured at the letter's February 2015 publication. - Jeffrey P. Bezos, 2015 Letter to Shareholders (2015 reporting period; published 2016), Failure, long-tailed business returns, and Type 1 versus Type 2 decisions. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2015-Letter-to-Shareholders.PDF Evidence note: Retrospective management account. The ten-percent and one-hundred-times illustration explains an asymmetric payoff shape, not a measured base rate for Amazon experiments. - Howard Marks, Fewer Losers, or More Winners? (September 12, 2023), Putting It in Brief; Not Risk Avoidance; The Role of Risk Bearing; What About in Practice?. https://www.oaktreecapital.com/insights/memo/fewer-losers-or-more-winners Evidence note: A conceptual and retrospective memo. It distinguishes bond upside, strategies needing winners, and risk control. It does not establish that a reader can identify investment bargains. - Mark Leonard, Constellation Software Inc. 2017 President's Letter (2018-04-20), Performance metrics; retained-capital policy; perpetual-owner objective; business-unit count. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-april-2018-final.pdf Evidence note: Management report. Statements about scalability and governance are Leonard's judgments, not independent causal findings. - Mark Leonard, Constellation Software Inc. 2021 President's Letter (2021-02-15), Complete letter, especially dividend policy, large VMS acquisitions, and head-office role. https://www.csisoftware.com/wp-content/uploads/2026/04/letter-to-shareholders-february-15-2021.pdf Evidence note: Records a board policy decision and management intentions as of publication. The reported 16 percent invitation figure does not have an unambiguous denominator in the letter. It does not establish the results of later investments. ## Source directory ### Warren E. Buffett · Berkshire Hathaway 2012 shareholder letter Warren E. Buffett, Berkshire Hathaway 2012 shareholder letter (2012 reporting year; published 2013), Dividends, including reinvestment, acquisitions, repurchases, and the conditions behind the payout argument.. https://www.berkshirehathaway.com/letters/2012ltr.pdf Evidence note: Written by the allocator. His assessments of acquisition value and alternative outcomes are his judgments, not independent causal measurements. ### Jeffrey P. Bezos · Amazon 2016 shareholder letter Jeffrey P. Bezos, Amazon 2016 shareholder letter (2016 reporting year; published 2017), Resist Proxies and High-Velocity Decision Making.. https://www.aboutamazon.com/news/company-news/2016-letter-to-shareholders Evidence note: Use the signed Bezos letter, not the hosting page’s staff byline. Distinguish his reported practices from proof that they caused Amazon’s results. ### Mark Leonard · Constellation Software 2014 president’s letter Mark Leonard, Constellation Software 2014 president’s letter (2014 reporting year; published 2015), Pages 1–5, particularly intangible-asset condition, reliable capital, and business-unit decentralization.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: The URL reflects later file hosting, not a 2026 letter. Business figures in the original refer to their stated historical periods. Author explanations remain author explanations. ### Howard Marks · Fewer Losers, or More Winners? Howard Marks, Fewer Losers, or More Winners? (September 12, 2023), Putting It in Brief; Not Risk Avoidance; The Role of Risk Bearing; What About in Practice?. https://www.oaktreecapital.com/insights/memo/fewer-losers-or-more-winners Evidence note: A conceptual and retrospective memo. It distinguishes bond upside, strategies needing winners, and risk control. It does not establish that a reader can identify investment bargains. ### Warren E. Buffett · 1983 Chairman's Letter Warren E. Buffett, 1983 Chairman's Letter (1983 reporting period; published 1984), Goodwill and its Amortization, especially the See's comparison. https://www.berkshirehathaway.com/letters/1983.html Evidence note: Contemporary discussion of accounting and economic goodwill using historical See's figures. The acquisition rationale is partly retrospective. ### Warren E. Buffett · 1985 Chairman's Letter Warren E. Buffett, 1985 Chairman's Letter (March 4, 1986), Shutdown of Textile Business. https://www.berkshirehathaway.com/letters/1985.html Evidence note: Contemporary account written after the July 1985 closure decision. It reports Buffett's own diagnosis and hindsight about earlier delay. ### Warren E. Buffett · 1986 Chairman's Letter Warren E. Buffett, 1986 Chairman's Letter (1986 reporting period; published 1987), Purchase-Price Accounting Adjustments and the owner-earnings discussion near the end of the letter. https://www.berkshirehathaway.com/letters/1986.html Evidence note: Explains Buffett's valuation adjustment using Scott Fetzer. Maintenance capital expenditure is explicitly an estimate, not an audited line item. ### Warren E. Buffett · 1993 Chairman's Letter Warren E. Buffett, 1993 Chairman's Letter (1993 reporting period; published 1994), Dexter Shoe. https://www.berkshirehathaway.com/letters/1993.html Evidence note: Contemporary, favorable account after the November 7, 1993 merger. It is useful evidence of the case made then, not proof that all deliberations were disclosed. ### Warren E. Buffett · 2001 Chairman's Letter Warren E. Buffett, 2001 Chairman's Letter (2001 reporting period; published 2002), Non-insurance operating highlights, shoe operations and Dexter. https://www.berkshirehathaway.com/2001ar/2001letter.html Evidence note: Retrospective self-assessment identifying the purchase, use of stock, and delayed operational response as three separate errors. ### Warren E. Buffett · 2007 Chairman's Letter Warren E. Buffett, 2007 Chairman's Letter (2007 reporting period; published 2008), Businesses, The Great, the Good and the Gruesome. https://www.berkshirehathaway.com/letters/2007ltr.pdf Evidence note: Retrospective comparison of See's, FlightSafety, capital-intensive growth, and Dexter. The Dexter share value is measured at the letter's February 2008 publication, not in the 2007 reporting year or in 1993. ### Warren E. Buffett · 2014 Chairman's Letter Warren E. Buffett, 2014 Chairman's Letter (2014 reporting period; published 2015), Berkshire, Past, Present and Future, especially the textile, See's and Dexter passages. https://www.berkshirehathaway.com/letters/2014ltr.pdf Evidence note: Fifty-year retrospective. It supplies later cumulative figures and Buffett's revised interpretation, not a contemporaneous record of the earlier decisions. The Dexter share value is measured at the letter's February 2015 publication. ### Jeffrey P. Bezos · 1997 Letter to Shareholders Jeffrey P. Bezos, 1997 Letter to Shareholders (1997 reporting period; published 1998), It's All About the Long Term, Obsess Over Customers, Infrastructure, and Goals for 1998. https://www.aboutamazon.com/news/company-news/amazons-original-1997-letter-to-shareholders Evidence note: Contemporary statement of management policy and 1997 results. Management's claims about market leadership and future economics are its judgments at the time. ### Jeffrey P. Bezos · 1998 Letter to Shareholders Jeffrey P. Bezos, 1998 Letter to Shareholders (1998 reporting period; published 1999), A Recap of 1998, Our Customers, and Goals for 1999. https://s2.q4cdn.com/299287126/files/doc_financials/annual/Shareholderletter98.pdf Evidence note: Contemporary follow-up reporting customer, revenue, repeat-order, cash-flow, inventory and infrastructure figures. The outcome does not isolate the effect of the 1997 policy. ### Jeffrey P. Bezos · 2001 Letter to Shareholders Jeffrey P. Bezos, 2001 Letter to Shareholders (2001 reporting period; published 2002), Opening discussion of price, cost reduction, growth, fixed costs and cash flow. https://ir.aboutamazon.com/files/doc_financials/annual/2001_shareholderLetter.pdf Evidence note: Management's contemporary explanation of a reinforcing operating cycle after reporting its first profitable quarter on the letter's pro forma measures. ### Jeffrey P. Bezos · 2002 Letter to Shareholders Jeffrey P. Bezos, 2002 Letter to Shareholders (2002 reporting period; published 2003), Opening discussion of customer experience, price, satisfaction, the 100-book comparison, and free cash flow. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2002_shareholderLetter.pdf Evidence note: Contemporary company evidence. The 100-book basket, satisfaction score and management's free-cash-flow definition have specific scopes and do not prove a universal price rule. ### Jeffrey P. Bezos · 2004 Letter to Shareholders Jeffrey P. Bezos, 2004 Letter to Shareholders (April 2005), Our Most Important Financial Measure, Free Cash Flow Per Share, including the transportation-machine example. https://ir.aboutamazon.com/files/doc_financials/annual/2004_Annual_report.pdf Evidence note: Uses a deliberately simplified fictional machine to separate earnings growth from cash generation, followed by Amazon's 2004 company figures. ### Jeffrey P. Bezos · 2005 Letter to Shareholders Jeffrey P. Bezos, 2005 Letter to Shareholders (2005 reporting period; published 2006), Opening discussion of math-based and judgment-based decisions, price reductions, and the single detail page. https://ir.aboutamazon.com/files/doc_financials/annual/AMZN2005AnnualReport.pdf Evidence note: Contemporary account of Amazon's decision policy with retrospective results for the Marketplace choice made in 2000. Long-run price effects were explicitly not measurable in advance. ### Jeffrey P. Bezos · 2009 Letter to Shareholders Jeffrey P. Bezos, 2009 Letter to Shareholders (April 2010), Opening results and the 452 goals for 2010. https://s2.q4cdn.com/299287126/files/doc_financials/annual/AMZN_Shareholder-Letter-2009-(final).pdf Evidence note: Reports the content of one annual planning set. It shows Amazon's chosen management emphasis, not proof that financial outputs can be ignored. ### Jeffrey P. Bezos · 2014 Letter to Shareholders Jeffrey P. Bezos, 2014 Letter to Shareholders (2014 reporting period; published 2015), Opening business-quality test and the Marketplace, Prime, FBA and AWS sections. https://ir.aboutamazon.com/files/doc_financials/annual/AMAZON-2014-Shareholder-Letter.pdf Evidence note: Retrospective account of several initiatives after they had shown traction. Management's business-quality assessments and causal explanations remain judgments. ### Jeffrey P. Bezos · 2015 Letter to Shareholders Jeffrey P. Bezos, 2015 Letter to Shareholders (2015 reporting period; published 2016), Failure, long-tailed business returns, and Type 1 versus Type 2 decisions. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2015-Letter-to-Shareholders.PDF Evidence note: Retrospective management account. The ten-percent and one-hundred-times illustration explains an asymmetric payoff shape, not a measured base rate for Amazon experiments. ### Howard Marks · The Tide Goes Out Howard Marks, The Tide Goes Out (2008-03-18), Leveraging and Inflating; Mark-to-Market Accounting; Should Does Not Equal Will; Leverage and Risk. https://www.oaktreecapital.com/insights/memo/the-tide-goes-out Evidence note: Contemporaneous analysis written before the worst phase of the financial crisis. Examples and conclusions express the author's assessment and are not an independent performance study. ### Howard Marks · Nobody Knows Howard Marks, Nobody Knows (2008-09-19), Does the Market Know?; The End of the Financial System; final bear-market-stage discussion. https://www.oaktreecapital.com/insights/memo/nobody-knows Evidence note: Contemporaneous memo after Lehman's bankruptcy. It records a decision premise and a search for bargains, but no security-level trade list. ### Howard Marks · The Limits to Negativism Howard Marks, The Limits to Negativism (2008-10-15), Final discussion of third-stage pessimism and Oaktree purchases. https://www.oaktreecapital.com/insights/memo/the-limits-to-negativism Evidence note: Contemporaneous practitioner report. It describes purchase direction and averaging down, but not positions, allocations, or later investment results. ### Howard Marks · Which Way Now? Howard Marks, Which Way Now? (2020-03-31), The Positive Case; the negative case; The Government Programs; Summing Up. https://www.oaktreecapital.com/insights/memo/which-way-now Evidence note: Dated scenario analysis and author judgment during the pandemic shock. It is not an Oaktree transaction ledger. ### Howard Marks · Uncertainty Howard Marks, Uncertainty (2020-05-11), All We Don't Know; In Praise of Doubt; confidence and position-sizing discussion. https://www.oaktreecapital.com/insights/memo/uncertainty Evidence note: The discussion of 2008 includes a retrospective practitioner report. The memo distinguishes evidence-based confidence from certainty. ### Howard Marks · The Anatomy of a Rally Howard Marks, The Anatomy of a Rally (2020-06-18), Market chronology; psychology of the rally; final assessment questions. https://www.oaktreecapital.com/insights/memo/the-anatomy-of-a-rally Evidence note: Contemporaneous market assessment after the spring rally. The judgment is not evidence of later excess returns. ### Howard Marks · Coming into Focus Howard Marks, Coming into Focus (2020-10-13), The Power of Interest Rates; crisis comparison; final portfolio-calibration discussion. https://www.oaktreecapital.com/insights/memo/coming-into-focus Evidence note: Retrospective report on Oaktree's pre-crisis posture and March buying, followed by a dated October assessment. It does not disclose portfolio allocations. ### Mark Leonard · Constellation Software Inc. 2013 President's Letter Mark Leonard, Constellation Software Inc. 2013 President's Letter (2014-04-30), Maintenance revenue; valuation sensitivity; TSS-scale acquisition; capital-funding discussion. https://www.csisoftware.com/wp-content/uploads/2026/04/presidentletter_2013.pdf Evidence note: Management report using company models and assumptions. Model outputs are not independently verified forecasts. ### Mark Leonard · Constellation Software Inc. 2014 President's Letter Mark Leonard, Constellation Software Inc. 2014 President's Letter (2015-04-06), Invested capital; debt duration; ROIC; TSS contribution to maintenance revenue. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf Evidence note: Company-reported figures include management-defined non-IFRS measures explained in the letter's glossary. ### Mark Leonard · Constellation Software Inc. 2015 President's Letter Mark Leonard, Constellation Software Inc. 2015 President's Letter (2016-04-26), Invested Capital; Return on Invested Capital; Organic Net Revenue Growth; Combined Ratio. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf Evidence note: Management report that labels some comparisons as incomplete or tentative. Company IRR records are not public. ### Mark Leonard · Constellation Software Inc. 2016 President's Letter Mark Leonard, Constellation Software Inc. 2016 President's Letter (2017-04-25), Maintaining Investment Discipline; post-acquisition reviews; investment capacity; human-scale business units. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf Evidence note: Retrospective management account of a rejected transaction and internal processes. The target, exact hurdle, and forecast model are undisclosed. ### Mark Leonard · Constellation Software Inc. 2017 President's Letter Mark Leonard, Constellation Software Inc. 2017 President's Letter (2018-04-20), Performance metrics; retained-capital policy; perpetual-owner objective; business-unit count. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-april-2018-final.pdf Evidence note: Management report. Statements about scalability and governance are Leonard's judgments, not independent causal findings. ### Mark Leonard · Constellation Software Inc. 2021 President's Letter Mark Leonard, Constellation Software Inc. 2021 President's Letter (2021-02-15), Complete letter, especially dividend policy, large VMS acquisitions, and head-office role. https://www.csisoftware.com/wp-content/uploads/2026/04/letter-to-shareholders-february-15-2021.pdf Evidence note: Records a board policy decision and management intentions as of publication. The reported 16 percent invitation figure does not have an unambiguous denominator in the letter. It does not establish the results of later investments. ### Warren E. Buffett · Buffett Partnership letters compilation Warren E. Buffett, Buffett Partnership letters compilation (1957-1970), Complete 152-page compilation, especially the 1957 and 1958 letters, January 18, 1963 Ground Rules and Dempster discussion, January 25 and October 9, 1967 letters, January 22 and May 29, 1969 letters, and February 25, 1970 letter. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: A third-party-hosted compilation of reproduced original letters. It is not an official Berkshire archive. Datelines and signatures were checked in the documents. All 29 located documents dated 1957 through 1970 were read in full; the earlier 1956 original referenced in the 1957 letter was not present. ### Warren E. Buffett; signed by Berkshire officers as noted · Early Berkshire Hathaway annual reports Warren E. Buffett; signed by Berkshire officers as noted, Early Berkshire Hathaway annual reports (1965-1976 reporting years), Independent archive of hosted original-report facsimiles. Exact report locators used here: 1965 report, printed pages 1-2, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1965-Berkshire-AR.pdf; 1966 report, printed pages 1, 3-6, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1966-Berkshire-AR.pdf; 1967 report, printed pages 1, 3-4, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1967-Berkshire-AR.pdf; 1975 report, printed pages 1 and 7-8, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1975-Berkshire-AR.pdf; 1976 report, printed pages 1-2, https://theoraclesclassroom.com/wp-content/uploads/2019/09/1976-Berkshire-AR.pdf. Berkshire’s official individual-letter archive begins at 1977.. https://theoraclesclassroom.com/archives/ Evidence note: These are independently hosted reproductions of the original reports, checked against the local original PDFs. The 1965-1969 reports are signed by Malcolm G. Chace Jr. and/or Kenneth V. Chace, while the local originals state that Buffett wrote the letter. The 1970-1976 reports are signed by Buffett. Berkshire’s official page supplies individual letters only from 1977 and a broader 1965-2024 compilation. ### Warren E. Buffett · Berkshire Hathaway 1974 annual letter Warren E. Buffett, Berkshire Hathaway 1974 annual letter (1974 reporting year; signed March 31, 1975), Insurance underwriting, Home and Auto, liquidity and Blue Chip Stamps. https://www.berkshirehathaway.com/letters/letters.html Evidence note: Contemporary report during an insurance downturn. The full text was available locally; the public URL is Berkshire's archive page because it does not expose an individual 1974 file. ### Warren E. Buffett · How Inflation Swindles the Equity Investor Warren E. Buffett, How Inflation Swindles the Equity Investor (May 1, 1977), The sticky equity coupon, retained earnings and capital required merely to maintain physical output. https://fortune.com/article/buffett-how-inflation-swindles-the-equity-investor-fortune-classics-1977/ Evidence note: Original publisher page. The article develops an argument that was contemporaneous with Berkshire's capital-intensive textile experience. ### Warren E. Buffett · The Superinvestors of Graham-and-Doddsville Warren E. Buffett, The Superinvestors of Graham-and-Doddsville (May 17, 1984), Accessible article text and performance tables. https://business.columbia.edu/insights/chazen-global-insights/superinvestors-graham-and-doddsville Evidence note: Columbia Business School's publication of Buffett's argument against a pure-chance explanation for the records of value investors he had identified in advance. ### Warren E. Buffett · 1989 Chairman's Letter Warren E. Buffett, 1989 Chairman's Letter (1989 reporting year; signed March 2, 1990), Mistakes of the First Twenty-five Years and the institutional imperative. https://www.berkshirehathaway.com/letters/1989.html Evidence note: Retrospective classification of commission, omission and institutional errors after twenty-five years of Berkshire control. ### Warren E. Buffett · 1991 Chairman's Letter Warren E. Buffett, 1991 Chairman's Letter (1991 reporting year; signed February 28, 1992), Salomon, omission errors, USAir and fixed-income securities. https://www.berkshirehathaway.com/letters/1991.html Evidence note: Buffett's own report after taking the interim chair at Salomon. It does not replace the hearing record or independent reporting on the scandal. ### Warren E. Buffett · Salomon Brothers securities-trading investigation testimony Warren E. Buffett, Salomon Brothers securities-trading investigation testimony (September 4-5, 1991), Opening apology, cooperation commitments, compliance standard and response to lawmakers. https://www.youtube.com/watch?v=MtaeGt3KwuA Evidence note: Public recording of Buffett's testimony before a House subcommittee. The hearing was reviewed in part; the course does not claim a complete transcript read. ### Warren E. Buffett · 1998 Chairman's Letter Warren E. Buffett, 1998 Chairman's Letter (1998 reporting year; signed March 1, 1999), General Re merger, share issuance and intrinsic value versus book value. https://www.berkshirehathaway.com/letters/1998pdf.pdf Evidence note: Contemporary favorable account of the General Re acquisition. Later letters identify underwriting, reserving and acquisition-currency errors that were not visible here. ### Warren E. Buffett; edited by Carol Loomis · Mr. Buffett on the Stock Market Warren E. Buffett; edited by Carol Loomis, Mr. Buffett on the Stock Market (November 22, 1999), Interest rates, corporate profits and the two seventeen-year market periods. https://fortune.com/article/warren-buffett-on-stock-market/ Evidence note: Original publisher page. It records Buffett's late-1999 valuation argument and should not be rewritten as a precise forecast of the market's turning date. ### Warren E. Buffett · 2000 Chairman's Letter Warren E. Buffett, 2000 Chairman's Letter (2000 reporting year; signed February 28, 2001), Speculation, technology businesses and limits of competence. https://www.berkshirehathaway.com/2000ar/2000letter.html Evidence note: Contemporary explanation of why Berkshire avoided most technology investments. It also admits earlier failures to understand several supposedly familiar industries. ### Warren E. Buffett · 2002 Chairman's Letter Warren E. Buffett, 2002 Chairman's Letter (2002 reporting year; signed February 21, 2003), General Re reserving, aggregation risk and derivatives. https://www.berkshirehathaway.com/letters/2002pdf.pdf Evidence note: Follow-up after the 2001 letter. It documents another year of reserve corrections and Buffett's account of what he failed to detect before the General Re merger. ### Warren E. Buffett · 2008 Chairman's Letter Warren E. Buffett, 2008 Chairman's Letter (2008 reporting year; signed February 27, 2009), Financial panic, government action, liquidity and investments made during disarray. https://www.berkshirehathaway.com/letters/2008ltr.pdf Evidence note: Contemporary report after Berkshire's worst annual decline in per-share book value to that date. ### Warren E. Buffett · Buy American. I Am. Warren E. Buffett, Buy American. I Am. (October 17, 2008), Personal equity-allocation disclosure and the distinction between forecasting business conditions and buying productive assets. https://www.nytimes.com/2008/10/17/opinion/17buffett.html Evidence note: Original publisher page. Access restrictions prevented a fresh complete page extraction, so the work is not marked read in full. ### Warren E. Buffett · 2009 Chairman's Letter Warren E. Buffett, 2009 Chairman's Letter (2009 reporting year; signed February 26, 2010), BNSF acquisition, stock consideration and liquidity, printed pages 15-17; signed February 26, 2010. https://www.berkshirehathaway.com/letters/2009ltr.pdf Evidence note: Contemporary account of the cash Berkshire deployed in 2008-2009 and the tradeoff created by using Berkshire shares for part of BNSF. ### Warren E. Buffett · 2014 Chairman's Letter Warren E. Buffett, 2014 Chairman's Letter (2014 reporting year; signed February 28, 2015), Tesco exit and delayed response, printed pages 17-18; Buffett retrospective, printed pages 24-37. https://www.berkshirehathaway.com/letters/2014ltr.pdf Evidence note: The Tesco discussion is a contemporary admission about the 2012-2014 holding and exit. The fifty-year review later in the same document is retrospective. ### Warren E. Buffett · 2015 Chairman's Letter Warren E. Buffett, 2015 Chairman's Letter (2015 reporting year; signed February 27, 2016), Precision Castparts acquisition, printed pages 4-5; BHE regulatory assumptions, printed pages 12-13. https://www.berkshirehathaway.com/letters/2015ltr.pdf Evidence note: Contemporary enthusiastic account of Precision Castparts before the later impairment. ### Warren E. Buffett · 2016 Chairman's Letter Warren E. Buffett, 2016 Chairman's Letter (2016 reporting year; signed February 25, 2017), Dexter and General Re acquisition currency, BNSF share issuance and future returns. https://www.berkshirehathaway.com/letters/2016ltr.pdf Evidence note: Later comparison of three stock-funded acquisitions, showing why the instrument used to pay can be a separate decision from the asset purchased. ### Warren E. Buffett · 2019 Chairman's Letter Warren E. Buffett, 2019 Chairman's Letter (2019 reporting year; signed February 22, 2020), Retained earnings, disappointing controlled businesses and insurance risk. https://www.berkshirehathaway.com/letters/2019ltr.pdf Evidence note: Written before the pandemic's full effects on Berkshire were known. ### Warren E. Buffett · 2020 Chairman's Letter Warren E. Buffett, 2020 Chairman's Letter (2020 reporting year; signed February 27, 2021), Precision Castparts $11 billion write-down and Buffett responsibility statement, printed pages 3-4. https://www.berkshirehathaway.com/letters/2020ltr.pdf Evidence note: Buffett attributes the Precision Castparts overpayment to his own optimism rather than deception by the seller. ### Warren E. Buffett · 2022 Chairman's Letter Warren E. Buffett, 2022 Chairman's Letter (2022 reporting year; signed February 25, 2023), A dozen good decisions, luck, Coca-Cola and American Express, and Berkshire's 1967 insurance pivot. https://www.berkshirehathaway.com/letters/2022ltr.pdf Evidence note: A late retrospective that deliberately compresses a long record. The course tests it against contemporary documents rather than accepting the compression as the whole story. ### Warren E. Buffett · 2023 Chairman's Letter Warren E. Buffett, 2023 Chairman's Letter (2023 reporting year; signed February 24, 2024), BHE regulatory and wildfire reassessment, printed pages 13-14; Charlie Munger tribute, printed pages 5-6. https://www.berkshirehathaway.com/letters/2023ltr.pdf Evidence note: First annual letter after Munger's death. Buffett's account of their relationship is retrospective and affectionate; the contemporaneous letters show the revisions unfolding more gradually. ### Warren E. Buffett · 2024 Chairman's Letter Warren E. Buffett, 2024 Chairman's Letter (2024 reporting year; signed February 22, 2025), Mistakes, delayed correction, GEICO, cash, Berkshire's sixty-year record and succession. https://www.berkshirehathaway.com/letters/2024ltr.pdf Evidence note: Buffett's final Berkshire annual letter. It was reviewed from the complete local original. ### Warren E. Buffett, Greg Abel and Ajit Jain · 2025 Berkshire Hathaway annual meeting Warren E. Buffett, Greg Abel and Ajit Jain, 2025 Berkshire Hathaway annual meeting (May 3, 2025), Buffett's closing succession announcement and meeting record. https://buffett.cnbc.com/2025-berkshire-hathaway-annual-meeting/ Evidence note: CNBC's authorized archive page. The meeting was reviewed in part, with emphasis on the succession announcement. ### Warren E. Buffett · Thanksgiving Message to Fellow Shareholders Warren E. Buffett, Thanksgiving Message to Fellow Shareholders (November 10, 2025), Complete message, including succession, philanthropy, delayed personnel action and final reflections. https://berkshirehathaway.com/news/nov1025.pdf Evidence note: Latest Buffett-authored material located on Berkshire's official site as of September 9, 2026. Read in full. ### Greg Abel · 2025 Berkshire Hathaway annual letter Greg Abel, 2025 Berkshire Hathaway annual letter (2025 reporting year; published February 28, 2026), Opening, authorship, culture and stewardship. https://www.berkshirehathaway.com/letters/2025ltr.pdf Evidence note: This is Greg Abel's first annual letter as Berkshire CEO. It is successor evidence, not Buffett-authored material. ### Warren E. Buffett · Warren Buffett shareholder-letter archive Warren E. Buffett, Warren Buffett shareholder-letter archive (1965-2024 reporting years), Individual Buffett letters for 1977-2024 and official notice of the complete 1965-2024 book. https://www.berkshirehathaway.com/letters/letters.html Evidence note: Berkshire official archive. It identifies the Buffett series separately from Greg Abel's successor letters and states that the complete printed collection covers 1965-2024. ### Warren E. Buffett · The Security I Like Best Warren E. Buffett, The Security I Like Best (December 6, 1951), Commercial and Financial Chronicle archive; exact article scan located but not completely extracted in this pass. https://archive.org/details/pub_commercial-and-financial-chronicle Evidence note: The original periodical archive was located, but the exact article was not verified by a complete direct read. The course therefore uses no quotation from it and labels the detailed early GEICO account as Buffett's 1995 retrospective. ### Warren E. Buffett · 1995 Chairman's Letter Warren E. Buffett, 1995 Chairman's Letter (1995 reporting year; published 1996), GEICO section, including Buffett's retrospective account of 1951-1952, the 1976 return and the 1995 agreement to acquire the remaining shares. https://www.berkshirehathaway.com/letters/1995.html Evidence note: Official original. This is later self-reporting about the 1951 episode, not contemporaneous evidence of what the unread 1951 article said in full. ### Warren E. Buffett · Second Annual Letter to Limited Partners Warren E. Buffett, Second Annual Letter to Limited Partners (1957 reporting year; circulated 1958), Reproduced original, printed pages 1-3. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. The reproduced original identifies itself as the second annual letter and includes a fragment of the unavailable prior letter. ### Warren E. Buffett · 1958 partnership letter Warren E. Buffett, 1958 partnership letter (1958 reporting year; circulated 1959), Reproduced original, printed pages 4-6. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Includes the complete Commonwealth Trust account and the stated expectation for performance in a rising market. ### Warren E. Buffett · Ground Rules and 1962 partnership review Warren E. Buffett, Ground Rules and 1962 partnership review (January 18, 1963), Reproduced original, printed pages 32-41, especially Dempster on pages 37-39. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Contemporary account of Dempster before and after the April 1962 management change. ### Warren E. Buffett · 1963 partnership review Warren E. Buffett, 1963 partnership review (January 18, 1964), Reproduced original, printed pages 51-62, especially the Dempster outcome on pages 59-61. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full in the local original compilation, printed pages 51-62. The source records the expanding capital base, performance comparisons and Dempster follow-through. ### Warren E. Buffett · Partnership objective-revision letter Warren E. Buffett, Partnership objective-revision letter (October 9, 1967), Reproduced original, printed pages 111-114. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Buffett explains why scale, scarce bargains and a changed market led him to reduce the partnership objective. ### Warren E. Buffett · 1965 partnership review Warren E. Buffett, 1965 partnership review (January 20, 1966), Reproduced original, printed pages 85-94. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Buffett limits new partners, describes scale as an emerging constraint, formalizes concentration limits and records his worst investment experience to date. ### Warren E. Buffett · The First Decade partnership review Warren E. Buffett, The First Decade partnership review (January 25, 1967), Reproduced original, printed pages 100-107. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Buffett states that the first decade could not be duplicated, identifies the shrinking idea flow, and distinguishes method from market fashion. ### Warren E. Buffett · 1967 partnership review Warren E. Buffett, 1967 partnership review (January 24, 1968), Reproduced original, printed pages 115-119. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. The letter measures the first year after the objective revision and separates headline outperformance from a poor workout result and the lower-return control portfolio. ### Warren E. Buffett · 1968 partnership review Warren E. Buffett, 1968 partnership review (January 22, 1969), Reproduced original, printed pages 123-128. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Buffett calls the 58.8 percent partnership result for the 1968 reporting year a freak and repeats that idea quality and quantity were at an all-time low. ### Warren E. Buffett · Partnership retirement announcement Warren E. Buffett, Partnership retirement announcement (May 29, 1969), Reproduced original, printed pages 129-131. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Buffett accepts responsibility for the 1969 workout losses and announces liquidation rather than continue without first-class ideas. ### Warren E. Buffett · Controlled-company distribution letter Warren E. Buffett, Controlled-company distribution letter (December 5, 1969), Reproduced original, printed pages 137-140. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Buffett reports the weak textile economics, the redeployment into insurance and banking, and the planned distribution of Berkshire and Diversified Retailing shares. ### Warren E. Buffett · Partnership liquidation questions and answers Warren E. Buffett, Partnership liquidation questions and answers (December 26, 1969), Reproduced original, printed pages 141-144. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. Buffett explains why Berkshire textiles remained open despite poor returns and details the constraints on distributing illiquid control shares. ### Warren E. Buffett · Final partnership letter Warren E. Buffett, Final partnership letter (February 25, 1970), Reproduced original, printed pages 145-152. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf Evidence note: Read in full. The closing letter gives partners a complete bond-selection framework and marks the transition out of advisory work. ### Warren E. Buffett · Letter to Charles N. Huggins at See’s Candy Shops Warren E. Buffett, Letter to Charles N. Huggins at See’s Candy Shops (December 13, 1972), Independent hosted scan, numbered points 1-4 and continuation headed Page #2. Partial OCR read through the Buffett signature; the full five-page scan was not visually accessible.. https://theoraclesclassroom.com/wp-content/uploads/2019/10/1972-Buffett-Letter-to-Sees-Candies.pdf Evidence note: Read in part from OCR with visible errors. No quotation or numerical claim is drawn from it. The readable portion documents Buffett proposing bounded merchandising tests while crediting Huggins’s greater product knowledge. ### Warren E. Buffett · Letter to George D. Young about GEICO Warren E. Buffett, Letter to George D. Young about GEICO (July 22, 1976), Independent hosted scan, numbered reasons on pages 1-3 and allocation of survival and underwriting responsibility at the end of page 3. Partial OCR ends before the fourth scan page.. https://theoraclesclassroom.com/wp-content/uploads/2020/10/Buffett-Letter-about-GEICO-1976.pdf Evidence note: Read in part from visibly imperfect OCR. No quotation or numerical claim is drawn from it. The readable portion separates sentiment, GEICO’s cost position, survival judgment and underwriting responsibility. ### Warren E. Buffett · Early Commercial & Financial Chronicle investment write-ups Warren E. Buffett, Early Commercial & Financial Chronicle investment write-ups (Early career; exact constituent dates unverified), Independent archive identifies a five-page scan. Constituent article titles and dates could not be verified because no readable extraction or local download was available.. https://theoraclesclassroom.com/wp-content/uploads/2020/10/Buffett-Old-invesetment-writeups-Commercial-_-Financial-Chronicle.pdf Evidence note: Located, not read. This item does not close the separate 1951 GEICO article gap. ### Jeffrey P. Bezos · 1999 Letter to Shareholders Jeffrey P. Bezos, 1999 Letter to Shareholders (1999 reporting period; published 2000), A Recap of 1999, Goals for 2000, distribution capacity and platform expansion. https://s2.q4cdn.com/299287126/files/doc_financials/annual/Shareholderletter99.pdf Evidence note: Contemporary management account of hypergrowth, category expansion and infrastructure burden. Later outcomes show that several named initiatives failed. ### Jeffrey P. Bezos · 2000 Letter to Shareholders Jeffrey P. Bezos, 2000 Letter to Shareholders (2000 reporting period; published 2001), Opening share-price discussion, platform investments and Goal for 2001. https://s2.q4cdn.com/299287126/files/doc_financials/annual/00ar_letter.pdf Evidence note: Contemporary admission that the land-rush metaphor had understated the time and difficulty required for single-category companies to reach scale. ### Jeffrey P. Bezos · 2003 Letter to Shareholders Jeffrey P. Bezos, 2003 Letter to Shareholders (2003 reporting period; published 2004), Long-term owner discussion, negative reviews and Instant Order Update. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2003_-Shareholder_-Letter041304.pdf Evidence note: Contemporary owner-versus-tenant frame following Amazon's first full-year profit. The examples describe deliberate short-term sales sacrifices. ### Jeffrey P. Bezos · 2006 Letter to Shareholders Jeffrey P. Bezos, 2006 Letter to Shareholders (2006 reporting period; published 2007), Planting Seeds section and discussion of physical stores, FBA, AWS and new categories. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2006.PDF Evidence note: Contemporary new-business test. The rejection of physical stores records a position that later changed when management saw differentiation. ### Jeffrey P. Bezos · 2007 Letter to Shareholders Jeffrey P. Bezos, 2007 Letter to Shareholders (2007 reporting period; published 2008), Kindle development, launch demand and supply response. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2007letter.pdf Evidence note: Contemporary account of a product developed for more than three years and then constrained by demand that exhausted launch inventory in 5.5 hours. ### Jeffrey P. Bezos · 2008 Letter to Shareholders Jeffrey P. Bezos, 2008 Letter to Shareholders (2008 reporting period; published 2009), Working backward, durable customer needs and recession-era investment. https://s2.q4cdn.com/299287126/files/doc_financials/annual/Amazon_SH_Letter_2008.pdf Evidence note: Management's recession-era account of choosing opportunities from customer needs rather than from existing skills. ### Jeffrey P. Bezos · 2010 Letter to Shareholders Jeffrey P. Bezos, 2010 Letter to Shareholders (2010 reporting period; published 2011), Service architecture, technology examples and connection to free cash flow. https://s2.q4cdn.com/299287126/files/doc_financials/annual/117006_ltr_ltr2.pdf Evidence note: Management account of technical capabilities developed under operating scale. It links engineering work to cash flow without isolating causal contribution. ### Jeffrey P. Bezos · 2011 Letter to Shareholders Jeffrey P. Bezos, 2011 Letter to Shareholders (2011 reporting period; published 2012), Self-service platforms, AWS, Marketplace, KDP and FBA. https://s2.q4cdn.com/299287126/files/doc_financials/annual/letter.PDF Evidence note: Management's case that Amazon platforms let outside participants bypass older gatekeepers. Later institutional evidence tests Amazon's own gatekeeper role. ### Jeffrey P. Bezos · 2012 Letter to Shareholders Jeffrey P. Bezos, 2012 Letter to Shareholders (2012 reporting period; published 2013), Proactive customer invention, blind alleys, device economics and AWS iteration. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2012-Shareholder-Letter.pdf Evidence note: Management acknowledges self-inflicted mistakes and blind alleys while describing continuing investment in devices and cloud services. ### Jeffrey P. Bezos · 2013 Letter to Shareholders Jeffrey P. Bezos, 2013 Letter to Shareholders (2013 reporting period; published 2014), Initiative catalogue, experiments, Amazon Fresh, fulfilment-centre improvement and employee programs. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2013-Letter-to-Shareholders.pdf Evidence note: A wide management inventory of initiatives, experimentation and operating practices. It gives selected results, not a complete denominator of attempts or costs. ### Jeffrey P. Bezos · 2017 Letter to Shareholders Jeffrey P. Bezos, 2017 Letter to Shareholders (2017 reporting period; published 2018), High standards, operating standards admission, six-page memos and Whole Foods. https://s2.q4cdn.com/299287126/files/doc_financials/annual/Amazon_Shareholder_Letter.pdf Evidence note: Bezos states that colleagues taught him operating standards he initially lacked. The Whole Foods acquisition also revises his 2006 physical-store position. ### Jeffrey P. Bezos · 2018 Letter to Shareholders Jeffrey P. Bezos, 2018 Letter to Shareholders (2018 reporting period; published 2019), Third-party sales, wandering, Fire Phone transfer, wage challenge and employee programs. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2018-Letter-to-Shareholders.pdf Evidence note: Retrospective comparison of first-party and third-party growth, plus management's account of retaining people and capability after the Fire Phone failure. ### Jeffrey P. Bezos · 2019 Letter to Shareholders Jeffrey P. Bezos, 2019 Letter to Shareholders (2019 reporting period; published April 2020), COVID-19 response, testing work, temporary pay, jobs and climate commitments. https://s2.q4cdn.com/299287126/files/doc_financials/2020/ar/2019-Shareholder-Letter.pdf Evidence note: Management's early-pandemic account, written while operating conditions were changing quickly. It does not independently assess worker experience or adequacy of protection. ### Jeffrey P. Bezos · 2020 Letter to Shareholders Jeffrey P. Bezos, 2020 Letter to Shareholders (2020 reporting period; published April 2021), Stakeholder value estimates, Bessemer discussion, safety commitments and final CEO reflection. https://s2.q4cdn.com/299287126/files/doc_financials/2021/ar/Amazon-2020-Shareholder-Letter-and-1997-Shareholder-Letter.pdf Evidence note: Bezos's final CEO letter. It responds to worker criticism and makes new employee and safety commitments while acknowledging uncertainty in some stakeholder-value estimates. ### Jeffrey P. Bezos · Jeff Bezos interview Jeffrey P. Bezos, Jeff Bezos interview (May 4, 2001), Full Academy of Achievement interview transcript. https://achievement.org/achiever/jeffrey-p-bezos/ Evidence note: Early retrospective account with specific credit to MacKenzie Bezos, Shel Kaphan, Bezos's parents, investors and early coworkers. Motivational claims remain his account. ### Jeffrey P. Bezos · 2010 Baccalaureate Remarks Jeffrey P. Bezos, 2010 Baccalaureate Remarks (May 30, 2010), Complete Princeton transcript. https://www.princeton.edu/news/2010/05/30/2010-baccalaureate-remarks Evidence note: Later moral retelling of childhood and the decision to start Amazon. Use as retrospective framing rather than contemporaneous evidence of 1994 motives. ### Jeffrey P. Bezos and Walter Mossberg · Jeff Bezos live from Code 2016 Jeffrey P. Bezos and Walter Mossberg, Jeff Bezos live from Code 2016 (May 31, 2016), Recode liveblog of the full stage interview; complete 81-minute audio separately located. https://live.recode.net/jeff-bezos-2016-code/ Evidence note: Read in part. The liveblog is not a verbatim transcript and the complete audio was not directly reviewed, so no omitted answer is treated as evidence. ### Jeffrey P. Bezos and David M. Rubenstein · Interview at the Economic Club of Washington Jeffrey P. Bezos and David M. Rubenstein, Interview at the Economic Club of Washington (September 13, 2018), Complete edited transcript. https://www.economicclub.org/sites/default/files/transcripts/Jeff_Bezos_Edited_Transcript.pdf Evidence note: Later retrospective account that distributes credit among mentors, colleagues and family, and identifies exceptional luck in AWS's competitive lead. ### Jeffrey P. Bezos · Written Testimony before the House Judiciary Subcommittee Jeffrey P. Bezos, Written Testimony before the House Judiciary Subcommittee (July 29, 2020), Complete written testimony. https://docs.house.gov/meetings/JU/JU05/20200729/110883/HHRG-116-JU05-Wstate-BezosJ-20200729.pdf Evidence note: Founder narrative delivered under antitrust scrutiny. It supplies specific claims but is advocacy, not an independent assessment of Amazon's market power. ### Jeff Bezos; introduction by Walter Isaacson · Invent and Wander: The Collected Writings of Jeff Bezos Jeff Bezos; introduction by Walter Isaacson, Invent and Wander: The Collected Writings of Jeff Bezos (November 17, 2020), Harvard Business Review Press publisher page. https://store.hbr.org/product/invent-and-wander-the-collected-writings-of-jeff-bezos-with-an-introduction-by-walter-isaacson/10466 Evidence note: Located, not read. No authorized full copy was available in the supplied library. Accessible component letters and speeches were reviewed separately. ### Brad Stone · The Everything Store: Jeff Bezos and the Age of Amazon Brad Stone, The Everything Store: Jeff Bezos and the Age of Amazon (October 15, 2013), Little, Brown and Company publisher page. https://www.hachettebookgroup.com/titles/brad-stone/the-everything-store/9780316219259/ Evidence note: Located, not read. The publisher page establishes scope and bibliographic details, not evidence for the chapters. ### Brad Stone · Amazon Unbound: Jeff Bezos and the Invention of a Global Empire Brad Stone, Amazon Unbound: Jeff Bezos and the Invention of a Global Empire (May 11, 2021), Simon & Schuster publisher page. https://www.simonandschuster.com/books/Amazon-Unbound/Brad-Stone/9781982132620 Evidence note: Located, not read. The publisher description was not used as reported evidence. ### Jodi Kantor and David Streitfeld · Inside Amazon: Wrestling Big Ideas in a Bruising Workplace Jodi Kantor and David Streitfeld, Inside Amazon: Wrestling Big Ideas in a Bruising Workplace (August 15, 2015), Complete reported article, including Amazon's responses. https://www.nytimes.com/2015/08/16/technology/inside-amazon-wrestling-big-ideas-in-a-bruising-workplace.html Evidence note: Reporting based on more than 100 current and former employees. Accounts vary, but the article documents costs and contest around the culture Bezos praised. ### U.S. House Judiciary Subcommittee on Antitrust, Commercial and Administrative Law · Investigation of Competition in Digital Markets: Majority Staff Report and Recommendations U.S. House Judiciary Subcommittee on Antitrust, Commercial and Administrative Law, Investigation of Competition in Digital Markets: Majority Staff Report and Recommendations (October 6, 2020), Amazon overview and selected marketplace, seller-data, fee, private-label, acquisition and logistics passages. https://www.govinfo.gov/content/pkg/GOVPUB-Y4_J89_1-PURL-gpo145949/pdf/GOVPUB-Y4_J89_1-PURL-gpo145949.pdf Evidence note: Read in part. The majority staff report draws on documents and market participants but is not a judicial finding and was not read outside the Amazon-relevant portions. ### National Labor Relations Board · Amazon.com Services LLC, Case 10-RC-269250 National Labor Relations Board, Amazon.com Services LLC, Case 10-RC-269250 (2021), Official case page, initial tally and election-result records. https://www.nlrb.gov/case/10-RC-269250 Evidence note: Read in part. The official case and tally pages were reviewed, but the complete docket and later regional decision were not read in full. ### Andy Jassy · 2021 Letter to Shareholders Andy Jassy, 2021 Letter to Shareholders (2021 reporting period; published 2022), Author byline, closing signature and appended 1997-letter label. https://www.aboutamazon.com/news/company-news/2021-letter-to-shareholders Evidence note: The current-period letter is authored by Andy Jassy. The appended 1997 letter is separately authored by Jeffrey P. Bezos. Only authorship and the reprint boundary were checked for this source; no full-reading claim is made. ### Andy Jassy · 2022 Letter to Shareholders Andy Jassy, 2022 Letter to Shareholders (2022 reporting period; published 2023), Author byline, closing signature and appended 1997-letter label. https://www.aboutamazon.com/news/company-news/amazon-ceo-andy-jassy-2022-letter-to-shareholders Evidence note: The current-period letter is authored by Andy Jassy. The appended 1997 letter is separately authored by Jeffrey P. Bezos. Only authorship and the reprint boundary were checked for this source; no full-reading claim is made. ### Andy Jassy · 2023 Letter to Shareholders Andy Jassy, 2023 Letter to Shareholders (2023 reporting period; published 2024), Author byline, closing signature and appended 1997-letter label. https://www.aboutamazon.com/news/company-news/amazon-ceo-andy-jassy-2023-letter-to-shareholders Evidence note: The current-period letter is authored by Andy Jassy. The appended 1997 letter is separately authored by Jeffrey P. Bezos. Only authorship and the reprint boundary were checked for this source; no full-reading claim is made. ### Andy Jassy · 2024 Letter to Shareholders Andy Jassy, 2024 Letter to Shareholders (2024 reporting period; published 2025), Author byline, closing signature and appended 1997-letter label. https://www.aboutamazon.com/news/company-news/amazon-ceo-andy-jassy-2024-letter-to-shareholders Evidence note: The current-period letter is authored by Andy Jassy. The appended 1997 letter is separately authored by Jeffrey P. Bezos. Only authorship and the reprint boundary were checked for this source; no full-reading claim is made. ### Howard Marks · The Route to Performance Howard Marks, The Route to Performance (1990-10-12), Printed pp. 1-2, the fourteen-year pension-plan record and the concluding avoidance-of-losers passage. https://www.oaktreecapital.com/docs/default-source/memos/1990-10-12-the-route-to-performance.pdf?sfvrsn=33bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · First Quarter Performance Howard Marks, First Quarter Performance (1991-04-11), Printed pp. 1-2, opening pendulum discussion and the closing three-step response to extremes. https://www.oaktreecapital.com/docs/default-source/memos/1991-04-11-first-quarter-client-performance.pdf?sfvrsn=d7bd0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Microeconomics 101: Supply, Demand and Convertibles Howard Marks, Microeconomics 101: Supply, Demand and Convertibles (1992-10-08), Complete original memo dated October 8, 1992. https://www.oaktreecapital.com/docs/default-source/memos/1992-10-08-microeconomics.pdf?sfvrsn=43bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The Value of Predictions, or Where'd All This Rain Come From? Howard Marks, The Value of Predictions, or Where'd All This Rain Come From? (1993-02-15), Complete memo, the seven-step forecasting chain and the closing tests for whether a forecast is actionable. https://www.oaktreecapital.com/docs/default-source/memos/1993-02-15-the-value-of-predictions-or-where-39-d-all-this-rain-come-from.pdf?sfvrsn=6fbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Random Thoughts on the Identification of Investment Opportunities Howard Marks, Random Thoughts on the Identification of Investment Opportunities (1994-01-24), Complete original memo dated January 24, 1994. https://www.oaktreecapital.com/docs/default-source/memos/1994-01-24-random-thoughts-on-the-identification-of-inv-opportunities.pdf?sfvrsn=17bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Risk in Today's Markets Howard Marks, Risk in Today's Markets (1994-02-17), Complete original memo dated February 17, 1994. https://www.oaktreecapital.com/docs/default-source/memos/1994-02-17-risk-in-todays-markets.pdf?sfvrsn=dfbd0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · "Risk in Today's Markets" Revisited Howard Marks, "Risk in Today's Markets" Revisited (1994-04-11), Complete original memo dated April 11, 1994. https://www.oaktreecapital.com/docs/default-source/memos/1994-04-11-risk-in-todays-markets-revisted.pdf?sfvrsn=2bbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · "How Does an Inefficient Market Get That Way?" Howard Marks, "How Does an Inefficient Market Get That Way?" (1994-07-15), Complete original memo dated July 15, 1994. https://www.oaktreecapital.com/docs/default-source/memos/1994-07-15-how-does-an-inefficient-market-get-that-way.pdf?sfvrsn=dbbd0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · How the Game Should Be Played Howard Marks, How the Game Should Be Played (1995-05-26), Complete memo, the baseball analogy and the closing discussion of Oaktree’s game plan. https://www.oaktreecapital.com/docs/default-source/memos/1995-05-26-how-the-game-should-be-played.pdf?sfvrsn=d3bd0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The Value of Predictions II (or "Give That Man a Cigar") Howard Marks, The Value of Predictions II (or "Give That Man a Cigar") (1996-07-22), Complete original memo dated July 22, 1996. https://www.oaktreecapital.com/docs/default-source/memos/1996-07-22-the-value-of-predictions-ii-or-give-that-man-a-cigar.pdf?sfvrsn=27bc0f65_8 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Will It Be Different This Time? Howard Marks, Will It Be Different This Time? (1996-11-25), Complete original memo dated November 25, 1996. https://www.oaktreecapital.com/docs/default-source/memos/1996-11-25-will-it-be-different-this-time.pdf?sfvrsn=23bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Are You An Investor or a Speculator? Howard Marks, Are You An Investor or a Speculator? (1997-09-03), Complete memo, passages defining investment versus speculation and applying the distinction to technology stocks. https://www.oaktreecapital.com/docs/default-source/memos/are-you-an-investor-or-a-speculator.pdf?sfvrsn=1e37cf65_2 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Who Knew? Howard Marks, Who Knew? (1998-01-08), Complete original memo dated January 8, 1998. https://www.oaktreecapital.com/docs/default-source/memos/1998-01-08-who-knew.pdf?sfvrsn=2fbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Genius Isn't Enough (and Other Lessons from Long-Term Capital Management) Howard Marks, Genius Isn't Enough (and Other Lessons from Long-Term Capital Management) (1998-10-09), Complete original memo dated October 9, 1998. https://www.oaktreecapital.com/docs/default-source/memos/1998-10-09-genius-isn-39-t-enough.pdf?sfvrsn=eac20f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · How's the Market? Howard Marks, How's the Market? (1999-04-15), Complete original memo dated April 15, 1999. https://www.oaktreecapital.com/docs/default-source/memos/1999-04-15-hows-the-market.pdf?sfvrsn=3fbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · bubble.com Howard Marks, bubble.com (2000-01-02), Complete memo, the opening bubble diagnosis and the numbered discussion of what happened, why, and what follows. https://www.oaktreecapital.com/docs/default-source/memos/2000-01-02-bubble.pdf?sfvrsn=37bc0f65_5 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Irrational Exuberance Howard Marks, Irrational Exuberance (2000-05-01), Complete memo, the Robertson, Soros-Druckenmiller, and Brinson cases and the closing discussion of pressure while early. https://www.oaktreecapital.com/docs/default-source/memos/2000-05-01-irrational-exuberance.pdf?sfvrsn=cfbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Investment Miscellany Howard Marks, Investment Miscellany (2000-11-16), Complete original memo dated November 16, 2000. https://www.oaktreecapital.com/docs/default-source/memos/2000-11-16-investment-miscellany.pdf?sfvrsn=1fbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · We're Not In 1999 Anymore, Toto Howard Marks, We're Not In 1999 Anymore, Toto (2000-12-31), Complete memo, opening correction retrospective and closing valuation and prospective-return discussion. https://www.oaktreecapital.com/docs/default-source/memos/2000-12-31-were-not-in-1999-anymore.pdf?sfvrsn=3bbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Safety First . . . But Where? Howard Marks, Safety First . . . But Where? (2001-04-10), Complete original memo dated April 10, 2001. https://www.oaktreecapital.com/docs/default-source/memos/2001-04-10-safety-first-but-where.pdf?sfvrsn=49c10f65_8 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · What's It All About, Alpha? Howard Marks, What's It All About, Alpha? (2001-07-11), Complete memo, opening Wharton-Chicago synthesis and the alpha, beta, and efficient-market discussions. https://www.oaktreecapital.com/docs/default-source/memos/2001-07-11-whats-it-all-about-alpha.pdf?sfvrsn=13bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Notes from New York Howard Marks, Notes from New York (2001-09-16), Complete original memo dated September 16, 2001. https://www.oaktreecapital.com/docs/default-source/memos/2001-09-16-notes-from-ny.pdf?sfvrsn=fbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · What Lies Ahead? Howard Marks, What Lies Ahead? (2001-10-04), Printed pp. 6-7, the “A Bear’s Eye View” response and closing discussion of caution, lost upside, and investor agility. https://www.oaktreecapital.com/docs/default-source/memos/2001-10-04-what-lies-ahead.pdf?sfvrsn=bbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · You Can't Predict. You Can Prepare. Howard Marks, You Can't Predict. You Can Prepare. (2001-11-20), Complete memo, discussions of staying power, cash, early caution, and preparing for unknowable cycles. https://www.oaktreecapital.com/docs/default-source/memos/2001-11-20-you-cant-predict-you-can-prepare.pdf?sfvrsn=bc00f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Learning From Enron Howard Marks, Learning From Enron (2002-03-14), Complete original memo dated March 14, 2002. https://www.oaktreecapital.com/docs/default-source/memos/2002-03-14-learning-from-enron.pdf?sfvrsn=77bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The Realist's Creed Howard Marks, The Realist's Creed (2002-05-31), Complete original memo dated May 31, 2002. https://www.oaktreecapital.com/docs/default-source/memos/2002-05-31-the-realists-creed.pdf?sfvrsn=1bbc0f65_5 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Quo Vadis? Howard Marks, Quo Vadis? (2002-07-26), Complete original memo dated July 26, 2002. https://www.oaktreecapital.com/docs/default-source/memos/2002-07-26-quo-vadis.pdf?sfvrsn=67bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Etorre's Wisdom Howard Marks, Etorre's Wisdom (2002-09-04), Complete original memo dated September 4, 2002. https://www.oaktreecapital.com/docs/default-source/memos/2002-09-04-etorres-wisdom.pdf?sfvrsn=63bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Returns and How They Get That Way Howard Marks, Returns and How They Get That Way (2002-11-11), Complete original memo dated November 11, 2002. https://www.oaktreecapital.com/docs/default-source/memos/2002-11-11-returns-and-how-they-get-that-way.pdf?sfvrsn=4fbc0f65_8 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Whad’Ya Know? Howard Marks, Whad’Ya Know? (2003-03-11), Complete original memo dated March 11, 2003. https://www.oaktreecapital.com/docs/default-source/memos/2003-03-11-whad-ya-know.pdf?sfvrsn=abbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · What's Going On? Howard Marks, What's Going On? (2003-05-06), Complete original memo dated May 6, 2003. https://www.oaktreecapital.com/docs/default-source/memos/2003-05-16-whats-going-on.pdf?sfvrsn=73bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The Most Important Thing Howard Marks, The Most Important Thing (2003-07-01), Printed pp. 1-13, eighteen “most important thing” passages and closing account of the resulting Oaktree creed. https://www.oaktreecapital.com/docs/default-source/memos/2003-07-01-the-most-important-thing.pdf?sfvrsn=91c00f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · What’s Your Game Plan? Howard Marks, What’s Your Game Plan? (2003-09-05), Complete original memo dated September 5, 2003. https://www.oaktreecapital.com/docs/default-source/memos/2003-09-05-whats-your-game-plan.pdf?sfvrsn=7bbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The Feeling's Mutual Howard Marks, The Feeling's Mutual (2003-10-02), Complete original memo dated October 2, 2003. https://www.oaktreecapital.com/docs/default-source/memos/2003-10-02-the-feelings-mutual.pdf?sfvrsn=47bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Hey, Steward!! Howard Marks, Hey, Steward!! (2004-03-16), Complete original memo dated March 16, 2004. https://www.oaktreecapital.com/docs/default-source/memos/2004-03-16-hey-steward!!.pdf?sfvrsn=a7bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Us and Them Howard Marks, Us and Them (2004-05-07), Complete original memo dated May 7, 2004. https://www.oaktreecapital.com/docs/default-source/memos/2004-05-07-us-and-them.pdf?sfvrsn=bfbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The Happy Medium Howard Marks, The Happy Medium (2004-07-20), Complete original memo dated July 20, 2004. https://www.oaktreecapital.com/docs/default-source/memos/2004-07-21-the-happy-medium.pdf?sfvrsn=4bbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Hedge Funds: A Case for Caution Howard Marks, Hedge Funds: A Case for Caution (2004-10-06), Complete original memo dated October 6, 2004. https://www.oaktreecapital.com/docs/default-source/memos/2004-10-06-hedge-funds-a-case-for-caution.pdf?sfvrsn=57bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Risk and Return Today Howard Marks, Risk and Return Today (2004-10-27), Printed pp. 1-7, sections “The Market at Work - 2004 Version,” “Why a Flatter Line?,” and “Implications for Investing”. https://www.oaktreecapital.com/docs/default-source/memos/2004-10-27-risk-and-return-today.pdf?sfvrsn=53bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Oaktree at Ten Howard Marks, Oaktree at Ten (2005-04-11), Printed pp. 1-5, sections “Priorities,” “Clients,” “Performance,” “Growth,” and “Plans for the Future”. https://www.oaktreecapital.com/docs/default-source/memos/2005-04-11-oaktree-at-ten.pdf?sfvrsn=5fbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · There They Go Again Howard Marks, There They Go Again (2005-05-06), Complete original memo dated May 6, 2005. https://www.oaktreecapital.com/docs/default-source/memos/2005-05-06-there-they-go-again.pdf?sfvrsn=5bbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · A Case in Point Howard Marks, A Case in Point (2005-06-06), Complete original memo dated June 6, 2005. https://www.oaktreecapital.com/docs/default-source/memos/2005-06-06-a-case-in-point.pdf?sfvrsn=a3bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Hindsight First, Please (or, What Were They Thinking?) Howard Marks, Hindsight First, Please (or, What Were They Thinking?) (2005-10-17), Complete original memo dated October 17, 2005. https://www.oaktreecapital.com/docs/default-source/memos/2005-10-17-hindsight-first-please-or-what-where-they-thinking.pdf?sfvrsn=93bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Risk Howard Marks, Risk (2006-01-19), Complete original memo dated January 19, 2006. https://www.oaktreecapital.com/docs/default-source/memos/2006-01-19-risk.pdf?sfvrsn=afbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · It Is What It Is Howard Marks, It Is What It Is (2006-03-27), Complete original memo dated March 27, 2006. https://www.oaktreecapital.com/docs/default-source/memos/2006-03-27-it-is-what-it-is.pdf?sfvrsn=87bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Returns, Absolute Returns and Risk Howard Marks, Returns, Absolute Returns and Risk (2006-06-13), Complete original memo dated June 13, 2006. https://www.oaktreecapital.com/docs/default-source/memos/2006-06-13-returns-absolute-returns-and-risk.pdf?sfvrsn=b7bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · You Can’t Eat IRR Howard Marks, You Can’t Eat IRR (2006-07-12), Complete original memo dated July 12, 2006. https://www.oaktreecapital.com/docs/default-source/memos/2006-07-12-you-cant-eat-irr.pdf?sfvrsn=83bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Dare to Be Great Howard Marks, Dare to Be Great (2006-09-07), Complete original memo dated September 7, 2006. https://www.oaktreecapital.com/docs/default-source/memos/2006-09-07-dare-to-be-great.pdf?sfvrsn=b3bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The New Paradigm Howard Marks, The New Paradigm (2006-10-19), Complete original memo dated October 19, 2006. https://www.oaktreecapital.com/docs/default-source/memos/2006-10-19-the-new-paradigm.pdf?sfvrsn=8bbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Pigweed Howard Marks, Pigweed (2006-12-07), Complete original memo dated December 7, 2006. https://www.oaktreecapital.com/docs/default-source/memos/2006-12-07-pigweed.pdf?sfvrsn=bbbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The Race to the Bottom Howard Marks, The Race to the Bottom (2007-02-14), Complete memo, opening covenant and leverage evidence and closing “race to the bottom” diagnosis. https://www.oaktreecapital.com/docs/default-source/memos/2007-02-14-the-race-to-the-bottom.pdf?sfvrsn=9bbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Everyone Knows Howard Marks, Everyone Knows (2007-04-26), Complete original memo dated April 26, 2007. https://www.oaktreecapital.com/docs/default-source/memos/2007-04-26-everyone-knows.pdf?sfvrsn=d3bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · It’s All Good Howard Marks, It’s All Good (2007-07-16), Complete memo, passages on the flattened risk-return line, thin risk premiums, and reduced reserve capacity. https://www.oaktreecapital.com/docs/default-source/memos/2007-07-16-its-all-good.pdf?sfvrsn=8fbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · It’s All Good . . . Really? Howard Marks, It’s All Good . . . Really? (2007-07-30), Complete memo, opening admission that the credit turn arrived sooner than expected and the revised assessment that follows. https://www.oaktreecapital.com/docs/default-source/memos/2007-07-30-its-all-good-really.pdf?sfvrsn=97bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Now It’s All Bad? Howard Marks, Now It’s All Bad? (2007-09-10), Complete memo, opening reassessment of the credit reversal and closing discussion of whether pessimism had gone too far. https://www.oaktreecapital.com/docs/default-source/memos/2007-09-10-now-its-all-bad.pdf?sfvrsn=9fbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · No Different This Time – The Lessons of ‘07 Howard Marks, No Different This Time – The Lessons of ‘07 (2007-12-17), Complete original memo dated December 17, 2007. https://www.oaktreecapital.com/docs/default-source/memos/2007-12-17-no-different-this-time-the-lessons-of-07.pdf?sfvrsn=f3bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Now What? Howard Marks, Now What? (2008-01-10), Complete original memo dated January 10, 2008. https://www.oaktreecapital.com/docs/default-source/memos/2008-01-10-now-what.pdf?sfvrsn=d7bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Whodunit Howard Marks, Whodunit (2008-02-20), Complete original memo dated February 20, 2008. https://www.oaktreecapital.com/docs/default-source/memos/2008-02-20-whodunit.pdf?sfvrsn=e7bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The Aviary Howard Marks, The Aviary (2008-05-16), Complete original memo dated May 16, 2008. https://www.oaktreecapital.com/docs/default-source/memos/2008-05-16-the-aviary.pdf?sfvrsn=efbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Doesn’t Make Sense Howard Marks, Doesn’t Make Sense (2008-07-31), Complete original memo dated July 31, 2008. https://www.oaktreecapital.com/docs/default-source/memos/2008-07-31-doesnt-make-sense.pdf?sfvrsn=ebbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · What Worries Me Howard Marks, What Worries Me (2008-08-28), Complete original memo dated August 28, 2008. https://www.oaktreecapital.com/docs/default-source/memos/2008-08-28-what-worries-me.pdf?sfvrsn=f7bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Plan B Howard Marks, Plan B (2008-09-24), Complete original memo dated September 24, 2008. https://www.oaktreecapital.com/docs/default-source/memos/2008-09-24-plan-b.pdf?sfvrsn=ffbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Volatility + Leverage = Dynamite Howard Marks, Volatility + Leverage = Dynamite (2008-12-17), Printed pp. 10-12, “Are You Tall Enough to Use Leverage?” and Oaktree’s disclosure of four affected evergreen funds. https://www.oaktreecapital.com/docs/default-source/memos/2008-12-17-volatility-leverage-dynamite.pdf?sfvrsn=c7bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The Long View Howard Marks, The Long View (2009-01-09), Complete memo, “The Short View” and “The Long View” passages distinguishing the foreseen credit cycle from the missed systemic mechanism. https://www.oaktreecapital.com/docs/default-source/memos/2009-01-09-the-long-view.pdf?sfvrsn=c3bc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Will It Work? Howard Marks, Will It Work? (2009-03-05), Complete original memo dated March 5, 2009. https://www.oaktreecapital.com/docs/default-source/memos/2009-03-05-will-it-work.pdf?sfvrsn=dbbc0f65_8 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · So Much That’s False and Nutty Howard Marks, So Much That’s False and Nutty (2009-07-08), Complete original memo dated July 8, 2009. https://www.oaktreecapital.com/docs/default-source/memos/2009-07-08-so-much-thats-false-nutty.pdf?sfvrsn=23bb0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Touchstones Howard Marks, Touchstones (2009-11-10), Complete original memo dated November 10, 2009. https://www.oaktreecapital.com/docs/default-source/memos/2009-11-10-touchstones.pdf?sfvrsn=2fbb0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Getting Lucky Howard Marks, Getting Lucky (2014-01-16), Complete memo, the Wharton, Chicago, 1978 high-yield reassignment, and decision-quality passages. https://www.oaktreecapital.com/docs/default-source/memos/2014-01-16-getting-lucky.pdf?sfvrsn=c4b70f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · It’s not Easy Howard Marks, It’s not Easy (2015-09-09), Printed pp. 1-3, section “Second-Level Thinking” and the account of drafting the 2009 sample chapter. https://www.oaktreecapital.com/docs/default-source/memos/2015-09-09-its-not-easy.pdf?sfvrsn=47bb0f65_7 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Growing the Pie Howard Marks, Growing the Pie (2019-04-01), Complete original memo dated April 1, 2019. https://www.oaktreecapital.com/docs/default-source/memos/growing-the-pie.pdf?sfvrsn=423fab65_8 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · This Time It's Different Howard Marks, This Time It's Different (2019-06-12), Complete original memo dated June 12, 2019. https://www.oaktreecapital.com/docs/default-source/memos/this-time-its-different.pdf?sfvrsn=fc19af65_10 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · On the Other Hand Howard Marks, On the Other Hand (2019-07-26), Complete original memo dated July 26, 2019. https://www.oaktreecapital.com/docs/default-source/memos/on-the-other-hand.pdf?sfvrsn=9b1a365_4 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Mysterious Howard Marks, Mysterious (2019-10-17), Complete original memo dated October 17, 2019. https://www.oaktreecapital.com/docs/default-source/memos/mysterious.pdf?sfvrsn=8a30a565_12 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · You Bet! Howard Marks, You Bet! (2020-01-13), Complete original memo dated January 13, 2020. https://www.oaktreecapital.com/docs/default-source/memos/you-bet.pdf?sfvrsn=785dbe65_8 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Nobody Knows II Howard Marks, Nobody Knows II (2020-03-03), Printed pp. 1-7, facts/inferences/guesses framework; p. 6, the seasonal-disease guess; p. 7, partial buying response. https://www.oaktreecapital.com/docs/default-source/memos/nobody-knows-ii.pdf?sfvrsn=108eb165_8 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Latest Update Howard Marks, Latest Update (2020-03-19), Printed pp. 1-6, worsening health and economic update and concluding case for incremental buying rather than calling the bottom. https://www.oaktreecapital.com/docs/default-source/memos/weekly.pdf?sfvrsn=cbf3b065_8 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Calibrating Howard Marks, Calibrating (2020-04-06), Printed pp. 1-8, the audit of March views; pp. 5-8, move from overweight defense toward offense and buy-on-the-way-down discussion. https://www.oaktreecapital.com/docs/default-source/memos/calibrating.pdf?sfvrsn=3e98b665_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Knowledge of the Future Howard Marks, Knowledge of the Future (2020-04-14), Complete original memo dated April 14, 2020. https://www.oaktreecapital.com/docs/default-source/memos/knowledge-of-the-future.pdf?sfvrsn=fa39b665_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Uncertainty II Howard Marks, Uncertainty II (2020-05-28), Complete memo, probability-tree discussion and concluding limits on conviction under pandemic uncertainty. https://www.oaktreecapital.com/docs/default-source/memos/uncertainty-ii.pdf?sfvrsn=d48a65_8 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Not Enough Howard Marks, Not Enough (2020-06-11), Complete original memo dated June 11, 2020. https://www.oaktreecapital.com/docs/default-source/memos/not-enough.pdf?sfvrsn=8f828c65_4 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Time for Thinking Howard Marks, Time for Thinking (2020-08-05), Complete original memo dated August 5, 2020. https://www.oaktreecapital.com/docs/default-source/memos/timeforthinking.pdf?sfvrsn=17818c65_8 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Something of Value Howard Marks, Something of Value (2021-01-11), Complete memo and appendix, sections on Marks’s personal value-investing journey and the dialogue with Andrew Marks. https://www.oaktreecapital.com/insights/memo/something-of-value Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · 2020 in Review Howard Marks, 2020 in Review (2021-03-04), Complete original memo dated March 4, 2021. https://www.oaktreecapital.com/insights/memo/2020-in-review Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Thinking About Macro Howard Marks, Thinking About Macro (2021-07-29), Complete original memo dated July 29, 2021. https://www.oaktreecapital.com/insights/memo/thinking-about-macro Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The Winds of Change Howard Marks, The Winds of Change (2021-11-23), Complete original memo dated November 23, 2021. https://www.oaktreecapital.com/insights/memo/the-winds-of-change Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Selling Out Howard Marks, Selling Out (2022-01-13), Complete original memo dated January 13, 2022. https://www.oaktreecapital.com/insights/memo/selling-out Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The Pendulum in International Affairs Howard Marks, The Pendulum in International Affairs (2022-03-23), Complete original memo dated March 23, 2022. https://www.oaktreecapital.com/insights/memo/the-pendulum-in-international-affairs Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Bull Market Rhymes Howard Marks, Bull Market Rhymes (2022-05-26), Complete original memo dated May 26, 2022. https://www.oaktreecapital.com/insights/memo/bull-market-rhymes Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Conversation at Panmure House Howard Marks, Conversation at Panmure House (2022-06-23), Complete interview transcript, discussion of cycle causality and the “Listening to the Cycle” working title. https://www.oaktreecapital.com/insights/memo/conversation-at-panmure-house Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · I Beg to Differ Howard Marks, I Beg to Differ (2022-07-26), Complete memo, sections on contrarianism, selling, and the Something of Value retrospective. https://www.oaktreecapital.com/insights/memo/i-beg-to-differ Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The Illusion of Knowledge Howard Marks, The Illusion of Knowledge (2022-09-08), Complete original memo dated September 8, 2022. https://www.oaktreecapital.com/insights/memo/the-illusion-of-knowledge Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · What Really Matters? Howard Marks, What Really Matters? (2022-11-22), Complete original memo dated November 22, 2022. https://www.oaktreecapital.com/insights/memo/what-really-matters Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Sea Change Howard Marks, Sea Change (2022-12-13), Complete memo, sections “Sea Change #1,” “Sea Change #2,” and the closing outlook and 2-to-4-percent range. https://www.oaktreecapital.com/insights/memo/sea-change Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Lessons from Silicon Valley Bank Howard Marks, Lessons from Silicon Valley Bank (2023-04-17), Complete original memo dated April 17, 2023. https://www.oaktreecapital.com/insights/memo/lessons-from-silicon-valley-bank Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Taking the Temperature Howard Marks, Taking the Temperature (2023-07-10), Printed p. 1, five-memo selection rule; pp. 3-4, reserve-fund deployment and $6 billion/$7.5 billion figures; pp. 8-11, method and limits. https://www.oaktreecapital.com/insights/memo/taking-the-temperature Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Further Thoughts on Sea Change Howard Marks, Further Thoughts on Sea Change (2023-10-11), Complete original memo dated October 11, 2023. https://www.oaktreecapital.com/insights/memo/further-thoughts-on-sea-change Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Easy Money Howard Marks, Easy Money (2024-01-09), Printed p. 1, ten years without reader response; p. 14, 3.0-to-3.5-percent guess over the next five to ten years. https://www.oaktreecapital.com/insights/memo/easy-money Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · AI Hurtles Ahead Howard Marks, AI Hurtles Ahead (2026-02-26), Official HTML, opening Claude tutorial and sections on investment implications, AI limitations, and risk posture. https://www.oaktreecapital.com/insights/memo/ai-hurtles-ahead Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · What’s Going on in Private Credit? Howard Marks, What’s Going on in Private Credit? (2026-04-09), Official HTML, sections “Direct Lending and Software,” “What Does the Market Know?,” and “What’s a Manager to Do?”; Oaktree disclosure gives 20-percent and 15-percent exposures. https://www.oaktreecapital.com/insights/memo/whats-going-on-in-private-credit Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Tell Me I’m Wrong Howard Marks, Tell Me I’m Wrong (2010-01-22), Complete original memo dated January 22, 2010. https://www.oaktreecapital.com/docs/default-source/memos/2010-01-22-tell-me-im-wrong.pdf?sfvrsn=2bbb0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · I’d Rather Be Wrong Howard Marks, I’d Rather Be Wrong (2010-03-17), Complete original memo dated March 17, 2010. https://www.oaktreecapital.com/docs/default-source/memos/2010-03-17-id-rather-be-wrong.pdf?sfvrsn=fbb0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Warning Flags Howard Marks, Warning Flags (2010-05-12), Complete original memo dated May 12, 2010. https://www.oaktreecapital.com/docs/default-source/memos/2010-05-12-warning-flags.pdf?sfvrsn=dfbc0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · It’s Greek to Me Howard Marks, It’s Greek to Me (2010-07-19), Complete original memo dated July 19, 2010. https://www.oaktreecapital.com/docs/default-source/memos/2010-07-19-its-greek-to-me.pdf?sfvrsn=27bb0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Hemlines Howard Marks, Hemlines (2010-09-10), Complete original memo dated September 10, 2010. https://www.oaktreecapital.com/docs/default-source/memos/2010-09-10-hemlines.pdf?sfvrsn=7bb0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Open and Shut Howard Marks, Open and Shut (2010-12-01), Complete original memo dated December 1, 2010. https://www.oaktreecapital.com/docs/default-source/memos/2010-12-01-open-and-shut.pdf?sfvrsn=67bb0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · All That Glitters Howard Marks, All That Glitters (2010-12-17), Complete original memo dated December 17, 2010. https://www.oaktreecapital.com/docs/default-source/memos/2010-12-17-all-that-glitters.pdf?sfvrsn=3bb0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · On Regulation Howard Marks, On Regulation (2011-03-02), Complete original memo dated March 2, 2011. https://www.oaktreecapital.com/docs/default-source/memos/2011-03-02-on-regulation.pdf?sfvrsn=37bb0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · How Quickly They Forget Howard Marks, How Quickly They Forget (2011-05-25), Complete original memo dated May 25, 2011. https://www.oaktreecapital.com/docs/default-source/memos/2011-05-25-how-quickly-they-forget.pdf?sfvrsn=33bb0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Down to the Wire Howard Marks, Down to the Wire (2011-07-21), Complete original memo dated July 21, 2011. https://www.oaktreecapital.com/docs/default-source/memos/2011-07-21-down-to-the-wire.pdf?sfvrsn=3fbb0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · What's Behind the Downturn? Howard Marks, What's Behind the Downturn? (2011-09-07), Complete original memo dated September 7, 2011. https://www.oaktreecapital.com/docs/default-source/memos/2011-09-07-whats-behind-the-downturn.pdf?sfvrsn=13bb0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · It's All Very Taxing Howard Marks, It's All Very Taxing (2011-11-16), Complete original memo dated November 16, 2011. https://www.oaktreecapital.com/docs/default-source/memos/2011-11-16-its-all-very-taxing.pdf?sfvrsn=3bbb0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · What Can We Do For You? Howard Marks, What Can We Do For You? (2012-01-10), Complete original memo dated January 10, 2012. https://www.oaktreecapital.com/docs/default-source/memos/2012-01-10-what-can-we-do-for-you.pdf?sfvrsn=7fbb0f65_8 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Assessing Performance Records A Case Study Howard Marks, Assessing Performance Records A Case Study (2012-02-15), Complete original memo dated February 15, 2012. https://www.oaktreecapital.com/docs/default-source/memos/2012-02-15-assessing-performance-records-a-case-study.pdf?sfvrsn=6bbb0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Déjà Vu All Over Again Howard Marks, Déjà Vu All Over Again (2012-03-19), Complete original memo dated March 19, 2012. https://www.oaktreecapital.com/docs/default-source/memos/2012-03-19-déjà-vu-all-over-again.pdf?sfvrsn=bbb0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · It's All A Big Mistake Howard Marks, It's All A Big Mistake (2012-06-20), Complete original memo dated June 20, 2012. https://www.oaktreecapital.com/docs/default-source/memos/2012-06-20-its-all-a-big-mistake.pdf?sfvrsn=17bb0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · On Uncertain Ground Howard Marks, On Uncertain Ground (2012-09-11), Complete original memo dated September 11, 2012. https://www.oaktreecapital.com/docs/default-source/memos/2012-09-11-on-uncertain-ground.pdf?sfvrsn=1fbb0f65_8 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · A Fresh Start (Hopefully) Howard Marks, A Fresh Start (Hopefully) (2012-11-19), Complete original memo dated November 19, 2012. https://www.oaktreecapital.com/docs/default-source/memos/2012-11-09-a-fresh-start-hopefully.pdf?sfvrsn=1bbb0f65_5 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Ditto Howard Marks, Ditto (2013-01-07), Complete original memo dated January 7, 2013. https://www.oaktreecapital.com/docs/default-source/memos/2013-01-07-ditto.pdf?sfvrsn=73bb0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks and Sheldon Stone · High Yield Bonds Today Howard Marks and Sheldon Stone, High Yield Bonds Today (2013-02-21), Complete original memo dated February 21, 2013. https://www.oaktreecapital.com/docs/default-source/memos/2013-02-21-high-yield-bonds-today.pdf?sfvrsn=63bb0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The Outlook For Equities Howard Marks, The Outlook For Equities (2013-03-13), Complete original memo dated March 13, 2013. https://www.oaktreecapital.com/docs/default-source/memos/2013-03-13-the-outlook-for-equities.pdf?sfvrsn=6fbb0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The Role of Confidence Howard Marks, The Role of Confidence (2013-08-05), Complete original memo dated August 5, 2013. https://www.oaktreecapital.com/docs/default-source/memos/2013-08-05-the-role-of-confidence.pdf?sfvrsn=7bbb0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The Race Is On Howard Marks, The Race Is On (2013-11-26), Complete original memo dated November 26, 2013. https://www.oaktreecapital.com/docs/default-source/memos/2013-11-26-the-race-is-on.pdf?sfvrsn=77bb0f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Dare to Be Great II Howard Marks, Dare to Be Great II (2014-04-08), Complete original memo dated April 8, 2014. https://www.oaktreecapital.com/docs/default-source/memos/2014-04-08-dare-to-be-great-ii.pdf?sfvrsn=44b60f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Risk Revisited Howard Marks, Risk Revisited (2014-09-03), Complete original memo dated September 3, 2014. https://www.oaktreecapital.com/docs/default-source/memos/2014-09-03-risk-revisited.pdf?sfvrsn=78b60f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The Lessons of Oil Howard Marks, The Lessons of Oil (2014-12-18), Complete original memo dated December 18, 2014. https://www.oaktreecapital.com/docs/default-source/memos/2014-12-18-the-lessons-of-oil.pdf?sfvrsn=40b60f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Liquidity Howard Marks, Liquidity (2015-03-25), Complete original memo dated March 25, 2015. https://www.oaktreecapital.com/docs/default-source/memos/2015-03-25-liquidity.pdf?sfvrsn=2dc70f65_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Risk Revisited Again Howard Marks, Risk Revisited Again (2015-06-08), Complete original memo dated June 8, 2015. https://www.oaktreecapital.com/docs/default-source/memos/2015-06-08-risk-revisited-again.pdf?sfvrsn=7bb70f65_8 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Inspiration from the World of Sports Howard Marks, Inspiration from the World of Sports (2015-10-22), Complete original memo dated October 22, 2015. https://www.oaktreecapital.com/docs/default-source/memos/2015-10-22-inspiration-from-the-world-of-sports.pdf?sfvrsn=aa1f0e65_7 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · On the Couch Howard Marks, On the Couch (2016-01-14), Complete original memo dated January 14, 2016. https://www.oaktreecapital.com/docs/default-source/memos/on-the-couch.pdf?sfvrsn=8a1e0165_7 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · What Does the Market Know? Howard Marks, What Does the Market Know? (2016-01-19), Complete original memo dated January 19, 2016. https://www.oaktreecapital.com/docs/default-source/memos/what-does-the-market-know.pdf?sfvrsn=cb7a0165_10 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Economic Reality Howard Marks, Economic Reality (2016-05-26), Complete original memo dated May 26, 2016. https://www.oaktreecapital.com/docs/default-source/memos/economic-reality.pdf?sfvrsn=49bc1865_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Political Reality Howard Marks, Political Reality (2016-08-17), Complete original memo dated August 17, 2016. https://www.oaktreecapital.com/docs/default-source/memos/political-reality.pdf?sfvrsn=39391265_2 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Implications of the Election Howard Marks, Implications of the Election (2016-11-07), Complete original memo dated November 7, 2016. https://www.oaktreecapital.com/docs/default-source/memos/implications-of-the-election_11716.pdf?sfvrsn=1389e965_4 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Go Figure! Howard Marks, Go Figure! (2016-11-14), Complete original memo dated November 14, 2016. https://www.oaktreecapital.com/docs/default-source/memos/go-figure.pdf?sfvrsn=ef38e965_7 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Expert Opinion Howard Marks, Expert Opinion (2017-01-10), Complete original memo dated January 10, 2017. https://www.oaktreecapital.com/docs/default-source/memos/expert-opinion.pdf?sfvrsn=ccffe365_4 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Lines in the Sand Howard Marks, Lines in the Sand (2017-04-18), Printed p. 1, client pressure and Oaktree’s adoption of subscription lines; p. 7, the internal guidelines process and its unresolved status. https://www.oaktreecapital.com/docs/default-source/memos/lines-in-the-sand.pdf?sfvrsn=bf5dfa65_2 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · There They Go Again... Again Howard Marks, There They Go Again... Again (2017-07-26), Printed p. 1, admission that caution since 2011 had not been right; pp. 19-22, limits and conditional response. https://www.oaktreecapital.com/docs/default-source/memos/there-they-go-again-again.pdf?sfvrsn=56d4f265_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Yet Again? Howard Marks, Yet Again? (2017-09-07), Printed pp. 4-8, revised Bitcoin framing after conversations; pp. 9-11, response to criticism that the July memo lacked prescriptions. https://www.oaktreecapital.com/docs/default-source/memos/yet-again.pdf?sfvrsn=3767f765_6 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Latest Thinking Howard Marks, Latest Thinking (2018-01-23), Complete original memo dated January 23, 2018. https://www.oaktreecapital.com/docs/default-source/memos/latest-thinking.pdf?sfvrsn=a77ccf65_2 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Investing Without People Howard Marks, Investing Without People (2018-06-18), Printed p. 1, nonexpert caveat; pp. 12 and 14-17, Andrew’s challenges, limits of active management, provisional human advantages, and the corrected Einstein attribution. https://www.oaktreecapital.com/docs/default-source/memos/investing-without-people.pdf?sfvrsn=7a5ec465_8 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The Seven Worst Words in the World Howard Marks, The Seven Worst Words in the World (2018-09-26), Complete original memo dated September 26, 2018. https://www.oaktreecapital.com/docs/default-source/memos/the-seven-worst-words-in-the-world.pdf?sfvrsn=6dc9dd65_4 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Political Reality Meets Economic Reality Howard Marks, Political Reality Meets Economic Reality (2019-01-30), Complete original memo dated January 30, 2019. https://www.oaktreecapital.com/docs/default-source/memos/political-reality-meets-economic-reality.pdf?sfvrsn=21c1d665_8 Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The Indispensability of Risk Howard Marks, The Indispensability of Risk (2024-04-17), Complete original memo dated April 17, 2024. https://www.oaktreecapital.com/insights/memo/the-indispensability-of-risk Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The Impact of Debt Howard Marks, The Impact of Debt (2024-05-08), Complete original memo dated May 8, 2024. https://www.oaktreecapital.com/insights/memo/the-impact-of-debt Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The Folly of Certainty Howard Marks, The Folly of Certainty (2024-07-17), Complete original memo dated July 17, 2024. https://www.oaktreecapital.com/insights/memo/the-folly-of-certainty Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Mr. Market Miscalculates Howard Marks, Mr. Market Miscalculates (2024-08-22), Printed p. 5, correction after Buffett denied making the repeatedly attributed profit-growth warning. https://www.oaktreecapital.com/insights/memo/mr-market-miscalculates Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Shall We Repeal the Laws of Economics? Howard Marks, Shall We Repeal the Laws of Economics? (2024-09-19), Complete original memo dated September 19, 2024. https://www.oaktreecapital.com/insights/memo/shall-we-repeal-the-laws-of-economics Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Ruminating on Asset Allocation Howard Marks, Ruminating on Asset Allocation (2024-10-22), Complete original memo dated October 22, 2024. https://www.oaktreecapital.com/insights/memo/ruminating-on-asset-allocation Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · On Bubble Watch Howard Marks, On Bubble Watch (2025-01-07), Complete original memo dated January 7, 2025. https://www.oaktreecapital.com/insights/memo/on-bubble-watch Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Gimme Credit Howard Marks, Gimme Credit (2025-03-06), Complete original memo dated March 6, 2025. https://www.oaktreecapital.com/insights/memo/gimme-credit Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Nobody Knows (Yet Again) Howard Marks, Nobody Knows (Yet Again) (2025-04-09), Complete original memo dated April 9, 2025. https://www.oaktreecapital.com/insights/memo/nobody-knows-yet-again Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · More on Repealing the Laws of Economics Howard Marks, More on Repealing the Laws of Economics (2025-06-18), Complete original memo dated June 18, 2025. https://www.oaktreecapital.com/insights/memo/more-on-repealing-the-laws-of-economics Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The Calculus of Value Howard Marks, The Calculus of Value (2025-08-14), Complete original memo dated August 14, 2025. https://www.oaktreecapital.com/insights/memo/the-calculus-of-value Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The Best of . . . Howard Marks, The Best of . . . (2025-10-12), Complete original memo dated October 12, 2025. https://www.oaktreecapital.com/insights/memo/the-best-of Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · A Look Under the Hood Howard Marks, A Look Under the Hood (2025-10-28), Complete original memo dated October 28, 2025. https://www.oaktreecapital.com/insights/memo/a-look-under-the-hood Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Cockroaches in the Coal Mine Howard Marks, Cockroaches in the Coal Mine (2025-11-06), Complete original memo dated November 6, 2025. https://www.oaktreecapital.com/insights/memo/cockroaches-in-the-coal-mine Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · Is It a Bubble? Howard Marks, Is It a Bubble? (2025-12-09), Complete original memo dated December 9, 2025. https://www.oaktreecapital.com/insights/memo/is-it-a-bubble Evidence note: Read directly in the official Oaktree complete collection or official memo page. ### Howard Marks · The Most Important Thing: Uncommon Sense for the Thoughtful Investor Howard Marks, The Most Important Thing: Uncommon Sense for the Thoughtful Investor (2011-05-01), Publisher page, description, contents, and listed Chapter 2 excerpt. https://cup.columbia.edu/book/the-most-important-thing/9780231153683/ Evidence note: Publisher record inspected. The full book was not locally available and was not read. ### Howard Marks · The Most Important Thing Illuminated Howard Marks, The Most Important Thing Illuminated (2013-01-01), Publisher page, description and contents. https://cup.columbia.edu/book/the-most-important-thing-illuminated/9780231162845/ Evidence note: Publisher record inspected. The full annotated edition was not locally available and was not read. ### Howard Marks · Mastering the Market Cycle: Getting the Odds on Your Side Howard Marks, Mastering the Market Cycle: Getting the Odds on Your Side (2018-10-02), Publisher page. https://www.hachette.co.uk/titles/howard-marks/mastering-the-market-cycle/9781473695689/ Evidence note: Publisher record inspected. The full book was not locally available and was not read. ### Howard Marks, Bruce Karsh, and Sheldon Stone · 30 Years of Oaktree Howard Marks, Bruce Karsh, and Sheldon Stone, 30 Years of Oaktree (2025-04-30), Official eight-page transcript, pp. 1-3 on formation and early funds, pp. 4-6 on partnership and operating choices, and pp. 7-8 on succession. https://www.oaktreecapital.com/insights/insight-podcast/education/30-years-of-oaktree-with-howard-marks-bruce-karsh-and-sheldon-stone Evidence note: Read in full. The founders reconstruct their partnership, early funds, culture, mistakes avoided and succession aims. ### Mark Leonard · Constellation Software 2007 Q1 president’s letter Mark Leonard, Constellation Software 2007 Q1 president’s letter (2007-05-08), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q1_2007_shareholders_report.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. ### Mark Leonard · Constellation Software 2007 Q2 president’s letter Mark Leonard, Constellation Software 2007 Q2 president’s letter (2007 Q2 reporting period; undated letter), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q2_2007_shareholders_report.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. ### Mark Leonard · Constellation Software 2007 Q3 president’s letter Mark Leonard, Constellation Software 2007 Q3 president’s letter (2007 Q3 reporting period; undated letter), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q3_2007_shareholders_report.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. ### Mark Leonard · Constellation Software 2007 Q4 president’s letter Mark Leonard, Constellation Software 2007 Q4 president’s letter (2008-03-05), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q4_2007_shareholders_report.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. ### Mark Leonard · Constellation Software 2008 Q1 president’s letter Mark Leonard, Constellation Software 2008 Q1 president’s letter (2008-05-07), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q1_2008_shareholdersreportf49d4032f35645668472bbd0e4d94b0b.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. ### Mark Leonard · Constellation Software 2008 Q2 president’s letter Mark Leonard, Constellation Software 2008 Q2 president’s letter (2008-08-07), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q2_2008_shareholdersreport.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. ### Mark Leonard · Constellation Software 2008 Q3 president’s letter Mark Leonard, Constellation Software 2008 Q3 president’s letter (2008-11-06), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidentsletter50ed1eb201c049b287838ec5e943432b.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. ### Mark Leonard · Constellation Software 2008 Q4 president’s letter Mark Leonard, Constellation Software 2008 Q4 president’s letter (2009-03-04), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q4_2008presidentsletter.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. ### Mark Leonard · Constellation Software 2009 Q1 president’s letter Mark Leonard, Constellation Software 2009 Q1 president’s letter (2009-05-06), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q1_2009_presidentletter.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. ### Mark Leonard · Constellation Software 2009 annual president’s letter Mark Leonard, Constellation Software 2009 annual president’s letter (2010-03-25), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/shareholder-letter-2009.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. ### Mark Leonard · Constellation Software 2010 annual president’s letter Mark Leonard, Constellation Software 2010 annual president’s letter (2011-05-02), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/shareholder-letter-2010.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. ### Mark Leonard · Constellation Software 2011 annual president’s letter Mark Leonard, Constellation Software 2011 annual president’s letter (2012-05-02), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/2011_presidents_letter.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. ### Mark Leonard · Constellation Software 2012 annual president’s letter Mark Leonard, Constellation Software 2012 annual president’s letter (2013-05-01), Complete letter, including tables and metric glossary. Relevant passage and page are identified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-final.pdf Evidence note: Contemporaneous management account unless explicitly identified as retrospective. Reporting period differs from publication date. Company-defined returns are not standardized ROIC. ### Jamal Baksh with Tegus · May 26, 2022 · CFO Interview Jamal Baksh with Tegus, May 26, 2022 · CFO Interview (2022-04-06; posted May 26, 2022), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/April-6-2022-Tegus-interview-with-CFO.pdf Evidence note: Full seven-page company-reviewed interview read. Baksh is the speaker throughout; this is not Leonard’s testimony. It explains VMS Ventures, delegated acquisitions and limits to outside-sector investment. ### Constellation Software; named respondents vary · September 17, 2021 · Members Agreement, Acquisitions, NCI and Dividends Constellation Software; named respondents vary, September 17, 2021 · Members Agreement, Acquisitions, NCI and Dividends (2021-09-17), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-september-2021-final.pdf Evidence note: Full Q&A read. Leonard discusses Topicus option horizons and judging managers through customer and employee records. Jamal answers accounting and tax questions. ### Constellation Software; named respondents vary · August 17, 2020 · Margins, Bonus Plan, Contour, Organic Growth, Debentures, External Boards Constellation Software; named respondents vary, August 17, 2020 · Margins, Bonus Plan, Contour, Organic Growth, Debentures, External Boards (2020-08-17), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-august-2020-final.pdf Evidence note: Full Q&A read. Jamal explains pandemic margins, bonus metrics and large-deal thresholds. Leonard responds on debt availability and outside boards. ### Constellation Software; named respondents vary · NOTICE TO SHAREHOLDERS REGARDING PREVIOUSLY SUBMITTED QUESTIONS Constellation Software; named respondents vary, NOTICE TO SHAREHOLDERS REGARDING PREVIOUSLY SUBMITTED QUESTIONS (2019-06-14), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q-a-june-2019-lost-submissions.pdf Evidence note: Full administrative notice read. Records deletion of submitted questions; it contains no investing lesson attributed to Leonard. ### Mark Leonard and named Constellation respondents · September 19, 2018 · Additional sectors for investment, Buybacks, ROIC, Value Investing Mark Leonard and named Constellation respondents, September 19, 2018 · Additional sectors for investment, Buybacks, ROIC, Value Investing (2018-09-19), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-september-2018-final.pdf Evidence note: Full Q&A read. Leonard qualifies his buyback position and distinguishes beliefs from certainty; Jamal supplies the invested-capital reconciliation. ### Mark Leonard and named Constellation respondents · February 20, 2019 · Special Dividend Mark Leonard and named Constellation respondents, February 20, 2019 · Special Dividend (2019-02-20), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/q-a-february-2019-divrelated-final.pdf Evidence note: Full Q&A read. Leonard defends the 2019 special dividend, magnetic hurdles and employee sharing of excess returns. ### Mark Leonard and named Constellation respondents · November 5, 2018 · License revenue volatility, CSI’s biggest failure, Business systems Mark Leonard and named Constellation respondents, November 5, 2018 · License revenue volatility, CSI’s biggest failure, Business systems (2018-11-05), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-nov-5-2018-final-1.pdf Evidence note: Full Q&A read. Leonard names excess early equity and a business sale as failures. Jamal explains license volatility. ### Mark Leonard and named Constellation respondents · October 9, 2018 · Moat, SaaS, TAM, corporate culture Mark Leonard and named Constellation respondents, October 9, 2018 · Moat, SaaS, TAM, corporate culture (2018-10-09), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-oct-9-2018-final-1.pdf Evidence note: Full Q&A read. Leonard discusses SaaS value capture and plural cultures. The club-software example belongs to Barry Symons. ### Mark Leonard and named Constellation respondents · October 4, 2018 · Employee attraction, retention, motivation and engagement Mark Leonard and named Constellation respondents, October 4, 2018 · Employee attraction, retention, motivation and engagement (2018-10-04), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-oct-4-2018-final.pdf Evidence note: Full Q&A read. Leonard questions engagement measurement and discusses opportunity, empathy and supervisor quality. ### Mark Leonard and named Constellation respondents · August 3, 2018 · HPC’s, Blackboard, Customer Acquisition Economics, Margin Trajectory, Organic Growth Profile, Tax Rate, SaaS vs On-Premise Mark Leonard and named Constellation respondents, August 3, 2018 · HPC’s, Blackboard, Customer Acquisition Economics, Margin Trajectory, Organic Growth Profile, Tax Rate, SaaS vs On-Premise (2018-08-03), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-august-2018-final.pdf Evidence note: Full Q&A read. Leonard admits selection bias in the conglomerate study; other responses cover margins, growth, taxes and SaaS. ### Mark Leonard and named Constellation respondents · July 25, 2018 · Employee Retention, Consensus, IFTODH, Bias, Incentive Compensation Mark Leonard and named Constellation respondents, July 25, 2018 · Employee Retention, Consensus, IFTODH, Bias, Incentive Compensation (2018-07-25), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-july-25-2018-final.pdf Evidence note: Full Q&A read. Leonard discusses skeptics, formulaic incentives, downside borne by homebuilding managers and uncertainty in supposedly private facts. ### Mark Leonard and named Constellation respondents · June 28, 2018 · Excess Capital, Dividends, Buybacks, Leverage Mark Leonard and named Constellation respondents, June 28, 2018 · Excess Capital, Dividends, Buybacks, Leverage (Questions received through 2018-06-26; posted June 28), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-june-2018-final.pdf Evidence note: Full Q&A read. Leonard distinguishes reliable moderate debt, dividends and moral concerns over buybacks. ### Constellation Software; named respondents vary · May 11, 2018 · Internal Competition, TSS Acquisitions Constellation Software; named respondents vary, May 11, 2018 · Internal Competition, TSS Acquisitions (Questions received through 2018-05-11), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-may-14-2018-final.pdf Evidence note: Full Q&A read. Corporate responses distinguish customer competition from acquisition coordination and explain TSS minority interests. ### Constellation Software; named respondents vary · April 19, 2018 · M&A, Cyber, Competition, Emulators, TSS Minority Interest, ANI Calculation, Organizational Structure Constellation Software; named respondents vary, April 19, 2018 · M&A, Cyber, Competition, Emulators, TSS Minority Interest, ANI Calculation, Organizational Structure (Questions received through 2018-04-19), Complete document; use the named question and respondent specified in the reading.. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-april-20-2018-final-1.pdf Evidence note: Full Q&A read. Unsigned corporate responses address acquisition capacity, minority claims, accounting assumptions and organizational flexibility. ### Constellation Software · Mark Leonard resignation and Mark Miller appointment Constellation Software, Mark Leonard resignation and Mark Miller appointment (2025-09-25), Announcement body and separately attributed statements.. https://www.csisoftware.com/constellation-software-inc.-announces-the-resignation-of-mark-leonard-and-appointment-of-mark-miller-as-president-of-constellation-software/ Evidence note: Company announcement establishes the leadership change and stated health reason. Expressions of confidence are expectations, not proof of succession outcomes. ### Constellation Software · Mark Leonard decision not to stand for board re-election Constellation Software, Mark Leonard decision not to stand for board re-election (2026-03-27), Announcement body and separately attributed statements.. https://www.csisoftware.com/constellation-software-inc-announces-mark-leonards-decision-to-not-stand-for-re-election-to-board-of-directors/ Evidence note: Company announcement gives the end-of-term plan and continuing advisory role focused on PEMS. The company’s retrospective praise is not an independent assessment. ## Works inventory · research checked 2026-09-09 Reading status describes the course research, not your personal reading progress. Located and unavailable works have not been counted as full reads. ### Warren Buffett · 1957 partnership letter Author or speaker: Warren E. Buffett Date or period: 1957 reporting year; circulated 1958 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · 1958 partnership letter Author or speaker: Warren E. Buffett Date or period: 1958 reporting year; circulated 1959 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · 1959 partnership letter Author or speaker: Warren E. Buffett Date or period: 1959 reporting year; signed February 20, 1960 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · 1960 partnership letter Author or speaker: Warren E. Buffett Date or period: 1960 reporting year; signed January 30, 1961 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · Partnership letter Author or speaker: Warren E. Buffett Date or period: July 22, 1961 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · 1961 partnership letter Author or speaker: Warren E. Buffett Date or period: 1961 reporting year; signed January 24, 1962 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · Partnership letter Author or speaker: Warren E. Buffett Date or period: July 6, 1962 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · Partnership letter Author or speaker: Warren E. Buffett Date or period: November 1, 1962 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · Ground Rules and Dempster partnership letter Author or speaker: Warren E. Buffett Date or period: January 18, 1963 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · Partnership letter Author or speaker: Warren E. Buffett Date or period: July 10, 1963 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · Partnership letter Author or speaker: Warren E. Buffett Date or period: November 6, 1963 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · 1963 partnership letter Author or speaker: Warren E. Buffett Date or period: 1963 reporting year; signed January 18, 1964 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · Partnership letter Author or speaker: Warren E. Buffett Date or period: July 8, 1964 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · 1964 partnership letter Author or speaker: Warren E. Buffett Date or period: 1964 reporting year; signed January 18, 1965 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · Partnership letter Author or speaker: Warren E. Buffett Date or period: July 9, 1965 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · Partnership letter Author or speaker: Warren E. Buffett Date or period: November 1, 1965 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · 1965 partnership letter Author or speaker: Warren E. Buffett Date or period: 1965 reporting year; signed January 20, 1966 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · Partnership letter Author or speaker: Warren E. Buffett Date or period: July 12, 1966 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · 1966 partnership letter Author or speaker: Warren E. Buffett Date or period: 1966 reporting year; signed January 25, 1967 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · Partnership letter Author or speaker: Warren E. Buffett Date or period: July 12, 1967 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · Partnership letter Author or speaker: Warren E. Buffett Date or period: October 9, 1967 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · 1967 partnership letter Author or speaker: Warren E. Buffett Date or period: 1967 reporting year; signed January 24, 1968 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · Partnership letter Author or speaker: Warren E. Buffett Date or period: July 11, 1968 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · 1968 partnership letter Author or speaker: Warren E. Buffett Date or period: 1968 reporting year; signed January 22, 1969 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · Partnership letter Author or speaker: Warren E. Buffett Date or period: May 29, 1969 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · Partnership letter Author or speaker: Warren E. Buffett Date or period: October 9, 1969 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · Partnership letter Author or speaker: Warren E. Buffett Date or period: December 5, 1969 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · Partnership letter Author or speaker: Warren E. Buffett Date or period: December 26, 1969 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · Final partnership letter Author or speaker: Warren E. Buffett Date or period: February 25, 1970 Status: Read in full Read directly in full from the dateline through the closing and any included appendix. The complete document was reviewed, not only the passages used in the chapters. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · 1965 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett; signed by Malcolm G. Chace Jr. and/or Kenneth V. Chace Date or period: 1965 reporting year; published 1966 Status: Read in full Read directly in full from the opening through the signature. The 1965 report was reviewed as a complete contemporary account of the textile, insurance, acquisition and capital-allocation position for that reporting year. The report is signed by a Chace officer and the local original attributes the letter text to Buffett. https://www.berkshirehathaway.com/letters/letters.html ### Warren Buffett · 1966 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett; signed by Malcolm G. Chace Jr. and/or Kenneth V. Chace Date or period: 1966 reporting year; published 1967 Status: Read in full Read directly in full from the opening through the signature. The 1966 report was reviewed as a complete contemporary account of the textile, insurance, acquisition and capital-allocation position for that reporting year. The report is signed by a Chace officer and the local original attributes the letter text to Buffett. https://www.berkshirehathaway.com/letters/letters.html ### Warren Buffett · 1967 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett; signed by Malcolm G. Chace Jr. and/or Kenneth V. Chace Date or period: 1967 reporting year; published 1968 Status: Read in full Read directly in full from the opening through the signature. The 1967 report was reviewed as a complete contemporary account of the textile, insurance, acquisition and capital-allocation position for that reporting year. The report is signed by a Chace officer and the local original attributes the letter text to Buffett. https://www.berkshirehathaway.com/letters/letters.html ### Warren Buffett · 1968 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett; signed by Malcolm G. Chace Jr. and/or Kenneth V. Chace Date or period: 1968 reporting year; published 1969 Status: Read in full Read directly in full from the opening through the signature. The 1968 report was reviewed as a complete contemporary account of the textile, insurance, acquisition and capital-allocation position for that reporting year. The report is signed by a Chace officer and the local original attributes the letter text to Buffett. https://www.berkshirehathaway.com/letters/letters.html ### Warren Buffett · 1969 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett; signed by Malcolm G. Chace Jr. and/or Kenneth V. Chace Date or period: 1969 reporting year; published 1970 Status: Read in full Read directly in full from the opening through the signature. The 1969 report was reviewed as a complete contemporary account of the textile, insurance, acquisition and capital-allocation position for that reporting year. The report is signed by a Chace officer and the local original attributes the letter text to Buffett. https://www.berkshirehathaway.com/letters/letters.html ### Warren Buffett · 1970 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1970 reporting year; published 1971 Status: Read in full Read directly in full from the opening through the signature. The 1970 report was reviewed as a complete contemporary account of the textile, insurance, acquisition and capital-allocation position for that reporting year. https://www.berkshirehathaway.com/letters/letters.html ### Warren Buffett · 1971 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1971 reporting year; published 1972 Status: Read in full Read directly in full from the opening through the signature. The 1971 report was reviewed as a complete contemporary account of the textile, insurance, acquisition and capital-allocation position for that reporting year. https://www.berkshirehathaway.com/letters/letters.html ### Warren Buffett · 1972 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1972 reporting year; published 1973 Status: Read in full Read directly in full from the opening through the signature. The 1972 report was reviewed as a complete contemporary account of the textile, insurance, acquisition and capital-allocation position for that reporting year. https://www.berkshirehathaway.com/letters/letters.html ### Warren Buffett · 1973 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1973 reporting year; published 1974 Status: Read in full Read directly in full from the opening through the signature. The 1973 report was reviewed as a complete contemporary account of the textile, insurance, acquisition and capital-allocation position for that reporting year. https://www.berkshirehathaway.com/letters/letters.html ### Warren Buffett · 1974 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1974 reporting year; published 1975 Status: Read in full Read directly in full from the opening through the signature. The 1974 report was reviewed as a complete contemporary account of the textile, insurance, acquisition and capital-allocation position for that reporting year. https://www.berkshirehathaway.com/letters/letters.html ### Warren Buffett · 1975 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1975 reporting year; published 1976 Status: Read in full Read directly in full from the opening through the signature. The 1975 report was reviewed as a complete contemporary account of the textile, insurance, acquisition and capital-allocation position for that reporting year. https://www.berkshirehathaway.com/letters/letters.html ### Warren Buffett · 1976 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1976 reporting year; published 1977 Status: Read in full Read directly in full from the opening through the signed closing in the locally extracted original annual letter. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/letters.html ### Warren Buffett · 1977 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1977 reporting year; published 1978 Status: Read in full Read directly in full from the opening through the signed closing in the locally extracted original annual letter. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/1977.html ### Warren Buffett · 1978 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1978 reporting year; published 1979 Status: Read in full Read directly in full from the opening through the signed closing in the locally extracted original annual letter. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/1978.html ### Warren Buffett · 1979 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1979 reporting year; published 1980 Status: Read in full Read directly in full from the opening through the signed closing in the locally extracted original annual letter. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/1979.html ### Warren Buffett · 1980 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1980 reporting year; published 1981 Status: Read in full Read directly in full from the opening through the signed closing in the locally extracted original annual letter. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/1980.html ### Warren Buffett · 1981 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1981 reporting year; published 1982 Status: Read in full Read directly in full from the opening through the signed closing in the locally extracted original annual letter. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/1981.html ### Warren Buffett · 1982 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1982 reporting year; published 1983 Status: Read in full Read directly in full from the opening through the signed closing in the locally extracted original annual letter. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/1982.html ### Warren Buffett · 1983 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1983 reporting year; published 1984 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/1983.html ### Warren Buffett · 1984 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1984 reporting year; published 1985 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/1984.html ### Warren Buffett · 1985 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1985 reporting year; published 1986 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/1985.html ### Warren Buffett · 1986 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1986 reporting year; published 1987 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/1986.html ### Warren Buffett · 1987 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1987 reporting year; published 1988 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/1987.html ### Warren Buffett · 1988 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1988 reporting year; published 1989 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/1988.html ### Warren Buffett · 1989 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1989 reporting year; published 1990 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/1989.html ### Warren Buffett · 1990 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1990 reporting year; published 1991 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/1990.html ### Warren Buffett · 1991 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1991 reporting year; published 1992 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/1991.html ### Warren Buffett · 1992 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1992 reporting year; published 1993 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/1992.html ### Warren Buffett · 1993 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1993 reporting year; published 1994 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/1993.html ### Warren Buffett · 1994 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1994 reporting year; published 1995 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/1994.html ### Warren Buffett · 1995 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1995 reporting year; published 1996 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/1995.html ### Warren Buffett · 1996 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1996 reporting year; published 1997 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/1996.html ### Warren Buffett · 1997 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1997 reporting year; published 1998 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/1997.html ### Warren Buffett · 1998 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1998 reporting year; published 1999 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/1998pdf.pdf ### Warren Buffett · 1999 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 1999 reporting year; published 2000 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/final1999pdf.pdf ### Warren Buffett · 2000 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2000 reporting year; published 2001 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/2000pdf.pdf ### Warren Buffett · 2001 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2001 reporting year; published 2002 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/2001ar/2001letter.html ### Warren Buffett · 2002 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2002 reporting year; published 2003 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/2002pdf.pdf ### Warren Buffett · 2003 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2003 reporting year; published 2004 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/2003ltr.pdf ### Warren Buffett · 2004 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2004 reporting year; published 2005 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/2004ltr.pdf ### Warren Buffett · 2005 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2005 reporting year; published 2006 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/2005ltr.pdf ### Warren Buffett · 2006 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2006 reporting year; published 2007 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/2006ltr.pdf ### Warren Buffett · 2007 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2007 reporting year; published 2008 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/2007ltr.pdf ### Warren Buffett · 2008 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2008 reporting year; published 2009 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/2008ltr.pdf ### Warren Buffett · 2009 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2009 reporting year; published 2010 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/2009ltr.pdf ### Warren Buffett · 2010 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2010 reporting year; published 2011 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/2010ltr.pdf ### Warren Buffett · 2011 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2011 reporting year; published 2012 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/2011ltr.pdf ### Warren Buffett · 2012 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2012 reporting year; published 2013 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/2012ltr.pdf ### Warren Buffett · 2013 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2013 reporting year; published 2014 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/2013ltr.pdf ### Warren Buffett · 2014 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2014 reporting year; published 2015 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/2014ltr.pdf ### Warren Buffett · 2015 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2015 reporting year; published 2016 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/2015ltr.pdf ### Warren Buffett · 2016 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2016 reporting year; published 2017 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/2016ltr.pdf ### Warren Buffett · 2017 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2017 reporting year; published 2018 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/2017ltr.pdf ### Warren Buffett · 2018 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2018 reporting year; published 2019 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/2018ltr.pdf ### Warren Buffett · 2019 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2019 reporting year; published 2020 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/2019ltr.pdf ### Warren Buffett · 2020 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2020 reporting year; published 2021 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/2020ltr.pdf ### Warren Buffett · 2021 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2021 reporting year; published 2022 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/2021ltr.pdf ### Warren Buffett · 2022 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2022 reporting year; published 2023 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/2022ltr.pdf ### Warren Buffett · 2023 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2023 reporting year; published 2024 Status: Read in full Read directly in full from the opening through the signed closing and any Buffett-authored appendix in the locally extracted original annual report. The review included every section, not only passages used in the course. https://www.berkshirehathaway.com/letters/2023ltr.pdf ### Warren Buffett · 2024 Berkshire Hathaway annual letter Author or speaker: Warren E. Buffett Date or period: 2024 reporting year; published 2025 Status: Read in full Read directly in full from the complete local original. It makes mistakes, delayed correction, Berkshire's sixty-year record and succession central to Buffett's final annual letter. https://www.berkshirehathaway.com/letters/2024ltr.pdf ### Warren Buffett · Thanksgiving Message to Fellow Shareholders Author or speaker: Warren E. Buffett Date or period: November 10, 2025 Status: Read in full Read directly in full. Buffett covers succession, accelerated giving and a delayed response to cognitive decline among managers. It is the latest Buffett-authored material located on Berkshire's official site as of September 9, 2026. https://berkshirehathaway.com/news/nov1025.pdf ### Warren Buffett · How Inflation Swindles the Equity Investor Author or speaker: Warren E. Buffett Date or period: May 1, 1977 Status: Read in part Read the original publisher's accessible text on sticky returns on equity, dividends and the capital needed merely to preserve physical output. A fresh complete extraction was not available, so no full-read claim is made. https://fortune.com/article/buffett-how-inflation-swindles-the-equity-investor-fortune-classics-1977/ ### Warren Buffett · The Superinvestors of Graham-and-Doddsville Author or speaker: Warren E. Buffett Date or period: May 17, 1984 Status: Read in part Read Columbia's accessible article text and performance tables. It supplies Buffett's case against pure chance and his 1984 account of investment risk, but the page extraction was not complete enough to support a full-read claim. https://business.columbia.edu/insights/chazen-global-insights/superinvestors-graham-and-doddsville ### Warren Buffett · Mr. Buffett on the Stock Market Author or speaker: Warren E. Buffett; edited by Carol Loomis Date or period: November 22, 1999 Status: Read in part Read the original publisher's available text on the two seventeen-year periods, interest rates, profits and expectations. The work is used as a valuation argument, not a precise market-timing forecast. https://fortune.com/article/warren-buffett-on-stock-market/ ### Warren Buffett · Buy American. I Am. Author or speaker: Warren E. Buffett Date or period: October 17, 2008 Status: Read in part Located at the original publisher and reviewed through accessible text records. Robots restrictions prevented a fresh complete extraction, so this is not claimed as a full read. https://www.nytimes.com/2008/10/17/opinion/17buffett.html ### Warren Buffett · Salomon Brothers congressional testimony Author or speaker: Warren E. Buffett Date or period: September 4-5, 1991 Status: Read in part Reviewed the recorded opening, apology, cooperation commitment and conduct standard. A complete official hearing transcript was not located and no full-record claim is made. https://www.youtube.com/watch?v=MtaeGt3KwuA ### Warren Buffett · Warren Buffett Archive annual meetings Author or speaker: Warren E. Buffett, Charlie Munger and other Berkshire participants Date or period: 1994-2025 archive span Status: Read in part Inventoried the authorized CNBC series and reviewed selected succession and development passages, including the 2025 announcement. The thirty-three-meeting archive was not watched or read in full. https://buffett.cnbc.com/annual-meetings/ ### Warren Buffett · 2025 Berkshire Hathaway annual letter Author or speaker: Greg Abel Date or period: 2025 reporting year; published February 28, 2026 Status: Read in part Read the opening authorship, stewardship and culture passages. This is inventoried to mark the boundary of Buffett authorship, not treated as another Buffett work. https://www.berkshirehathaway.com/letters/2025ltr.pdf ### Warren Buffett · The Security I Like Best Author or speaker: Warren E. Buffett Date or period: December 6, 1951 Status: Located, not read The Commercial and Financial Chronicle archive was located, but a verified complete scan of the exact article was not extracted in this pass. The course does not quote it or claim a full read. https://archive.org/details/pub_commercial-and-financial-chronicle ### Warren Buffett · The Essays of Warren Buffett: Lessons for Corporate America Author or speaker: Warren E. Buffett; arranged by Lawrence A. Cunningham Date or period: 1997 onward; multiple editions Status: Unavailable No authorized full local or freely accessible copy was available. Because the collection rearranges annual-letter material by topic, it was inventoried but not used as evidence of chronology. https://www.panmacmillan.com/authors/lawrence-a-cunningham/the-essays-of-warren-buffett-8th-ed/9781804092774 ### Warren Buffett · Tap Dancing to Work Author or speaker: Carol J. Loomis, with twelve Buffett-authored pieces Date or period: 2012 Status: Unavailable The publisher record confirms that the collection contains twelve Buffett-authored pieces and Loomis commentary. No full copy was available locally, so its contents are not claimed as read. https://www.penguinrandomhouse.com/books/312515/tap-dancing-to-work-by-collected-and-expanded-by-carol-j-loomis/ ### Warren Buffett · The Snowball: Warren Buffett and the Business of Life Author or speaker: Alice Schroeder Date or period: 2008 Status: Unavailable The publisher page and author interview were reviewed, but the 976-page full text was not locally available. Personal-development claims that require this biography remain outside the taught narrative. https://www.penguinrandomhouse.com/books/162310/the-snowball-by-alice-schroeder/9780553905496/ ### Warren Buffett · Buffett: The Making of an American Capitalist Author or speaker: Roger Lowenstein Date or period: 1995 Status: Unavailable The publisher description was located, but the full book was not available. It is an independent biographical gap, especially for family, interpersonal and early-career context. https://www.penguinrandomhouse.com/books/103994/buffett-by-roger-lowenstein/ ### Warren Buffett · Warren Buffett's Ground Rules Author or speaker: Jeremy C. Miller Date or period: 2016 Status: Unavailable The publisher page and preview were located. The book was not read in full; direct partnership letters, not Miller's synthesis, support the early chapters. https://profilebooks.com/work/warren-buffetts-ground-rules/ ### Warren Buffett · Berkshire Beyond Buffett Author or speaker: Lawrence A. Cunningham Date or period: 2014 Status: Unavailable The publisher page was located, but no full text was available. Claims about Berkshire culture after Buffett therefore rely on primary succession documents and remain narrower than a full organizational study. https://cup.columbia.edu/book/berkshire-beyond-buffett/9780231170048/ ### Warren Buffett · Berkshire Hathaway Letters to Shareholders: 1965-2024 Author or speaker: Warren E. Buffett Date or period: Published 2025 Status: Located, not read The official complete printed collection was located through Berkshire's archive page. It was not obtained. Its contents duplicate the annual series inventoried from local originals, but the book itself is not claimed as read. https://shop.ingramspark.com/b/084?params=RWMLgPMF5zpy6nosCqjAEDrnEGv4uQf6GDilKrIVWXt ### Warren Buffett · First annual partnership letter referenced by the 1957 letter Author or speaker: Warren E. Buffett Date or period: 1956 reporting year Status: Unavailable The 1957 document labels itself the second annual letter and reproduces part of the prior letter, but the 1956 original is not present in the located compilation. The quoted fragment was read only within the 1957 letter. https://focusedcompounding.com/wp-content/uploads/2018/04/Complete_Buffett_partnership_letters-1957-70.pdf ### Warren Buffett · Talk to University of Florida MBA students Author or speaker: Warren E. Buffett Date or period: 1998 Status: Located, not read The hosting institution confirms the 1998 visit, but a complete original recording or institution transcript was not reviewed. No chapter relies on popular transcript reproductions of the talk. https://warrington.ufl.edu/news/warren-buffett-to-mba-students-this-is-what-sets-apart-a-big-winner-from-the-rest-of-the-pack/ ### Warren Buffett · Financial Crisis Inquiry Commission testimony on credit ratings Author or speaker: Warren E. Buffett Date or period: June 2, 2010 Status: Located, not read The preserved FCIC hearing page, official transcript link and video were located. Buffett provided no written statement, and the complete oral session was not read or watched in this pass. https://fcic.law.stanford.edu/hearings/testimony/credibility-of-credit-ratings-the-investment-decisions ### Warren Buffett · Letter to Charles N. Huggins at See’s Candy Shops Author or speaker: Warren E. Buffett Date or period: December 13, 1972 Status: Read in part Read the complete partial OCR returned for the first two letter pages through the Buffett signature. The five-page scan itself was not fully accessible or visually verified, and OCR errors preclude quotations and precise numerical use. https://theoraclesclassroom.com/wp-content/uploads/2019/10/1972-Buffett-Letter-to-Sees-Candies.pdf ### Warren Buffett · Letter to George D. Young about GEICO Author or speaker: Warren E. Buffett Date or period: July 22, 1976 Status: Read in part Read the complete partial OCR returned for pages 1-3. It ends before the fourth scan page and visibly corrupts figures, so it supports only bounded paraphrase and no precise quantities or quotations. https://theoraclesclassroom.com/wp-content/uploads/2020/10/Buffett-Letter-about-GEICO-1976.pdf ### Warren Buffett · Early Commercial & Financial Chronicle investment write-ups Author or speaker: Warren E. Buffett Date or period: Early career; exact constituent dates unverified Status: Located, not read The independent archive lists a five-page scan, but local and browser retrieval failed and no readable extraction was returned. Constituent titles and dates remain unverified. https://theoraclesclassroom.com/wp-content/uploads/2020/10/Buffett-Old-invesetment-writeups-Commercial-_-Financial-Chronicle.pdf ### Jeff Bezos · 1997 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 1997 reporting period; published 1998 Status: Read in full Complete signed current-period letter read. It declares the long-term tradeoffs and records customer, capacity, hiring and capital facts available before later success. https://www.aboutamazon.com/news/company-news/amazons-original-1997-letter-to-shareholders ### Jeff Bezos · 1998 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 1998 reporting period; published 1999 Status: Read in full Complete signed current-period letter read. It shows demand shifting the constraint toward systems, distribution, hiring and operating process. https://s2.q4cdn.com/299287126/files/doc_financials/annual/Shareholderletter98.pdf ### Jeff Bezos · 1999 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 1999 reporting period; published 2000 Status: Read in full Complete signed current-period letter read. It records hypergrowth, category proliferation and an increase from roughly 300,000 to more than 5 million square feet of distribution capacity in less than 12 months. https://s2.q4cdn.com/299287126/files/doc_financials/annual/Shareholderletter99.pdf ### Jeff Bezos · 2000 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 2000 reporting period; published 2001 Status: Read in full Complete signed current-period letter read. It admits losses and revises the land-rush assumption after the share-price collapse. https://s2.q4cdn.com/299287126/files/doc_financials/annual/00ar_letter.pdf ### Jeff Bezos · 2001 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 2001 reporting period; published 2002 Status: Read in full Complete signed current-period letter read. It distinguishes the growth years from almost two years of cost work and reports the first profitable quarter on Amazon's pro forma measures. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2001_shareholderLetter.pdf ### Jeff Bezos · 2002 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 2002 reporting period; published 2003 Status: Read in full Complete signed current-period letter read. It reports bounded price, service and free-cash-flow evidence for the revised operating loop. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2002_shareholderLetter.pdf ### Jeff Bezos · 2003 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 2003 reporting period; published 2004 Status: Read in full Complete signed current-period letter read from the official archive because the local 2003 original was absent. It reframes selected short-term sales sacrifices as owner behaviour. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2003_-Shareholder_-Letter041304.pdf ### Jeff Bezos · 2004 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 2004 reporting period; published 2005 Status: Read in full Complete signed current-period letter read. It makes free cash flow per share the financial control and shows why earnings growth can consume value. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2004_shareholderLetter.pdf ### Jeff Bezos · 2005 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 2005 reporting period; published 2006 Status: Read in full Complete signed current-period letter read. It distinguishes model-heavy operating choices from long-horizon judgment and supplies later Marketplace evidence. https://s2.q4cdn.com/299287126/files/doc_financials/annual/shareholderletter2005.pdf ### Jeff Bezos · 2006 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 2006 reporting period; published 2007 Status: Read in full Complete signed current-period letter read. It defines a new-business test and records a physical-store conclusion later changed under new facts. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2006.PDF ### Jeff Bezos · 2007 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 2007 reporting period; published 2008 Status: Read in full Complete signed current-period letter read. Kindle's long development was followed by a 5.5-hour launch sellout and a new supply constraint. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2007letter.pdf ### Jeff Bezos · 2008 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 2008 reporting period; published 2009 Status: Read in full Complete signed current-period letter read. Recession pressure sharpens the working-backward test and durable customer needs. https://s2.q4cdn.com/299287126/files/doc_financials/annual/Amazon_SH_Letter_2008.pdf ### Jeff Bezos · 2009 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 2009 reporting period; published 2010 Status: Read in full Complete signed current-period letter read. A 452-goal planning set shows input ownership and revision without proving that outputs can be ignored. https://s2.q4cdn.com/299287126/files/doc_financials/annual/AMZN_Shareholder-Letter-2009-(final).pdf ### Jeff Bezos · 2010 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 2010 reporting period; published 2011 Status: Read in full Complete signed current-period letter read. It explains how operating strain produced service architecture and reusable technical capability. https://s2.q4cdn.com/299287126/files/doc_financials/annual/117006_ltr_ltr2.pdf ### Jeff Bezos · 2011 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 2011 reporting period; published 2012 Status: Read in full Complete signed current-period letter read. It presents Amazon platforms as a way for sellers, authors and developers to bypass older gatekeepers. https://s2.q4cdn.com/299287126/files/doc_financials/annual/letter.PDF ### Jeff Bezos · 2012 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 2012 reporting period; published 2013 Status: Read in full Complete signed current-period letter read. It acknowledges blind alleys and self-inflicted mistakes while documenting continuing device and AWS iteration. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2012-Shareholder-Letter.pdf ### Jeff Bezos · 2013 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 2013 reporting period; published 2014 Status: Read in full Complete signed current-period letter read. It records five-year incubation, rising experiment volume, fulfilment improvements and employee programs. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2013-Letter-to-Shareholders.pdf ### Jeff Bezos · 2014 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 2014 reporting period; published 2015 Status: Read in full Complete signed current-period letter read. It retells Marketplace, Prime and AWS after success while acknowledging that Auctions and zShops attracted almost no customers. https://s2.q4cdn.com/299287126/files/doc_financials/annual/AMAZON-2014-Shareholder-Letter.pdf ### Jeff Bezos · 2015 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 2015 reporting period; published 2016 Status: Read in full Complete signed current-period letter read. It responds to organizational scale through failure portfolios, reversible choices and faster execution. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2015-Letter-to-Shareholders.PDF ### Jeff Bezos · 2016 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 2016 reporting period; published 2017 Status: Read in full Complete signed current-period letter read. Day 1, proxy resistance and decision velocity are mature-company responses rather than proof of timeless practice. https://www.aboutamazon.com/news/company-news/2016-letter-to-shareholders ### Jeff Bezos · 2017 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 2017 reporting period; published 2018 Status: Read in full Complete signed current-period letter read. Bezos admits colleagues taught him operating standards and records Amazon's differentiated move into physical retail. https://s2.q4cdn.com/299287126/files/doc_financials/annual/Amazon_Shareholder_Letter.pdf ### Jeff Bezos · 2018 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 2018 reporting period; published 2019 Status: Read in full Complete signed current-period letter read. It preserves transferable capability after Fire Phone, compares third-party and first-party growth, and reports the $15 wage decision. https://s2.q4cdn.com/299287126/files/doc_financials/annual/2018-Letter-to-Shareholders.pdf ### Jeff Bezos · 2019 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 2019 reporting period; published 2020 Status: Read in full Complete signed current-period letter read. Published in April 2020, it records Amazon's early pandemic response from management's perspective. https://s2.q4cdn.com/299287126/files/doc_financials/2020/ar/2019-Shareholder-Letter.pdf ### Jeff Bezos · 2020 Letter to Shareholders Author or speaker: Jeffrey P. Bezos Date or period: 2020 reporting period; published 2021 Status: Read in full Complete signed current-period letter read. Bezos widens the stakeholder frame and promises an employer and safety revision in his final CEO letter. https://s2.q4cdn.com/299287126/files/doc_financials/2021/ar/Amazon-2020-Shareholder-Letter-and-1997-Shareholder-Letter.pdf ### Jeff Bezos · Jeff Bezos interview Author or speaker: Jeffrey P. Bezos Date or period: May 4, 2001 Status: Read in full Complete transcript read. It supplies an early retrospective with partner credit, funding risk, operational unpreparedness and a packing-process correction from a coworker. https://achievement.org/achiever/jeffrey-p-bezos/ ### Jeff Bezos · 2010 Baccalaureate Remarks Author or speaker: Jeffrey P. Bezos Date or period: May 30, 2010 Status: Read in full Complete transcript read. It shows how Bezos later organized childhood and Amazon's founding around choices, kindness and regret. https://www.princeton.edu/news/2010/05/30/2010-baccalaureate-remarks ### Jeff Bezos · Jeff Bezos live from Code 2016 Author or speaker: Jeffrey P. Bezos and Walter Mossberg Date or period: May 31, 2016 Status: Read in part The Recode liveblog was reviewed and the complete 81-minute audio was located. The audio was not listened to or transcribed, so the record is partial and no missing answer supports a chapter claim. https://live.recode.net/jeff-bezos-2016-code/ ### Jeff Bezos · Interview at the Economic Club of Washington Author or speaker: Jeffrey P. Bezos and David M. Rubenstein Date or period: September 13, 2018 Status: Read in full Complete 30-page edited transcript read. Bezos credits mentors, family and employees, describes Prime's frightening early model and identifies luck in AWS's competitive lead. https://www.economicclub.org/sites/default/files/transcripts/Jeff_Bezos_Edited_Transcript.pdf ### Jeff Bezos · Written Testimony before the House Judiciary Subcommittee Author or speaker: Jeffrey P. Bezos Date or period: July 29, 2020 Status: Read in full Complete written testimony read. It is a specific founder account delivered under antitrust scrutiny and is treated as advocacy rather than independent proof. https://docs.house.gov/meetings/JU/JU05/20200729/110883/HHRG-116-JU05-Wstate-BezosJ-20200729.pdf ### Jeff Bezos · Invent and Wander: The Collected Writings of Jeff Bezos Author or speaker: Jeff Bezos; introduction by Walter Isaacson Date or period: November 17, 2020 Status: Located, not read Publisher record located. No authorized full copy was available in the supplied library. Individually accessible component letters and speeches were reviewed separately. https://store.hbr.org/product/invent-and-wander-the-collected-writings-of-jeff-bezos-with-an-introduction-by-walter-isaacson/10466 ### Jeff Bezos · The Everything Store: Jeff Bezos and the Age of Amazon Author or speaker: Brad Stone Date or period: October 15, 2013 Status: Located, not read Official publisher page located. The full book was not present in the supplied library, and its description was not used as evidence. https://www.hachettebookgroup.com/titles/brad-stone/the-everything-store/9780316219259/ ### Jeff Bezos · Amazon Unbound: Jeff Bezos and the Invention of a Global Empire Author or speaker: Brad Stone Date or period: May 11, 2021 Status: Located, not read Official publisher page located. The full book was not present in the supplied library, and its description was not used as evidence. https://www.simonandschuster.com/books/Amazon-Unbound/Brad-Stone/9781982132620 ### Jeff Bezos · Inside Amazon: Wrestling Big Ideas in a Bruising Workplace Author or speaker: Jodi Kantor and David Streitfeld Date or period: August 15, 2015 Status: Read in full Complete preserved article read, including Amazon's replies. It reports varied employee experiences and documents workload, health, family and internal-competition costs that complicate the high-standards narrative. https://www.nytimes.com/2015/08/16/technology/inside-amazon-wrestling-big-ideas-in-a-bruising-workplace.html ### Jeff Bezos · Investigation of Competition in Digital Markets: Majority Staff Report and Recommendations Author or speaker: U.S. House Judiciary Subcommittee on Antitrust, Commercial and Administrative Law Date or period: October 6, 2020 Status: Read in part Amazon overview and selected marketplace, data, fee, private-label, acquisition, pricing and logistics passages reviewed. The report is a majority staff assessment, not a judicial finding, and the complete 451-page report was not read. https://www.govinfo.gov/content/pkg/GOVPUB-Y4_J89_1-PURL-gpo145949/pdf/GOVPUB-Y4_J89_1-PURL-gpo145949.pdf ### Jeff Bezos · Amazon.com Services LLC, Case 10-RC-269250 Author or speaker: National Labor Relations Board Date or period: 2021 Status: Read in part Official case, initial tally and election-result records reviewed. The complete docket and later regional decision were not read, so the record is used only as a bounded institutional challenge. https://www.nlrb.gov/case/10-RC-269250 ### Jeff Bezos · 2021 Letter to Shareholders · successor authorship check Author or speaker: Andy Jassy Date or period: 2021 reporting period; published 2022 Status: Read in part Only the byline, closing signature and appended 1997-letter boundary were checked. This is succession context, not a Bezos-authored letter or a full read of Jassy’s work. https://www.aboutamazon.com/news/company-news/2021-letter-to-shareholders ### Jeff Bezos · 2022 Letter to Shareholders · successor authorship check Author or speaker: Andy Jassy Date or period: 2022 reporting period; published 2023 Status: Read in part Only the byline, closing signature and appended 1997-letter boundary were checked. This is succession context, not a Bezos-authored letter or a full read of Jassy’s work. https://www.aboutamazon.com/news/company-news/amazon-ceo-andy-jassy-2022-letter-to-shareholders ### Jeff Bezos · 2023 Letter to Shareholders · successor authorship check Author or speaker: Andy Jassy Date or period: 2023 reporting period; published 2024 Status: Read in part Only the byline, closing signature and appended 1997-letter boundary were checked. This is succession context, not a Bezos-authored letter or a full read of Jassy’s work. https://www.aboutamazon.com/news/company-news/amazon-ceo-andy-jassy-2023-letter-to-shareholders ### Jeff Bezos · 2024 Letter to Shareholders · successor authorship check Author or speaker: Andy Jassy Date or period: 2024 reporting period; published 2025 Status: Read in part Only the byline, closing signature and appended 1997-letter boundary were checked. This is succession context, not a Bezos-authored letter or a full read of Jassy’s work. https://www.aboutamazon.com/news/company-news/amazon-ceo-andy-jassy-2024-letter-to-shareholders ### Howard Marks · The Route to Performance Author or speaker: Howard Marks Date or period: 1990-10-12 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/1990-10-12-the-route-to-performance.pdf?sfvrsn=33bc0f65_6 ### Howard Marks · First Quarter Performance Author or speaker: Howard Marks Date or period: 1991-04-11 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/1991-04-11-first-quarter-client-performance.pdf?sfvrsn=d7bd0f65_6 ### Howard Marks · Microeconomics 101: Supply, Demand and Convertibles Author or speaker: Howard Marks Date or period: 1992-10-08 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/1992-10-08-microeconomics.pdf?sfvrsn=43bc0f65_6 ### Howard Marks · The Value of Predictions, or Where'd All This Rain Come From? Author or speaker: Howard Marks Date or period: 1993-02-15 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/1993-02-15-the-value-of-predictions-or-where-39-d-all-this-rain-come-from.pdf?sfvrsn=6fbc0f65_6 ### Howard Marks · Random Thoughts on the Identification of Investment Opportunities Author or speaker: Howard Marks Date or period: 1994-01-24 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/1994-01-24-random-thoughts-on-the-identification-of-inv-opportunities.pdf?sfvrsn=17bc0f65_6 ### Howard Marks · Risk in Today's Markets Author or speaker: Howard Marks Date or period: 1994-02-17 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/1994-02-17-risk-in-todays-markets.pdf?sfvrsn=dfbd0f65_6 ### Howard Marks · "Risk in Today's Markets" Revisited Author or speaker: Howard Marks Date or period: 1994-04-11 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/1994-04-11-risk-in-todays-markets-revisted.pdf?sfvrsn=2bbc0f65_6 ### Howard Marks · "How Does an Inefficient Market Get That Way?" Author or speaker: Howard Marks Date or period: 1994-07-15 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/1994-07-15-how-does-an-inefficient-market-get-that-way.pdf?sfvrsn=dbbd0f65_6 ### Howard Marks · How the Game Should Be Played Author or speaker: Howard Marks Date or period: 1995-05-26 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/1995-05-26-how-the-game-should-be-played.pdf?sfvrsn=d3bd0f65_6 ### Howard Marks · The Value of Predictions II (or "Give That Man a Cigar") Author or speaker: Howard Marks Date or period: 1996-07-22 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/1996-07-22-the-value-of-predictions-ii-or-give-that-man-a-cigar.pdf?sfvrsn=27bc0f65_8 ### Howard Marks · Will It Be Different This Time? Author or speaker: Howard Marks Date or period: 1996-11-25 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/1996-11-25-will-it-be-different-this-time.pdf?sfvrsn=23bc0f65_6 ### Howard Marks · Are You An Investor or a Speculator? Author or speaker: Howard Marks Date or period: 1997-09-03 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/are-you-an-investor-or-a-speculator.pdf?sfvrsn=1e37cf65_2 ### Howard Marks · Who Knew? Author or speaker: Howard Marks Date or period: 1998-01-08 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/1998-01-08-who-knew.pdf?sfvrsn=2fbc0f65_6 ### Howard Marks · Genius Isn't Enough (and Other Lessons from Long-Term Capital Management) Author or speaker: Howard Marks Date or period: 1998-10-09 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/1998-10-09-genius-isn-39-t-enough.pdf?sfvrsn=eac20f65_6 ### Howard Marks · How's the Market? Author or speaker: Howard Marks Date or period: 1999-04-15 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/1999-04-15-hows-the-market.pdf?sfvrsn=3fbc0f65_6 ### Howard Marks · bubble.com Author or speaker: Howard Marks Date or period: 2000-01-02 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2000-01-02-bubble.pdf?sfvrsn=37bc0f65_5 ### Howard Marks · Irrational Exuberance Author or speaker: Howard Marks Date or period: 2000-05-01 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2000-05-01-irrational-exuberance.pdf?sfvrsn=cfbc0f65_6 ### Howard Marks · Investment Miscellany Author or speaker: Howard Marks Date or period: 2000-11-16 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2000-11-16-investment-miscellany.pdf?sfvrsn=1fbc0f65_6 ### Howard Marks · We're Not In 1999 Anymore, Toto Author or speaker: Howard Marks Date or period: 2000-12-31 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2000-12-31-were-not-in-1999-anymore.pdf?sfvrsn=3bbc0f65_6 ### Howard Marks · Safety First . . . But Where? Author or speaker: Howard Marks Date or period: 2001-04-10 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2001-04-10-safety-first-but-where.pdf?sfvrsn=49c10f65_8 ### Howard Marks · What's It All About, Alpha? Author or speaker: Howard Marks Date or period: 2001-07-11 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2001-07-11-whats-it-all-about-alpha.pdf?sfvrsn=13bc0f65_6 ### Howard Marks · Notes from New York Author or speaker: Howard Marks Date or period: 2001-09-16 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2001-09-16-notes-from-ny.pdf?sfvrsn=fbc0f65_6 ### Howard Marks · What Lies Ahead? Author or speaker: Howard Marks Date or period: 2001-10-04 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2001-10-04-what-lies-ahead.pdf?sfvrsn=bbc0f65_6 ### Howard Marks · You Can't Predict. You Can Prepare. Author or speaker: Howard Marks Date or period: 2001-11-20 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2001-11-20-you-cant-predict-you-can-prepare.pdf?sfvrsn=bc00f65_6 ### Howard Marks · Learning From Enron Author or speaker: Howard Marks Date or period: 2002-03-14 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2002-03-14-learning-from-enron.pdf?sfvrsn=77bc0f65_6 ### Howard Marks · The Realist's Creed Author or speaker: Howard Marks Date or period: 2002-05-31 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2002-05-31-the-realists-creed.pdf?sfvrsn=1bbc0f65_5 ### Howard Marks · Quo Vadis? Author or speaker: Howard Marks Date or period: 2002-07-26 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2002-07-26-quo-vadis.pdf?sfvrsn=67bc0f65_6 ### Howard Marks · Etorre's Wisdom Author or speaker: Howard Marks Date or period: 2002-09-04 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2002-09-04-etorres-wisdom.pdf?sfvrsn=63bc0f65_6 ### Howard Marks · Returns and How They Get That Way Author or speaker: Howard Marks Date or period: 2002-11-11 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2002-11-11-returns-and-how-they-get-that-way.pdf?sfvrsn=4fbc0f65_8 ### Howard Marks · Whad’Ya Know? Author or speaker: Howard Marks Date or period: 2003-03-11 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2003-03-11-whad-ya-know.pdf?sfvrsn=abbc0f65_6 ### Howard Marks · What's Going On? Author or speaker: Howard Marks Date or period: 2003-05-06 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2003-05-16-whats-going-on.pdf?sfvrsn=73bc0f65_6 ### Howard Marks · The Most Important Thing Author or speaker: Howard Marks Date or period: 2003-07-01 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2003-07-01-the-most-important-thing.pdf?sfvrsn=91c00f65_6 ### Howard Marks · What’s Your Game Plan? Author or speaker: Howard Marks Date or period: 2003-09-05 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2003-09-05-whats-your-game-plan.pdf?sfvrsn=7bbc0f65_6 ### Howard Marks · The Feeling's Mutual Author or speaker: Howard Marks Date or period: 2003-10-02 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2003-10-02-the-feelings-mutual.pdf?sfvrsn=47bc0f65_6 ### Howard Marks · Hey, Steward!! Author or speaker: Howard Marks Date or period: 2004-03-16 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2004-03-16-hey-steward!!.pdf?sfvrsn=a7bc0f65_6 ### Howard Marks · Us and Them Author or speaker: Howard Marks Date or period: 2004-05-07 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2004-05-07-us-and-them.pdf?sfvrsn=bfbc0f65_6 ### Howard Marks · The Happy Medium Author or speaker: Howard Marks Date or period: 2004-07-20 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2004-07-21-the-happy-medium.pdf?sfvrsn=4bbc0f65_6 ### Howard Marks · Hedge Funds: A Case for Caution Author or speaker: Howard Marks Date or period: 2004-10-06 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2004-10-06-hedge-funds-a-case-for-caution.pdf?sfvrsn=57bc0f65_6 ### Howard Marks · Risk and Return Today Author or speaker: Howard Marks Date or period: 2004-10-27 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2004-10-27-risk-and-return-today.pdf?sfvrsn=53bc0f65_6 ### Howard Marks · Oaktree at Ten Author or speaker: Howard Marks Date or period: 2005-04-11 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2005-04-11-oaktree-at-ten.pdf?sfvrsn=5fbc0f65_6 ### Howard Marks · There They Go Again Author or speaker: Howard Marks Date or period: 2005-05-06 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2005-05-06-there-they-go-again.pdf?sfvrsn=5bbc0f65_6 ### Howard Marks · A Case in Point Author or speaker: Howard Marks Date or period: 2005-06-06 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2005-06-06-a-case-in-point.pdf?sfvrsn=a3bc0f65_6 ### Howard Marks · Hindsight First, Please (or, What Were They Thinking?) Author or speaker: Howard Marks Date or period: 2005-10-17 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2005-10-17-hindsight-first-please-or-what-where-they-thinking.pdf?sfvrsn=93bc0f65_6 ### Howard Marks · Risk Author or speaker: Howard Marks Date or period: 2006-01-19 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2006-01-19-risk.pdf?sfvrsn=afbc0f65_6 ### Howard Marks · It Is What It Is Author or speaker: Howard Marks Date or period: 2006-03-27 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2006-03-27-it-is-what-it-is.pdf?sfvrsn=87bc0f65_6 ### Howard Marks · Returns, Absolute Returns and Risk Author or speaker: Howard Marks Date or period: 2006-06-13 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2006-06-13-returns-absolute-returns-and-risk.pdf?sfvrsn=b7bc0f65_6 ### Howard Marks · You Can’t Eat IRR Author or speaker: Howard Marks Date or period: 2006-07-12 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2006-07-12-you-cant-eat-irr.pdf?sfvrsn=83bc0f65_6 ### Howard Marks · Dare to Be Great Author or speaker: Howard Marks Date or period: 2006-09-07 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2006-09-07-dare-to-be-great.pdf?sfvrsn=b3bc0f65_6 ### Howard Marks · The New Paradigm Author or speaker: Howard Marks Date or period: 2006-10-19 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2006-10-19-the-new-paradigm.pdf?sfvrsn=8bbc0f65_6 ### Howard Marks · Pigweed Author or speaker: Howard Marks Date or period: 2006-12-07 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2006-12-07-pigweed.pdf?sfvrsn=bbbc0f65_6 ### Howard Marks · The Race to the Bottom Author or speaker: Howard Marks Date or period: 2007-02-14 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2007-02-14-the-race-to-the-bottom.pdf?sfvrsn=9bbc0f65_6 ### Howard Marks · Everyone Knows Author or speaker: Howard Marks Date or period: 2007-04-26 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2007-04-26-everyone-knows.pdf?sfvrsn=d3bc0f65_6 ### Howard Marks · It’s All Good Author or speaker: Howard Marks Date or period: 2007-07-16 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2007-07-16-its-all-good.pdf?sfvrsn=8fbc0f65_6 ### Howard Marks · It’s All Good . . . Really? Author or speaker: Howard Marks Date or period: 2007-07-30 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2007-07-30-its-all-good-really.pdf?sfvrsn=97bc0f65_6 ### Howard Marks · Now It’s All Bad? Author or speaker: Howard Marks Date or period: 2007-09-10 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2007-09-10-now-its-all-bad.pdf?sfvrsn=9fbc0f65_6 ### Howard Marks · No Different This Time – The Lessons of ‘07 Author or speaker: Howard Marks Date or period: 2007-12-17 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2007-12-17-no-different-this-time-the-lessons-of-07.pdf?sfvrsn=f3bc0f65_6 ### Howard Marks · Now What? Author or speaker: Howard Marks Date or period: 2008-01-10 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2008-01-10-now-what.pdf?sfvrsn=d7bc0f65_6 ### Howard Marks · Whodunit Author or speaker: Howard Marks Date or period: 2008-02-20 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2008-02-20-whodunit.pdf?sfvrsn=e7bc0f65_6 ### Howard Marks · The Tide Goes Out Author or speaker: Howard Marks Date or period: 2008-03-18 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2008-03-18-the-tide-goes-out.pdf?sfvrsn=e3bc0f65_6 ### Howard Marks · The Aviary Author or speaker: Howard Marks Date or period: 2008-05-16 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2008-05-16-the-aviary.pdf?sfvrsn=efbc0f65_6 ### Howard Marks · Doesn’t Make Sense Author or speaker: Howard Marks Date or period: 2008-07-31 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2008-07-31-doesnt-make-sense.pdf?sfvrsn=ebbc0f65_6 ### Howard Marks · What Worries Me Author or speaker: Howard Marks Date or period: 2008-08-28 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2008-08-28-what-worries-me.pdf?sfvrsn=f7bc0f65_6 ### Howard Marks · Nobody Knows Author or speaker: Howard Marks Date or period: 2008-09-19 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2008-09-19-nobody-knows.pdf?sfvrsn=cbbc0f65_6 ### Howard Marks · Plan B Author or speaker: Howard Marks Date or period: 2008-09-24 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2008-09-24-plan-b.pdf?sfvrsn=ffbc0f65_6 ### Howard Marks · The Limits to Negativism Author or speaker: Howard Marks Date or period: 2008-10-15 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2008-10-15-the-limits-to-negativism.pdf?sfvrsn=fbbc0f65_6 ### Howard Marks · Volatility + Leverage = Dynamite Author or speaker: Howard Marks Date or period: 2008-12-17 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2008-12-17-volatility-leverage-dynamite.pdf?sfvrsn=c7bc0f65_6 ### Howard Marks · The Long View Author or speaker: Howard Marks Date or period: 2009-01-09 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2009-01-09-the-long-view.pdf?sfvrsn=c3bc0f65_6 ### Howard Marks · Will It Work? Author or speaker: Howard Marks Date or period: 2009-03-05 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2009-03-05-will-it-work.pdf?sfvrsn=dbbc0f65_8 ### Howard Marks · So Much That’s False and Nutty Author or speaker: Howard Marks Date or period: 2009-07-08 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2009-07-08-so-much-thats-false-nutty.pdf?sfvrsn=23bb0f65_6 ### Howard Marks · Touchstones Author or speaker: Howard Marks Date or period: 2009-11-10 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2009-11-10-touchstones.pdf?sfvrsn=2fbb0f65_6 ### Howard Marks · Tell Me I’m Wrong Author or speaker: Howard Marks Date or period: 2010-01-22 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2010-01-22-tell-me-im-wrong.pdf?sfvrsn=2bbb0f65_6 ### Howard Marks · I’d Rather Be Wrong Author or speaker: Howard Marks Date or period: 2010-03-17 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2010-03-17-id-rather-be-wrong.pdf?sfvrsn=fbb0f65_6 ### Howard Marks · Warning Flags Author or speaker: Howard Marks Date or period: 2010-05-12 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2010-05-12-warning-flags.pdf?sfvrsn=dfbc0f65_6 ### Howard Marks · It’s Greek to Me Author or speaker: Howard Marks Date or period: 2010-07-19 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2010-07-19-its-greek-to-me.pdf?sfvrsn=27bb0f65_6 ### Howard Marks · Hemlines Author or speaker: Howard Marks Date or period: 2010-09-10 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2010-09-10-hemlines.pdf?sfvrsn=7bb0f65_6 ### Howard Marks · Open and Shut Author or speaker: Howard Marks Date or period: 2010-12-01 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2010-12-01-open-and-shut.pdf?sfvrsn=67bb0f65_6 ### Howard Marks · All That Glitters Author or speaker: Howard Marks Date or period: 2010-12-17 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2010-12-17-all-that-glitters.pdf?sfvrsn=3bb0f65_6 ### Howard Marks · On Regulation Author or speaker: Howard Marks Date or period: 2011-03-02 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2011-03-02-on-regulation.pdf?sfvrsn=37bb0f65_6 ### Howard Marks · How Quickly They Forget Author or speaker: Howard Marks Date or period: 2011-05-25 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2011-05-25-how-quickly-they-forget.pdf?sfvrsn=33bb0f65_6 ### Howard Marks · Down to the Wire Author or speaker: Howard Marks Date or period: 2011-07-21 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2011-07-21-down-to-the-wire.pdf?sfvrsn=3fbb0f65_6 ### Howard Marks · What's Behind the Downturn? Author or speaker: Howard Marks Date or period: 2011-09-07 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2011-09-07-whats-behind-the-downturn.pdf?sfvrsn=13bb0f65_6 ### Howard Marks · It's All Very Taxing Author or speaker: Howard Marks Date or period: 2011-11-16 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2011-11-16-its-all-very-taxing.pdf?sfvrsn=3bbb0f65_6 ### Howard Marks · What Can We Do For You? Author or speaker: Howard Marks Date or period: 2012-01-10 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2012-01-10-what-can-we-do-for-you.pdf?sfvrsn=7fbb0f65_8 ### Howard Marks · Assessing Performance Records A Case Study Author or speaker: Howard Marks Date or period: 2012-02-15 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2012-02-15-assessing-performance-records-a-case-study.pdf?sfvrsn=6bbb0f65_6 ### Howard Marks · Déjà Vu All Over Again Author or speaker: Howard Marks Date or period: 2012-03-19 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2012-03-19-déjà-vu-all-over-again.pdf?sfvrsn=bbb0f65_6 ### Howard Marks · It's All A Big Mistake Author or speaker: Howard Marks Date or period: 2012-06-20 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2012-06-20-its-all-a-big-mistake.pdf?sfvrsn=17bb0f65_6 ### Howard Marks · On Uncertain Ground Author or speaker: Howard Marks Date or period: 2012-09-11 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2012-09-11-on-uncertain-ground.pdf?sfvrsn=1fbb0f65_8 ### Howard Marks · A Fresh Start (Hopefully) Author or speaker: Howard Marks Date or period: 2012-11-19 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2012-11-09-a-fresh-start-hopefully.pdf?sfvrsn=1bbb0f65_5 ### Howard Marks · Ditto Author or speaker: Howard Marks Date or period: 2013-01-07 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2013-01-07-ditto.pdf?sfvrsn=73bb0f65_6 ### Howard Marks · High Yield Bonds Today Author or speaker: Howard Marks and Sheldon Stone Date or period: 2013-02-21 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2013-02-21-high-yield-bonds-today.pdf?sfvrsn=63bb0f65_6 ### Howard Marks · The Outlook For Equities Author or speaker: Howard Marks Date or period: 2013-03-13 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2013-03-13-the-outlook-for-equities.pdf?sfvrsn=6fbb0f65_6 ### Howard Marks · The Role of Confidence Author or speaker: Howard Marks Date or period: 2013-08-05 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2013-08-05-the-role-of-confidence.pdf?sfvrsn=7bbb0f65_6 ### Howard Marks · The Race Is On Author or speaker: Howard Marks Date or period: 2013-11-26 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2013-11-26-the-race-is-on.pdf?sfvrsn=77bb0f65_6 ### Howard Marks · Getting Lucky Author or speaker: Howard Marks Date or period: 2014-01-16 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2014-01-16-getting-lucky.pdf?sfvrsn=c4b70f65_6 ### Howard Marks · Dare to Be Great II Author or speaker: Howard Marks Date or period: 2014-04-08 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2014-04-08-dare-to-be-great-ii.pdf?sfvrsn=44b60f65_6 ### Howard Marks · Risk Revisited Author or speaker: Howard Marks Date or period: 2014-09-03 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2014-09-03-risk-revisited.pdf?sfvrsn=78b60f65_6 ### Howard Marks · The Lessons of Oil Author or speaker: Howard Marks Date or period: 2014-12-18 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2014-12-18-the-lessons-of-oil.pdf?sfvrsn=40b60f65_6 ### Howard Marks · Liquidity Author or speaker: Howard Marks Date or period: 2015-03-25 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2015-03-25-liquidity.pdf?sfvrsn=2dc70f65_6 ### Howard Marks · Risk Revisited Again Author or speaker: Howard Marks Date or period: 2015-06-08 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2015-06-08-risk-revisited-again.pdf?sfvrsn=7bb70f65_8 ### Howard Marks · It’s not Easy Author or speaker: Howard Marks Date or period: 2015-09-09 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2015-09-09-its-not-easy.pdf?sfvrsn=47bb0f65_7 ### Howard Marks · Inspiration from the World of Sports Author or speaker: Howard Marks Date or period: 2015-10-22 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/2015-10-22-inspiration-from-the-world-of-sports.pdf?sfvrsn=aa1f0e65_7 ### Howard Marks · On the Couch Author or speaker: Howard Marks Date or period: 2016-01-14 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/on-the-couch.pdf?sfvrsn=8a1e0165_7 ### Howard Marks · What Does the Market Know? Author or speaker: Howard Marks Date or period: 2016-01-19 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/what-does-the-market-know.pdf?sfvrsn=cb7a0165_10 ### Howard Marks · Economic Reality Author or speaker: Howard Marks Date or period: 2016-05-26 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/economic-reality.pdf?sfvrsn=49bc1865_6 ### Howard Marks · Political Reality Author or speaker: Howard Marks Date or period: 2016-08-17 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/political-reality.pdf?sfvrsn=39391265_2 ### Howard Marks · Implications of the Election Author or speaker: Howard Marks Date or period: 2016-11-07 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/implications-of-the-election_11716.pdf?sfvrsn=1389e965_4 ### Howard Marks · Go Figure! Author or speaker: Howard Marks Date or period: 2016-11-14 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/go-figure.pdf?sfvrsn=ef38e965_7 ### Howard Marks · Expert Opinion Author or speaker: Howard Marks Date or period: 2017-01-10 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/expert-opinion.pdf?sfvrsn=ccffe365_4 ### Howard Marks · Lines in the Sand Author or speaker: Howard Marks Date or period: 2017-04-18 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/lines-in-the-sand.pdf?sfvrsn=bf5dfa65_2 ### Howard Marks · There They Go Again... Again Author or speaker: Howard Marks Date or period: 2017-07-26 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/there-they-go-again-again.pdf?sfvrsn=56d4f265_6 ### Howard Marks · Yet Again? Author or speaker: Howard Marks Date or period: 2017-09-07 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/yet-again.pdf?sfvrsn=3767f765_6 ### Howard Marks · Latest Thinking Author or speaker: Howard Marks Date or period: 2018-01-23 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/latest-thinking.pdf?sfvrsn=a77ccf65_2 ### Howard Marks · Investing Without People Author or speaker: Howard Marks Date or period: 2018-06-18 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/investing-without-people.pdf?sfvrsn=7a5ec465_8 ### Howard Marks · The Seven Worst Words in the World Author or speaker: Howard Marks Date or period: 2018-09-26 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/the-seven-worst-words-in-the-world.pdf?sfvrsn=6dc9dd65_4 ### Howard Marks · Political Reality Meets Economic Reality Author or speaker: Howard Marks Date or period: 2019-01-30 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/political-reality-meets-economic-reality.pdf?sfvrsn=21c1d665_8 ### Howard Marks · Growing the Pie Author or speaker: Howard Marks Date or period: 2019-04-01 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/growing-the-pie.pdf?sfvrsn=423fab65_8 ### Howard Marks · This Time It's Different Author or speaker: Howard Marks Date or period: 2019-06-12 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/this-time-its-different.pdf?sfvrsn=fc19af65_10 ### Howard Marks · On the Other Hand Author or speaker: Howard Marks Date or period: 2019-07-26 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/on-the-other-hand.pdf?sfvrsn=9b1a365_4 ### Howard Marks · Mysterious Author or speaker: Howard Marks Date or period: 2019-10-17 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/mysterious.pdf?sfvrsn=8a30a565_12 ### Howard Marks · You Bet! Author or speaker: Howard Marks Date or period: 2020-01-13 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/you-bet.pdf?sfvrsn=785dbe65_8 ### Howard Marks · Nobody Knows II Author or speaker: Howard Marks Date or period: 2020-03-03 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/nobody-knows-ii.pdf?sfvrsn=108eb165_8 ### Howard Marks · Latest Update Author or speaker: Howard Marks Date or period: 2020-03-19 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/weekly.pdf?sfvrsn=cbf3b065_8 ### Howard Marks · Which Way Now? Author or speaker: Howard Marks Date or period: 2020-03-31 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/which-way-now.pdf?sfvrsn=f5f4b765_8 ### Howard Marks · Calibrating Author or speaker: Howard Marks Date or period: 2020-04-06 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/calibrating.pdf?sfvrsn=3e98b665_6 ### Howard Marks · Knowledge of the Future Author or speaker: Howard Marks Date or period: 2020-04-14 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/knowledge-of-the-future.pdf?sfvrsn=fa39b665_6 ### Howard Marks · Uncertainty Author or speaker: Howard Marks Date or period: 2020-05-11 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/uncertainty.pdf?sfvrsn=d913b465_28 ### Howard Marks · Uncertainty II Author or speaker: Howard Marks Date or period: 2020-05-28 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/uncertainty-ii.pdf?sfvrsn=d48a65_8 ### Howard Marks · Not Enough Author or speaker: Howard Marks Date or period: 2020-06-11 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/not-enough.pdf?sfvrsn=8f828c65_4 ### Howard Marks · The Anatomy of a Rally Author or speaker: Howard Marks Date or period: 2020-06-18 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/the-anatomy-of-a-rally.pdf?sfvrsn=f5828c65_4 ### Howard Marks · Time for Thinking Author or speaker: Howard Marks Date or period: 2020-08-05 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/docs/default-source/memos/timeforthinking.pdf?sfvrsn=17818c65_8 ### Howard Marks · Coming into Focus Author or speaker: Howard Marks Date or period: 2020-10-13 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/coming-into-focus ### Howard Marks · Something of Value Author or speaker: Howard Marks Date or period: 2021-01-11 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/something-of-value ### Howard Marks · 2020 in Review Author or speaker: Howard Marks Date or period: 2021-03-04 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/2020-in-review ### Howard Marks · Thinking About Macro Author or speaker: Howard Marks Date or period: 2021-07-29 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/thinking-about-macro ### Howard Marks · The Winds of Change Author or speaker: Howard Marks Date or period: 2021-11-23 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/the-winds-of-change ### Howard Marks · Selling Out Author or speaker: Howard Marks Date or period: 2022-01-13 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/selling-out ### Howard Marks · The Pendulum in International Affairs Author or speaker: Howard Marks Date or period: 2022-03-23 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/the-pendulum-in-international-affairs ### Howard Marks · Bull Market Rhymes Author or speaker: Howard Marks Date or period: 2022-05-26 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/bull-market-rhymes ### Howard Marks · Conversation at Panmure House Author or speaker: Howard Marks Date or period: 2022-06-23 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/conversation-at-panmure-house ### Howard Marks · I Beg to Differ Author or speaker: Howard Marks Date or period: 2022-07-26 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/i-beg-to-differ ### Howard Marks · The Illusion of Knowledge Author or speaker: Howard Marks Date or period: 2022-09-08 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/the-illusion-of-knowledge ### Howard Marks · What Really Matters? Author or speaker: Howard Marks Date or period: 2022-11-22 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/what-really-matters ### Howard Marks · Sea Change Author or speaker: Howard Marks Date or period: 2022-12-13 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/sea-change ### Howard Marks · Lessons from Silicon Valley Bank Author or speaker: Howard Marks Date or period: 2023-04-17 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/lessons-from-silicon-valley-bank ### Howard Marks · Taking the Temperature Author or speaker: Howard Marks Date or period: 2023-07-10 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/taking-the-temperature ### Howard Marks · Fewer Losers, or More Winners? Author or speaker: Howard Marks Date or period: 2023-09-12 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/fewer-losers-or-more-winners ### Howard Marks · Further Thoughts on Sea Change Author or speaker: Howard Marks Date or period: 2023-10-11 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/further-thoughts-on-sea-change ### Howard Marks · Easy Money Author or speaker: Howard Marks Date or period: 2024-01-09 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/easy-money ### Howard Marks · The Indispensability of Risk Author or speaker: Howard Marks Date or period: 2024-04-17 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/the-indispensability-of-risk ### Howard Marks · The Impact of Debt Author or speaker: Howard Marks Date or period: 2024-05-08 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/the-impact-of-debt ### Howard Marks · The Folly of Certainty Author or speaker: Howard Marks Date or period: 2024-07-17 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/the-folly-of-certainty ### Howard Marks · Mr. Market Miscalculates Author or speaker: Howard Marks Date or period: 2024-08-22 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/mr-market-miscalculates ### Howard Marks · Shall We Repeal the Laws of Economics? Author or speaker: Howard Marks Date or period: 2024-09-19 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/shall-we-repeal-the-laws-of-economics ### Howard Marks · Ruminating on Asset Allocation Author or speaker: Howard Marks Date or period: 2024-10-22 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/ruminating-on-asset-allocation ### Howard Marks · On Bubble Watch Author or speaker: Howard Marks Date or period: 2025-01-07 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/on-bubble-watch ### Howard Marks · Gimme Credit Author or speaker: Howard Marks Date or period: 2025-03-06 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/gimme-credit ### Howard Marks · Nobody Knows (Yet Again) Author or speaker: Howard Marks Date or period: 2025-04-09 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/nobody-knows-yet-again ### Howard Marks · More on Repealing the Laws of Economics Author or speaker: Howard Marks Date or period: 2025-06-18 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/more-on-repealing-the-laws-of-economics ### Howard Marks · The Calculus of Value Author or speaker: Howard Marks Date or period: 2025-08-14 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/the-calculus-of-value ### Howard Marks · The Best of . . . Author or speaker: Howard Marks Date or period: 2025-10-12 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/the-best-of ### Howard Marks · A Look Under the Hood Author or speaker: Howard Marks Date or period: 2025-10-28 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/a-look-under-the-hood ### Howard Marks · Cockroaches in the Coal Mine Author or speaker: Howard Marks Date or period: 2025-11-06 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/cockroaches-in-the-coal-mine ### Howard Marks · Is It a Bubble? Author or speaker: Howard Marks Date or period: 2025-12-09 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/is-it-a-bubble ### Howard Marks · AI Hurtles Ahead Author or speaker: Howard Marks Date or period: 2026-02-26 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/ai-hurtles-ahead ### Howard Marks · What’s Going on in Private Credit? Author or speaker: Howard Marks Date or period: 2026-04-09 Status: Read in full Read from the complete original memo. Development evidence, revisions and period claims were logged in the private reading record. https://www.oaktreecapital.com/insights/memo/whats-going-on-in-private-credit ### Howard Marks · The Complete Collection: Memos from Howard Marks Author or speaker: Howard Marks Date or period: 2025-12-31 Status: Read in part All dated memo texts contained in the 1,641-page official collection were read in full and tracked individually. The compilation remains Read in part because its front matter, contents pages, repeated legal disclosures and every decorative image were not reviewed page by page. Meaningful figures identified during the memo readings were inspected. https://www.oaktreecapital.com/docs/default-source/memos/the-complete-collection.pdf?sfvrsn=58102966_3 ### Howard Marks · The Most Important Thing: Uncommon Sense for the Thoughtful Investor Author or speaker: Howard Marks Date or period: 2011-05-01 Status: Located, not read The publisher description and contents were inspected. No authorized local full text was found, so the book is not claimed as read. https://cup.columbia.edu/book/the-most-important-thing/9780231153683/ ### Howard Marks · The Most Important Thing Illuminated Author or speaker: Howard Marks Date or period: 2013-01-01 Status: Located, not read The publisher description and contents were inspected. No authorized local full text was found, so the annotations and counterpoints remain an explicit gap. https://cup.columbia.edu/book/the-most-important-thing-illuminated/9780231162845/ ### Howard Marks · Mastering the Market Cycle: Getting the Odds on Your Side Author or speaker: Howard Marks Date or period: 2018-10-02 Status: Located, not read The publisher page was inspected. No authorized local full text was found. Original later interviews and memos were used only for the book’s formation and revisions. https://www.hachette.co.uk/titles/howard-marks/mastering-the-market-cycle/9781473695689/ ### Howard Marks · 30 Years of Oaktree Author or speaker: Howard Marks, Bruce Karsh, and Sheldon Stone Date or period: 2025-04-30 Status: Read in full Read the official eight-page transcript in full. It supplies formation, partnership, culture, operating pressure and succession evidence. https://www.oaktreecapital.com/insights/insight-podcast/education/30-years-of-oaktree-with-howard-marks-bruce-karsh-and-sheldon-stone ### Howard Marks · Cycles with Howard Marks Author or speaker: Howard Marks, Julio Herrera, Jordon Kruse, and Robert O’Leary Date or period: 2018-10-30 Status: Located, not read Official Oaktree program page located. The video was not reviewed in this pass. https://www.oaktreecapital.com/insights/insight-video/market-commentary/cycles-with-howard-marks ### Howard Marks · Celebrating 35 Years of Memos with Howard Marks Author or speaker: Howard Marks Date or period: 2025-10-12 Status: Located, not read Official Oaktree program page located. The video was not reviewed in this pass. https://www.oaktreecapital.com/insights/memo-video/celebrating-35-years-of-memos-with-howard-marks ### Mark Leonard · Constellation Software 2007 Q1 president’s letter Author or speaker: Mark Leonard Date or period: 2007-05-08 Status: Read in full Early initiative retrenchment; organic growth disappointment; expectation that capital will return to internal projects. Complete original letter read, including tables and glossary. https://www.csisoftware.com/wp-content/uploads/2026/04/q1_2007_shareholders_report.pdf ### Mark Leonard · Constellation Software 2007 Q2 president’s letter Author or speaker: Mark Leonard Date or period: 2007 Q2 reporting period; undated letter Status: Read in full Concern that initiative culling overshot; willingness to sacrifice margin for worthwhile growth. Complete original letter read, including tables and glossary. https://www.csisoftware.com/wp-content/uploads/2026/04/q2_2007_shareholders_report.pdf ### Mark Leonard · Constellation Software 2007 Q3 president’s letter Author or speaker: Mark Leonard Date or period: 2007 Q3 reporting period; undated letter Status: Read in full Growth target creates pressure to compromise acquisition standards; failed secondary offering. Complete original letter read, including tables and glossary. https://www.csisoftware.com/wp-content/uploads/2026/04/q3_2007_shareholders_report.pdf ### Mark Leonard · Constellation Software 2007 Q4 president’s letter Author or speaker: Mark Leonard Date or period: 2008-03-05 Status: Read in full Revised measurement definitions, excess cash incentives and limits of the intangible-asset assumption. Complete original letter read, including tables and glossary. https://www.csisoftware.com/wp-content/uploads/2026/04/q4_2007_shareholders_report.pdf ### Mark Leonard · Constellation Software 2008 Q1 president’s letter Author or speaker: Mark Leonard Date or period: 2008-05-07 Status: Read in full Patient ownership requires judging long-term trade-offs; minority investments complicate influence. Complete original letter read, including tables and glossary. https://www.csisoftware.com/wp-content/uploads/2026/04/q1_2008_shareholdersreportf49d4032f35645668472bbd0e4d94b0b.pdf ### Mark Leonard · Constellation Software 2008 Q2 president’s letter Author or speaker: Mark Leonard Date or period: 2008-08-07 Status: Read in full Reconsiders debt as acquisition opportunities improve; growth is optional and funding matters. Complete original letter read, including tables and glossary. https://www.csisoftware.com/wp-content/uploads/2026/04/q2_2008_shareholdersreport.pdf ### Mark Leonard · Constellation Software 2008 Q3 president’s letter Author or speaker: Mark Leonard Date or period: 2008-11-06 Status: Read in full MAJES uneconomic contracts and contingent exposure; strong results do not settle acquisition quality. Complete original letter read, including tables and glossary. https://www.csisoftware.com/wp-content/uploads/2026/04/presidentsletter50ed1eb201c049b287838ec5e943432b.pdf ### Mark Leonard · Constellation Software 2008 Q4 president’s letter Author or speaker: Mark Leonard Date or period: 2009-03-04 Status: Read in full Currency and bonus effects complicate reported profits; MAJES accounting and cash diverge. Complete original letter read, including tables and glossary. https://www.csisoftware.com/wp-content/uploads/2026/04/q4_2008presidentsletter.pdf ### Mark Leonard · Constellation Software 2009 Q1 president’s letter Author or speaker: Mark Leonard Date or period: 2009-05-06 Status: Read in full Worst recorded organic contraction; expected distressed sellers fail to materialize; cash-based MAJES check. Complete original letter read, including tables and glossary. https://www.csisoftware.com/wp-content/uploads/2026/04/q1_2009_presidentletter.pdf ### Mark Leonard · Constellation Software 2009 annual president’s letter Author or speaker: Mark Leonard Date or period: 2010-03-25 Status: Read in full Reduces expected organic growth after initiative evidence; retains many-vertical strategy despite board complexity concerns. Complete original letter read, including tables and glossary. https://www.csisoftware.com/wp-content/uploads/2026/04/shareholder-letter-2009.pdf ### Mark Leonard · Constellation Software 2010 annual president’s letter Author or speaker: Mark Leonard Date or period: 2011-05-02 Status: Read in full Contemporary account of strategic sale review and Leonard’s stated sadness at a possible sale. Complete original letter read, including tables and glossary. https://www.csisoftware.com/wp-content/uploads/2026/04/shareholder-letter-2010.pdf ### Mark Leonard · Constellation Software 2011 annual president’s letter Author or speaker: Mark Leonard Date or period: 2012-05-02 Status: Read in full Sale process raises short-term profits while harming long-term investment; explicitly revises view of stock-price management. Complete original letter read, including tables and glossary. https://www.csisoftware.com/wp-content/uploads/2026/04/2011_presidents_letter.pdf ### Mark Leonard · Constellation Software 2012 annual president’s letter Author or speaker: Mark Leonard Date or period: 2013-05-01 Status: Read in full Six-year initiative measurement experiment; small-unit scaling and financing constraints; dividend commitment. Complete original letter read, including tables and glossary. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-final.pdf ### Mark Leonard · Constellation Software 2013 annual president’s letter Author or speaker: Mark Leonard Date or period: 2014-04-30 Status: Read in full Analyst-model experiment reveals value attributed to future acquisitions; design of more reliable capital. Complete original letter read, including tables and glossary. https://www.csisoftware.com/wp-content/uploads/2026/04/presidentletter_2013.pdf ### Mark Leonard · Constellation Software 2014 annual president’s letter Author or speaker: Mark Leonard Date or period: 2015-04-06 Status: Read in full Personal workload and compensation change; early equity dilution reassessed; experiments, autonomy and SaaS uncertainty. Complete original letter read, including tables and glossary. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents_letter_2014.pdf ### Mark Leonard · Constellation Software 2015 annual president’s letter Author or speaker: Mark Leonard Date or period: 2016-04-26 Status: Read in full Hurdle-rate evidence, high-return bonus problems, selected conglomerate comparison, and limitations of the combined metric. Complete original letter read, including tables and glossary. https://www.csisoftware.com/wp-content/uploads/2026/04/pl_2015.pdf ### Mark Leonard · Constellation Software 2016 annual president’s letter Author or speaker: Mark Leonard Date or period: 2017-04-25 Status: Read in full Admits personal capacity limit and incomplete small-unit evidence; post-acquisition reviews, competition and management succession. Complete original letter read, including tables and glossary. https://www.csisoftware.com/wp-content/uploads/2026/04/2017-presidents-letter-1.pdf ### Mark Leonard · Constellation Software 2017 annual president’s letter Author or speaker: Mark Leonard Date or period: 2018-04-20 Status: Read in full Ends routine letters, reconsiders performance metric, defends board tenure, explores craftsmen and compounders. Complete original letter read, including tables and glossary. https://www.csisoftware.com/wp-content/uploads/2026/04/presidents-letter-april-2018-final.pdf ### Mark Leonard · Constellation Software 2021 policy president’s letter Author or speaker: Mark Leonard Date or period: 2021-02-15 Status: Read in full Explicit conversion on dividends and large-deal hurdle rates; new head-office capital stewardship mandate. Complete original letter read, including tables and glossary. https://www.csisoftware.com/wp-content/uploads/2026/04/letter-to-shareholders-february-15-2021.pdf ### Mark Leonard · May 26, 2022 · CFO Interview Author or speaker: Jamal Baksh with Tegus Date or period: 2022-04-06; posted May 26, 2022 Status: Read in full Full seven-page company-reviewed interview read. Baksh is the speaker throughout; this is not Leonard’s testimony. It explains VMS Ventures, delegated acquisitions and limits to outside-sector investment. https://www.csisoftware.com/wp-content/uploads/2026/04/April-6-2022-Tegus-interview-with-CFO.pdf ### Mark Leonard · September 17, 2021 · Members Agreement, Acquisitions, NCI and Dividends Author or speaker: Constellation Software; named respondents vary Date or period: 2021-09-17 Status: Read in full Full Q&A read. Leonard discusses Topicus option horizons and judging managers through customer and employee records. Jamal answers accounting and tax questions. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-september-2021-final.pdf ### Mark Leonard · August 17, 2020 · Margins, Bonus Plan, Contour, Organic Growth, Debentures, External Boards Author or speaker: Constellation Software; named respondents vary Date or period: 2020-08-17 Status: Read in full Full Q&A read. Jamal explains pandemic margins, bonus metrics and large-deal thresholds. Leonard responds on debt availability and outside boards. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-august-2020-final.pdf ### Mark Leonard · NOTICE TO SHAREHOLDERS REGARDING PREVIOUSLY SUBMITTED QUESTIONS Author or speaker: Constellation Software; named respondents vary Date or period: 2019-06-14 Status: Read in full Full administrative notice read. Records deletion of submitted questions; it contains no investing lesson attributed to Leonard. https://www.csisoftware.com/wp-content/uploads/2026/04/q-a-june-2019-lost-submissions.pdf ### Mark Leonard · September 19, 2018 · Additional sectors for investment, Buybacks, ROIC, Value Investing Author or speaker: Mark Leonard and named Constellation respondents Date or period: 2018-09-19 Status: Read in full Full Q&A read. Leonard qualifies his buyback position and distinguishes beliefs from certainty; Jamal supplies the invested-capital reconciliation. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-september-2018-final.pdf ### Mark Leonard · February 20, 2019 · Special Dividend Author or speaker: Mark Leonard and named Constellation respondents Date or period: 2019-02-20 Status: Read in full Full Q&A read. Leonard defends the 2019 special dividend, magnetic hurdles and employee sharing of excess returns. https://www.csisoftware.com/wp-content/uploads/2026/04/q-a-february-2019-divrelated-final.pdf ### Mark Leonard · November 5, 2018 · License revenue volatility, CSI’s biggest failure, Business systems Author or speaker: Mark Leonard and named Constellation respondents Date or period: 2018-11-05 Status: Read in full Full Q&A read. Leonard names excess early equity and a business sale as failures. Jamal explains license volatility. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-nov-5-2018-final-1.pdf ### Mark Leonard · October 9, 2018 · Moat, SaaS, TAM, corporate culture Author or speaker: Mark Leonard and named Constellation respondents Date or period: 2018-10-09 Status: Read in full Full Q&A read. Leonard discusses SaaS value capture and plural cultures. The club-software example belongs to Barry Symons. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-oct-9-2018-final-1.pdf ### Mark Leonard · October 4, 2018 · Employee attraction, retention, motivation and engagement Author or speaker: Mark Leonard and named Constellation respondents Date or period: 2018-10-04 Status: Read in full Full Q&A read. Leonard questions engagement measurement and discusses opportunity, empathy and supervisor quality. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-oct-4-2018-final.pdf ### Mark Leonard · August 3, 2018 · HPC’s, Blackboard, Customer Acquisition Economics, Margin Trajectory, Organic Growth Profile, Tax Rate, SaaS vs On-Premise Author or speaker: Mark Leonard and named Constellation respondents Date or period: 2018-08-03 Status: Read in full Full Q&A read. Leonard admits selection bias in the conglomerate study; other responses cover margins, growth, taxes and SaaS. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-august-2018-final.pdf ### Mark Leonard · July 25, 2018 · Employee Retention, Consensus, IFTODH, Bias, Incentive Compensation Author or speaker: Mark Leonard and named Constellation respondents Date or period: 2018-07-25 Status: Read in full Full Q&A read. Leonard discusses skeptics, formulaic incentives, downside borne by homebuilding managers and uncertainty in supposedly private facts. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-july-25-2018-final.pdf ### Mark Leonard · June 28, 2018 · Excess Capital, Dividends, Buybacks, Leverage Author or speaker: Mark Leonard and named Constellation respondents Date or period: Questions received through 2018-06-26; posted June 28 Status: Read in full Full Q&A read. Leonard distinguishes reliable moderate debt, dividends and moral concerns over buybacks. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-june-2018-final.pdf ### Mark Leonard · May 11, 2018 · Internal Competition, TSS Acquisitions Author or speaker: Constellation Software; named respondents vary Date or period: Questions received through 2018-05-11 Status: Read in full Full Q&A read. Corporate responses distinguish customer competition from acquisition coordination and explain TSS minority interests. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-may-14-2018-final.pdf ### Mark Leonard · April 19, 2018 · M&A, Cyber, Competition, Emulators, TSS Minority Interest, ANI Calculation, Organizational Structure Author or speaker: Constellation Software; named respondents vary Date or period: Questions received through 2018-04-19 Status: Read in full Full Q&A read. Unsigned corporate responses address acquisition capacity, minority claims, accounting assumptions and organizational flexibility. https://www.csisoftware.com/wp-content/uploads/2026/04/qa-april-20-2018-final-1.pdf ### Mark Leonard · Mark Leonard resignation and Mark Miller appointment Author or speaker: Constellation Software Date or period: 2025-09-25 Status: Read in full Company announcement establishes the leadership change and stated health reason. Expressions of confidence are expectations, not proof of succession outcomes. https://www.csisoftware.com/constellation-software-inc.-announces-the-resignation-of-mark-leonard-and-appointment-of-mark-miller-as-president-of-constellation-software/ ### Mark Leonard · Mark Leonard decision not to stand for board re-election Author or speaker: Constellation Software Date or period: 2026-03-27 Status: Read in full Company announcement gives the end-of-term plan and continuing advisory role focused on PEMS. The company’s retrospective praise is not an independent assessment. https://www.csisoftware.com/constellation-software-inc-announces-mark-leonards-decision-to-not-stand-for-re-election-to-board-of-directors/ ### Mark Leonard · Constellation Software Inc. 2026 Annual General Meeting Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2026; exact event date to verify Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.csisoftware.com/corporate-material/constellation-software-inc-2026-annual-general-meeting/ ### Mark Leonard · Constellation Software Inc. 2025 Fourth Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2026; exact event date to verify Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.csisoftware.com/corporate-material/constellation-software-inc-2025-fourth-quarter-results-webcast-2/ ### Mark Leonard · Conference Call to Discuss AI’s Impact on Software Businesses Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2025-09-22 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://edge.media-server.com/mmc/p/xea2cw62/ ### Mark Leonard · Constellation Software Inc. 2017 Fourth Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2018-02-15 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2018/02/15/fourth-quarter-2017-results-call/play ### Mark Leonard · Constellation Software Inc. 2017 Third Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2017-10-27 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2017/10/27/third-quarter-2017-results-call/play ### Mark Leonard · Constellation Software Inc. 2017 Second Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2017-07-27 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2017/07/27/second-quarter-2017-results-call/play ### Mark Leonard · Constellation Software Inc. 2017 First Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2017-04-28 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2017/04/28/first-quarter-2017-results-call/play ### Mark Leonard · Constellation Software Inc. 2016 Fourth Quarter and Year End Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2017-02-16 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2017/02/16/fourth-quarter-2016-results-call/play ### Mark Leonard · Constellation Software Inc. 2016 Third Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2016-10-27 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2016/10/27/third-quarter-2015-results-call/play ### Mark Leonard · Constellation Software Inc. 2016 Second Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2016-07-28 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2016/07/28/second-quarter-2015-results-call/play ### Mark Leonard · Constellation Software Inc. 2016 First Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2016-04-28 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2016/04/28/first-quarter-2016-results-call/play ### Mark Leonard · Constellation Software Inc. 2015 Fourth Quarter and Year End Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2016-02-19 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2016/02/19/fourth-quarter-2015-results-call/play ### Mark Leonard · Constellation Software Inc. 2015 Third Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2015-10-29 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2015/10/29/third-quarter-2015-results-call/play ### Mark Leonard · Constellation Software Inc. 2015 First Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2015-04-30 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2015/04/30/first-quarter-2015-results-call/play ### Mark Leonard · Constellation Software Inc. 2015 Second Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2015-07-30 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2015/07/30/second-quarter-2015-results-call/play ### Mark Leonard · Constellation Software Inc. 2014 Fourth Quarter and Year End Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2015-02-26 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2015/02/26/year-end-2014-results-call/play ### Mark Leonard · Constellation Software Inc. 2014 Third Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2014-11-03 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2014/11/03/q3-2014-results-call/play ### Mark Leonard · Constellation Software Inc. 2014 Second Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2014-08-01 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2014/08/01/q2-2014-results-call/play ### Mark Leonard · Constellation Software Inc. 2014 First Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2014-05-01 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2014/05/01/q1-2014-results-call/play ### Mark Leonard · Constellation Software Inc. 2013 Fourth Quarter and Year End Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2014-03-07 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2014/03/07/q4-2013-results-call/play ### Mark Leonard · Constellation Software Inc. 2013 Third Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2013-10-31 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2013/10/31/q3-2013-results-call/play ### Mark Leonard · Constellation Software Inc. 2013 Second Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2013-08-01 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2013/08/01/q2-2013-results-call/play ### Mark Leonard · Constellation Software Inc. 2013 First Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2013-05-02 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2013/05/02/q1-2013-results-call/play ### Mark Leonard · Constellation Software Inc. 2012 Fourth Quarter and Year End Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2013-03-07 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2013/03/07/q4-2012-results-call/play ### Mark Leonard · Constellation Software Inc. 2012 Third Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2012-11-02 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2012/11/02/q3-2012-results-call/play ### Mark Leonard · Constellation Software Inc. 2012 Second Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2012-08-02 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2012/08/02/q2-2012-results-call/play ### Mark Leonard · Constellation Software Inc. 2012 First Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2012-05-03 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2012/05/03/q1-2012-results-call/play ### Mark Leonard · Constellation Software Inc. 2011 Fourth Quarter and Year End Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2012-03-01 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2012/03/01/q4-2011-results-call/play ### Mark Leonard · Constellation Software Inc. 2011 Third Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2011-11-03 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2011/11/03/q3-2011-results-call/play ### Mark Leonard · Constellation Software Inc. 2011 Second Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2011-08-04 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2011/08/04/q2-2011-results-call/play ### Mark Leonard · Constellation Software Inc. 2011 First Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2011-05-05 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2011/05/05/q1-2011-results-call/play ### Mark Leonard · Constellation Software Inc. 2010 Fourth Quarter and Year End Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2011-03-03 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2011/03/03/q4-2010-results-call/play ### Mark Leonard · Constellation Software Inc. 2010 Third Quarter Results Webcast Author or speaker: Constellation Software; speaker attribution not yet verified Date or period: 2010-11-04 Status: Located, not read Listed by the official webcast archive. The recording and speaker turns have not been reviewed. No statements from it are attributed to Leonard in this course. Older players may require registration. https://www.gowebcasting.com/events/constellation-software-inc/2010/11/04/q3-2010-results-call/play