Munger checklist — v0.3.0

This profile carries the mental-models, business-quality, and psychology-of-misjudgment territory: invert the question before answering it, read incentives before reading the strategy, and judge the moat by the competitive mechanism that sustains it rather than by the statistics it produces. Cheapness is not the hurdle here — the best alternative use of the capital is. The 103 items run as one flat list; category headers are organizational only. The output of this lens is as often "too hard, next" as it is a verdict.

Items tagged [PRACTITIONER] are answered by the practitioner, never by the profile agent — the agent marks them practitioner-pending and moves on; where an Agent proxy is given, the agent evaluates the proxy instead.

Evidence grades: quote-backed = verbatim-verified primary source; teaching-derived = faithful to published teaching, paraphrased; grades signal how much weight the citation itself can bear.

Inversion & failure modes

M1. Inversion — where does this business die?

• Asks: Working from the risk-factor section (事業等のリスク), segment trends, and the reconciled figure table, name the four to six most plausible ways this company is destroyed within ten years. Check which of those attack vectors the filings (its own and competitors') show already in motion.
• Good: Each kill path is stated concretely — the usual categories are substitution, customer or supplier concentration, demographic decline in its end market, regulation, a price war by a named rival, and balance-sheet stress — and tested against numbers: the threatened segment's share of operating profit, concentration ratios trending the wrong way, unit volumes in the shrinking market. The thesis explicitly addresses the two most live vectors; "no plausible kill path found" is a red flag for shallow work, not a pass.
• Source: Talk One (Harvard School commencement, 1986, Poor Charlie's Almanack), after Jacobi — "Invert, always invert." (quote-backed)

M2. Permanent-loss exposure map

• Asks: Through which specific mechanisms could this holding produce a permanent — not quotational — loss: debt maturities, covenant breaches, dilution, customer concentration, litigation?
• Good: Builds the map from the balance sheet and notes — maturity ladder vs. free cash flow, covenant headroom, historical share-count creep, top-customer percentages, the legal-proceedings note — explicitly distinguishing price volatility from each identified path to impairment.
• Source: Poor Charlie's Almanack, Investing Principles Checklist (Risk — avoid permanent loss of capital). (teaching-derived)

M3. Inflation and rate exposure

• Asks: If inflation runs hot and rates stay elevated for five years, which lines of this company's statements get hurt, and what filed evidence shows it can pass costs through?
• Good: Identifies fixed vs. floating debt mix, lease escalators, the capex intensity of replacing assets in inflated currency, and the historical price/mix disclosure showing whether gross margin held through past cost spikes.
• Source: Investing Principles Checklist (Risk — beware inflation and interest-rate exposures). (teaching-derived)

M4. Second-order effects of the plan

• Asks: For management's stated growth plan, what are the second-order consequences — competitor response, channel cannibalization, capacity gluts, regulatory attention — and what do competitors' filings show is already in motion?
• Good: Reads rivals' filings for announced capacity and counter-moves, and names who actually captures the plan's benefit — recognizing the classic failure where every player makes the same "obviously good" investment and all the gains flow to customers.
• Source: Investing Principles Checklist (Analytic rigor — second-order and higher-level effects). (teaching-derived)

M5. Secular-change threat inventory

• Asks: Which technological or secular changes could destroy this business's economics within a decade, and does the company's own risk-factor and segment disclosure show the threat already arriving?
• Good: Names the specific wave, tracks its footprint in segment numbers over 3+ years, states whether it attacks the moat mechanism itself, and grades management's response by filed actions (capex redirection, divestitures) rather than rhetoric.
• Source: Investing Principles Checklist (Change — recognize and adapt to the true nature of the world around you). (teaching-derived)

M6. How the giants died

• Asks: How did the dominant companies of this industry's past actually die — the GM, Kodak, and department-store post-mortems for this sector — and is this company demonstrably exposed to any of the same killers?
• Good: At least three named historical failures with their specific mechanism (cost ratchet, technology substitution, channel shift), each mechanism checked against current filings with a figure-backed exposure judgment.
• Source: Daily Journal AGM Q&As 2014–2017 (GM, Kodak, retailing obituaries). (teaching-derived)

M7. Bureaucracy diseconomy

• Asks: Is there filing evidence this company has crossed into the disadvantages of scale — G&A growing faster than gross profit, thickening management layers, slowing product cycles, share losses to nimbler rivals in fast-moving niches?
• Good: Five-year trends of G&A vs. revenue, revenue per employee, and named sub-segments where smaller competitors are taking share; weighs scale benefits against bureaucratic sclerosis, since both coexist in the same firm.
• Source: Talk Two, "A Lesson on Elementary, Worldly Wisdom" (USC, 1994) — big, fat, dumb bureaucracies as the disadvantage of scale. (teaching-derived)

M8. Specialist counterattack

• Asks: Which narrowly focused specialists are attacking pieces of this large company's business, and do the specialists' filings show them winning inside their niches?
• Good: Names two or three focused competitors and compares niche-level growth and margins — the trade-journal lesson that scale in the wrong dimension loses to fanatical narrow focus — and identifies the segments where that defeat is already underway.
• Source: Talk Two (trade journals defeating the broad general-interest magazine). (teaching-derived)

M9. Technology: helps you or kills you

• Asks: When this company spends on new technology or efficiency, who has historically captured the savings — the owner, or the customer, because every competitor installs the same improvement?
• Good: Across at least one completed capex cycle, gross margin and ROIC held or improved after the spend; in commodity settings where identical equipment is available to all, the pitched "return on investment" is presumed to leak to customers, per the textile-loom lesson.
• Source: "The Art of Stock Picking" / Talk Two — "The great lesson in microeconomics is to discriminate between when technology is going to help you and when it's going to kill you." (quote-backed)

M10. Reputation heartbeat scan

• Asks: What does the filed record show about how this company treats its reputation — recalls, consent orders, customer litigation — and how much of the moat IS reputation?
• Good: Sweeps legal-proceedings and subsequent-events notes across five years, sizes the revenue that depends on trust (brands, fiduciary services, safety-critical products), and reasons through what a single scandal would do to it.
• Source: Investing Principles Checklist (Focus) — "reputation and integrity are your most valuable assets—and can be lost in a heartbeat." (quote-backed)

M11. Reliability failure screen

• Asks: Does this business reliably do what it says — guidance met, filings on time, internal controls clean, commitments delivered?
• Good: A track-record table of guidance vs. actuals over five years, late filings, control weaknesses, and auditor-opinion history; chronic unreliability disqualifies regardless of statistical cheapness, because unreliability alone can undo every virtue.
• Source: Talk One (prescriptions for guaranteed misery — unreliability foremost among them). (teaching-derived)

M12. Opaque-book instant flag

• Asks: Do the notes disclose large derivative positions, Level 3 assets, off-balance-sheet vehicles, or financing structures whose value cannot be independently reconstructed from the filings?
• Good: Derivatives notional and Level 3 assets each below a small, stated fraction of tangible equity (rule of thumb: Level 3 < 10%), no material off-balance-sheet obligations beyond disclosed leases; if the book can't be reconstructed, the item forces too-hard rather than a haircut.
• Source: Wesco/Berkshire meeting Q&As on derivatives accounting (early 2000s). (teaching-derived)

M13. Lender-earnings believability

• Asks: For a bank, insurer, or lender: are reported earnings validated by conservative reserving — or is the income statement a management estimate wearing a suit?
• Good: Multi-year reserve development neutral-to-favorable, charge-offs tracking provisioning assumptions, no reserve releases arriving exactly when needed to hit consensus, and growth not outrunning the industry in the soft part of the cycle.
• Source: Wesco AGM Q&As on financial-institution accounting (2007–2009 crisis era). (teaching-derived)

M14. Complexity-fraud proximity

• Asks: How many complexity flags does the report carry — restated segments, serial "one-time" charges, frequent auditor or CFO changes, acquisitive obfuscation of organic trends, multi-level holding structures?
• Good: At most one flag across five years of filings, organic growth reconstructable without management's help, same auditor ≥5 years with clean opinions; more than one flag shifts the burden of proof against the company.
• Source: Stanford Lawyer interview, 2009 — "Where you have complexity, by nature you can have fraud and mistakes." (quote-backed)

M15. Reaching-for-yield check

• Asks: Where the company holds an investment portfolio or float: is it invested boringly, or is management stretching for yield into credit, duration, or exotica at the wrong point in the cycle?
• Good: Investment-grade concentration, duration matched to liabilities, and no growing allocation to structured products or "alternative" yield vehicles as spreads compress; stated yield ambitions are modest.
• Source: Wesco AGM Q&As on float investing and the dangers of reaching for yield. (teaching-derived)

M16. Who-loses inversion

• Asks: Invert the revenue model: who is worse off when this company wins — customers, suppliers, employees, or regulators — and does that loser have the power to retaliate over a decade?
• Good: Revenue traceable to measurable customer benefit (win-win); where a loser exists, the filed litigation, regulatory, and political history shows the loser lacks a credible mechanism to claw the economics back.
• Source: Daily Journal AGM Q&As on win-win businesses vs. casinos and rip-offs (2017–2019). (teaching-derived)

M17. Fifty-percent-quotation survivability

• Asks: If the stock price halved with no change in the business, could the company and the position both survive and act — no covenant, refinancing, collateral, or forced-seller trigger anywhere in the chain?
• Good: No debt or convertible triggers tied to market cap or rating cliffs, refinancing wall more than 24 months out, and position sized so a 50% quotational drop invites adding, not explaining.
• Source: BBC interview, 2009 — "If you're not willing to react with equanimity to a market price decline of 50% two or three times a century… you deserve the mediocre result you're going to get." (quote-backed)

Moat mechanisms

M18. Name the moat's mechanism

• Asks: Which specific mechanism protects this company's returns — scale economies, network effects, switching costs, a brand/psychological franchise, a regulatory license, a cost position — and does a decade of the figure table show the mechanism itself holding (margins that survived a downcycle, held or gained share), not just its result?
• Good: One or two named mechanisms, each tied to its fingerprint in the figures: price increases carried through cost inflation, competitors' filings showing worse economics in the same segment, measured switching costs. High returns with no nameable mechanism are presumed mean-reverting; a claimed mechanism whose fingerprint is absent from ten years of figures also fails.
• Source: Talk Two, "A Lesson on Elementary, Worldly Wisdom" (USC, 1994) — the catalog of competitive advantages and their destruction. (teaching-derived)

M19. Volume cost curve

• Asks: Do the filings show unit costs falling as volume grows — gross margin expanding with scale, per-unit opex declining — consistent with a genuine scale-economies advantage?
• Good: A 5–10 year table of revenue vs. gross margin vs. opex ratio that separates true scale and experience-curve effects from one-time cost cuts, benchmarked against the largest competitor's filings.
• Source: Talk Two (advantages of scale — simple geometry and the experience curve). (teaching-derived)

M20. Advertising scale advantage

• Asks: Does advertising and marketing cost per unit of revenue undercut smaller rivals because a large fixed media cost is spread over more volume — and is the gap visible when you compare filings?
• Good: S&M as a percentage of revenue compared against smaller competitors, showing the big player getting more total impact at a lower percentage — the network-TV-era point translated to current media.
• Source: Talk Two (advantages of scale — advertising). (teaching-derived)

M21. Ubiquity as social proof

• Asks: Does being everywhere itself sell the product — is the brand so universally known that availability and familiarity do the selling against any unknown challenger? (Cross-note: the Social-Proof Tendency operating as moat rather than as error.)
• Good: Ubiquity evidence from filings — distribution reach, category share, decades of leadership — plus a concrete account of why a buyer defaults to the known brand rather than save a trivial amount on an unknown one, the Wrigley-vs-Glotz's lesson.
• Source: Talk Two (informational and social-proof advantages of scale — the Wrigley example). (teaching-derived)

M22. Cascade toward dominance

• Asks: Does this market structurally cascade toward one dominant firm — network effects, information goods, daily-newspaper economics — and is this company the emerging winner or the doomed number two?
• Good: Market-share trend rising over ≥5 years plus a named unit-cost or demand-side advantage that mechanically grows with share; applies the newspaper lesson that in winner-take-most markets the number-two position is not half as good — it can be worth nothing.
• Source: "The Art of Stock Picking" / Talk Two — "In some businesses, the very nature of things is to sort of cascade toward the overwhelming dominance of one firm." (quote-backed)

M23. Regulatory license moat

• Asks: Does a license, patent, concession, or regulatory approval keep competitors out — and what are its expiry, renewal terms, and political risk as disclosed in the filings?
• Good: Cites the exact instrument (broadcast license, patent-cliff schedule, utility franchise, airport slots) with dates and the renewal track record, plus the mirror risk: what regulation grants, regulation can revoke or reprice.
• Source: Talk Two (competitive-advantage taxonomy — franchises and government-granted positions). (teaching-derived)

M24. Structural cost position

• Asks: Is this company the demonstrably lowest-cost producer in its industry, with the gap coming from a durable source — location, process, vertical integration, culture — rather than temporary underinvestment?
• Good: A cross-company cost-per-unit comparison built from competing filings, traced to a named structural cause that survived at least one full industry cycle.
• Source: Talk Two (in commodity-like fields, winning by being the low-cost operator). (teaching-derived)

M25. Scale savings shared with customers

• Asks: Does the company hand a measurable part of its scale advantage back to customers, Costco-style, to deepen the moat — rather than harvesting it as peak margin?
• Good: Gross margin deliberately below the peer set while volume, membership, or retention grows and ROIC stays high through asset turns — all three legs shown from the figures; maximizing near-term markup on a scale advantage gets no credit here.
• Source: Daily Journal AGM Q&As and CNBC interviews on Costco (multiple years). (teaching-derived)

M26. Pricing power — exercised or banked

• Asks: Is there filing-level evidence customers pay a premium out of trust or habit — realized price increases at or above inflation with no volume or retention loss — or conspicuously unexercised power, See's/Disney style, sitting there unused?
• Good: Decomposes 5+ years of revenue growth into price vs. volume from the disclosures; either realized increases ≥ inflation with stable units and churn, or a demonstrated gap between price and delivered value that makes the unexercised power a hidden asset the figures don't yet show.
• Source: Talk Two (unexercised pricing power — See's Candies and Disney); "The Art of Stock Picking" (the Wrigley premium). (teaching-derived)

M27. Habit moat

• Asks: Is repeat purchase driven by conscious re-choice or by habit and default — auto-renewal, daily routine, switching hassle — and what event would have to occur to break the habit? (Cross-note: Inconsistency-Avoidance Tendency operating as moat.)
• Good: Locates the product inside a customer routine or default setting, then studies actual churn at natural break-points (price hike, redesign, personnel turnover), compared against habit-moat archetypes like razors, cola, and entrenched ERP.
• Source: Psychology of Human Misjudgment, tendency #5 — "The chains of habit are too light to be felt until they are too heavy to be broken." (quote-backed)

M28. Association moat

• Asks: What does this brand borrow by conditioned association — happiness, status, safety, expertise — how many decades of consistent advertising built the reflex, and what single event could poison it? (Cross-note: Mere-Association Tendency operating as moat.)
• Good: Names the specific association with its evidence (advertising spend held high for decades, repeat-purchase or per-capita consumption data), estimates what a challenger would need to replicate it, and identifies contamination risk — the Coke lesson that the trademarked association in customers' heads, not the physical product, is the asset.
• Source: Psychology of Human Misjudgment, tendency #10; Talk Four (the Coca-Cola exercise — Pavlovian conditioning as business design). (teaching-derived)

M29. Reciprocation as the sales engine

• Asks: Does the business acquire customers or influence gatekeepers through reciprocation — free samples, free tiers, gifts and lunches for doctors, brokers, or purchasing agents — and is that engine durable, scalable, and legal? (Cross-note: Reciprocation Tendency as moat and as hazard.)
• Good: Quantifies free-to-paid conversion where reciprocation targets end users (usually legitimate), and flags reciprocation aimed at agents spending other people's money as a regulatory and ethical time bomb — noting Sam Walton's refusal to let purchasing agents accept so much as a hot dog, a hygiene that is itself a cost advantage.
• Source: Psychology of Human Misjudgment, tendency #9 (Reciprocation). (teaching-derived)

M30. Social proof as the product

• Asks: Is "everyone else uses it" the actual product here — network effects, marketplaces, fashion, review-driven platforms — and if so, what happens to value when the crowd's direction flips? (Cross-note: Social-Proof Tendency as moat.)
• Good: Separates true network effects, where utility mechanically rises with users (exchanges, standards, marketplaces), from fashion-driven social proof, where value equals current coolness and can invert overnight; assigns fashion moats a much shorter duration in the valuation.
• Source: Psychology of Human Misjudgment, tendency #15 (Social Proof). (teaching-derived)

M31. Lollapalooza moat count

• Asks: How many psychological tendencies does this business harness simultaneously on the customer side — habit plus association plus social proof plus reciprocation plus doubt-avoidance?
• Good: Enumerates each tendency with its concrete mechanism in the model (the Coca-Cola pattern: conditioned reflex, happy associations from a century of advertising, social proof of ubiquity, daily habit); one-tendency businesses are fragile, four-plus in combination approach unassailable.
• Source: Psychology of Human Misjudgment, tendency #25 (extreme outcomes come from tendencies in confluence). (teaching-derived)

M32. Autocatalytic growth drivers

• Asks: Which self-reinforcing loops does this business contain — where growth itself causes more growth — and are both ends of each loop visible in the filed numbers?
• Good: Names each loop concretely (more users → better product → more users; more volume → lower unit cost → lower price → more volume) with its two filing-observable measures moving together over 5+ years; distinguishes genuine autocatalysis from market growth the company merely rode.
• Source: Talk Four (seeking autocatalytic effects in the Coca-Cola exercise). (teaching-derived)

M33. Surfing test

• Asks: Is this company riding a large, durable wave — a technology, demographic, or regulatory shift — from the early-scale position that surfing rewards?
• Good: Identifies the wave with independently verifiable size and the company's early-mover share metrics across successive filings; distinguishes riding the wave from arriving as late foam after the economics are competed away.
• Source: Talk Two (the "surfing" model — riding the big wave of a new development). (teaching-derived)

M34. Moat direction, not moat existence

• Asks: Is the moat measurably wider than it was five years ago — or are we admiring a GM/Kodak-style moat while it evaporates?
• Good: At least three of four trend lines favorable across five years: relative market share, realized price vs. peers, customer retention/repeat rate, unit-cost gap vs. the best competitor; any single deteriorating line requires explicit rebuttal.
• Source: Daily Journal AGM Q&As 2014–2017 (GM, Kodak, department stores as vanished moats). (teaching-derived)

Quality & opportunity cost

M35. Opportunity-cost hurdle

• Asks: Is this company clearly better than the best alternative use of the same capital — the strongest existing holding, the best watchlist candidate, the boring index, or simply waiting — compared line-by-line rather than against a formulaic discount rate?
• Good: A named alternative appears in the write-up with a side-by-side on yield, growth, durability, and price, and the candidate wins explicitly; no DCF that hides the hurdle inside a discount-rate assumption. If it does not beat the best thing already available, the verdict is pass no matter how attractive it looks in isolation.
• Source: Investing Principles Checklist (Allocation) — "highest and best use is always measured by the next best use (opportunity cost)"; Berkshire 2003 Q&A rejecting cost-of-capital theater. (quote-backed)

M36. Long-run return gravity

• Asks: Does the expected holding return rest on the business's own return on capital plus reinvestment — not on multiple re-rating?
• Good: Ten-year median ROIC mid-teens or better with demonstrated reinvestment at those rates; thesis return ≈ ROIC × reinvestment rate + shareholder yield with the exit multiple held at or below today's, and a stated price ceiling beyond which the convergence arithmetic stops working.
• Source: "The Art of Stock Picking" — "Over the long term, it's hard for a stock to earn a much better return than the business which underlies it earns." (quote-backed)

M37. Great-at-fair, not fair-at-great

• Asks: Is this a great business at a fair price, or a mediocre business at a bargain price — and which half of that sentence does the figure table actually support?
• Good: Durability evidence (≥8 of 10 years with ROIC above cost of capital, stable or rising gross margin) AND price within a stated fair band; a statistically cheap company with sub-par returns on capital fails regardless of discount.
• Source: Munger aphorism, Wesco/Berkshire Q&As — "A great business at a fair price is superior to a fair business at a great price." (quote-backed)

M38. Cash you can actually take out

• Asks: Over the past decade, what fraction of reported net income converted to free cash available to owners, after real capex, working-capital build, and capitalized development?
• Good: Ten-year cumulative FCF / cumulative net income ≥ ~80% from the figure rows, or a demonstrated case that the gap is value-creating reinvestment with its unit economics shown; the business whose profit permanently sits in the equipment yard fails.
• Source: Berkshire Hathaway AGM 2003 Q&A — "The second earns 12%, but all the excess cash must be reinvested — there's never any cash… 'There's all of my profit.' We hate that kind of business." (quote-backed)

M39. Deferred gratification on the P&L

• Asks: Can you point to a specific, quantified place where management accepted lower near-term reported earnings for long-term value — expensed growth investment, pricing restraint, conservative revenue recognition?
• Good: At least one identifiable P&L drag with its future payoff quantified from filings (growth spend inside SG&A with cohort economics, below-peer pricing with above-peer retention), discussed by management in owner terms.
• Source: Daily Journal AGM 2019 Q&A (low expectations and long attention spans — deferred gratification applied to capital). (teaching-derived)

M40. Raisins mixed with turds

• Asks: Does a genuinely great core segment come packaged with a value-destroying segment large enough to spoil consolidated economics — with the price implicitly paying for the raisins while ignoring the turds?
• Good: Segment-level revenue/EBIT reconciliation in the figure table; bad segments are either small (<15% of capital employed), credibly discontinuable, or fully charged for in the valuation. If segment capital cannot be disentangled from the filings, the item fails toward too-hard.
• Source: Berkshire AGM 2000 Q&A — "If you mix raisins with turds, they're still turds." (quote-backed)

M41. Value, progress, wealth — not price, activity, size

• Asks: Over the last decade, did growth in revenue and assets translate into growth in per-share value drivers, or did the company get bigger while per-share wealth stagnated?
• Good: A ten-year table of revenue, share count, per-share earnings, and per-share equity build showing whether size outran wealth; names the dilution events or value-destructive acquisitions where activity substituted for progress.
• Source: Investing Principles Checklist (Analytic rigor) — "Determine value apart from price; progress apart from activity; wealth apart from size." (quote-backed)

M42. Retained-earnings productivity

• Asks: For each unit of capital this company retained over the past 5–10 years, how much incremental earnings did it produce — does capital allocation show skill or mere habit?
• Good: Computes incremental return as change in earnings ÷ cumulative retained capital from the figure table; scores each allocation channel (capex, M&A, buybacks, dividends) by results, treating serial goodwill impairments as the tuition bill of bad allocation.
• Source: Investing Principles Checklist (Allocation — proper allocation of capital is an investor's number-one job). (teaching-derived)

M43. Uninterrupted-compounding fit

• Asks: Do the reinvestment runway and durability justify holding a decade without interruption — or does the thesis secretly require selling to a greater fool within a few years?
• Good: Runway evidence from filings: addressable-market share, history of reinvesting at high rates, moat durability. An idea that only works with an exit-multiple assumption is classified as a trade, not a compounder, and treated accordingly.
• Source: Talk Two + Investing Principles Checklist (Patience — never interrupt compounding unnecessarily); "sit-on-your-ass investing." (quote-backed)

M44. Pari-mutuel odds check

• Asks: Granting that the business is excellent, are the odds still mispriced — or is the quality already fully paid for at today's multiple, like the standout horse at unplayable odds?
• Good: Converts the current price into implied expectations (growth and margins reverse-engineered against the reconciled figures) and compares those to the evidence — the job is not to spot the best horse but to find the mispriced bet.
• Source: Talk Two (the stock market as a pari-mutuel system). (teaching-derived)

M45. Business quality before manager quality

• Asks: Does the thesis survive replacing current management with merely average operators?
• Good: Re-runs the economics at peer-median margins and capital allocation: still adequate → pass. If the case depends on an exceptional operator, that dependence is stated explicitly and tested against M57 and the succession disclosures.
• Source: "The Art of Stock Picking" — "Averaged out, betting on the quality of a business is better than betting on the quality of management." (quote-backed)

Incentives & management character

M46. What is management actually paid to do?

• Asks: From the officer-compensation and governance disclosures, what exactly is each senior manager paid for — which metrics, over what horizon, with what upside — and what would a rational person do to maximize that package regardless of shareholder outcomes? Whose other incentives touch the reported figures (a listed parent, cross-shareholding partners as captive customers)?
• Good: A table of the actual comp metrics with the distortion each invites (buybacks to juice EPS, empire-building M&A for revenue targets), checked against the company's own history; pay and celebrated KPIs that can only be won by growing per-share owner value. Misaligned comp is close to disqualifying regardless of how the numbers look.
• Source: Psychology of Human Misjudgment, tendency #1 — "Never, ever, think about something else when you should be thinking about the power of incentives." (quote-backed)

M47. Commission chain in the revenue engine

• Asks: Who earns a commission or fee anywhere between this product and its end buyer — salespeople, brokers, doctors, consultants, channel partners — and is the product bought because it is best, or because someone is paid to push it? Is the seller's take large relative to the value delivered?
• Good: Traces the full chain and applies the Xerox test (the inferior machine outsold the better one purely on commission structure); separates revenue that would survive removing the incentive from revenue that would not, remembering that everywhere there is a large commission there is a high probability of a rip-off.
• Source: Psychology of Human Misjudgment, tendency #1 (the Xerox example); Daily Journal AGM Q&As on commission-driven finance products (2017–2019). (teaching-derived)

M48. Anti-gaming controls

• Asks: Where can employees game the company's own metrics and systems — sales quotas, loan volume, cost-plus contracts, adjusted KPIs — and does management design temptation out, cash-register-style, rather than merely exhorting people to be good?
• Good: Identifies the single most gameable internal metric and checks the control environment around it: audit quality, whistleblower record, whether past cheating was punished promptly and visibly. Quota schemes are the predicted failure mode; prevention-by-design is the fix.
• Source: Psychology of Human Misjudgment, tendency #1 — "Perhaps the most important rule in management is 'Get the incentives right.'" (quote-backed)

M49. Comp-ratchet and reset flag

• Asks: Over the last five years of comp disclosures, has the board re-set missed targets, repriced underwater options, or let a compensation consultant ratchet pay to peer percentiles — the envy ratchet where every board pays above median?
• Good: Zero repricings or lowered targets after misses, no automatic "75th percentile of peers" language, CEO pay growth roughly tracking per-share value growth; in people businesses, no envy-driven defections or infighting as the predictable output of visible pay gaps.
• Source: Psychology of Human Misjudgment, tendency #8 (Envy/Jealousy); Wesco Q&As on compensation consultants ("I'd rather throw a viper down my shirtfront"). (teaching-derived)

M50. EBITDA-speak and the honest bridge

• Asks: Where do management's preferred metrics (adjusted EBITDA, non-GAAP EPS, "core" earnings) diverge from GAAP and from your own reconciled figure — and which figure does the thesis actually use?
• Good: A bridge table with every adjustment accepted or rejected for a stated reason; the gap under ~20% of GAAP income and "one-time" items that do not recur three years running. A story that only works in EBITDA fails; the thesis math uses your figure, never theirs.
• Source: Investing Principles Checklist (Intellectual humility — never fool yourself); Wesco/Berkshire Q&As (substitute "bullshit earnings" for EBITDA). (teaching-derived)

M51. Stock comp is compensation

• Asks: What are earnings after treating stock-based compensation as the cash cost it is, and after netting buybacks against option/RSU dilution?
• Good: An SBC-adjusted cash-earnings row plus a five-year diluted share-count trend; per-share earnings still grow attractively after charging SBC at grant value and counting only net — not gross — buybacks.
• Source: Wesco AGM Q&As during the option-expensing fights ("if options aren't compensation, what are they, and if compensation isn't an expense, what is it?"). (teaching-derived)

M52. Buyback price discipline

• Asks: Do repurchases cluster when the shares are cheap and stop when they are dear, Singleton-style — or do they mechanically offset dilution at any price?
• Good: A year-by-year table of buyback spend against that year's valuation multiple showing inverse correlation; no debt-funded repurchases at peak multiples; a stated valuation condition for buybacks somewhere in the filed record.
• Source: Wesco/Daily Journal Q&As praising Henry Singleton's capital allocation. (teaching-derived)

M53. Serial-acquisition empire test

• Asks: If the company grows by acquisition, what return has acquired capital actually earned — and is the CEO paid or celebrated for size rather than returns?
• Good: Cumulative acquisition spend vs. incremental operating profit clears the opportunity-cost hurdle; no pattern of goodwill impairments trailing deals by 2–4 years; the confident deal-time rhetoric checked against realized results, since most acquisitions are done out of over-optimism and empire-building.
• Source: Wesco AGM Q&As on why acquisitions fail buyers; Psychology of Human Misjudgment, tendency #12 (Excessive Self-Regard). (teaching-derived)

M54. Character screen

• Asks: What does the filed record show about the character of the people you would be compounding with — related-party transactions, restatements, auditor changes, regulator actions, executive turnover, pledged shares, insider loans, self-dealing at prior companies?
• Good: Works the proxy and filings item by item: a clean answer cites the absence of each red flag after actually checking, a dirty answer names transactions and dates. One material grime instance fails the item — no averaging, because no price compensates for questionable character.
• Source: Investing Principles Checklist (Risk — avoid people of questionable character); Wesco/Daily Journal Q&As ("we've never made a good deal with a bad person"). (teaching-derived)

M55. Track one — interest map of the ecosystem

• Asks: For every actor in this company's ecosystem — customers, suppliers, distributors, employees, regulators, managers — what does each rationally want, and does the business model align those interests or fight them?
• Good: A literal table of actor → rational interest → filing evidence (supplier concentration notes, contract terms, insider ownership); flags any actor whose self-interest, rationally pursued, erodes the company.
• Source: Talk Four ("Practical Thought About Practical Thought?" — two-track analysis, track one). (teaching-derived)

M56. Salesmanship-in-the-report screen

• Asks: Which parts of the annual report exist to inform and which exist to sell — and once the promotional layer (adjusted metrics, TAM slides, glossy narrative) is stripped away, what do the bare audited numbers say?
• Good: Physically separates the audited statements and notes from the narrative sections; lists every load-bearing claim that appears only in the narrative with no audited counterpart. The thesis relies exclusively on the audited layer — the document's authors respond to incentives too.
• Source: Talk Four (two-track analysis applied to communications and persuasion). (teaching-derived)

M57. Edison-and-Welch founder test

• Asks: Where the case does rest on a founder/operator: is there documented evidence of both technical problem-solving (products or costs others couldn't achieve) and execution (targets set in prior filings that were met)?
• Good: At least three verifiable achievements from the dated public record — cost curves bent, launch dates hit, guidance met — plus meaningful personal ownership held through drawdowns rather than sold into strength.
• Source: Fortune interview, 2009, on BYD's Wang Chuanfu — "a combination of Thomas Edison and Jack Welch." (quote-backed)

M58. Seamless web of deserved trust

• Asks: Does the company run on deserved trust — decentralized authority, low bureaucracy, long-tenured operators — or on procedure, litigation, and internal policing?
• Good: Filing evidence of low executive turnover (median officer tenure >7 years), a thin corporate center relative to revenue, absence of chronic whistleblower/litigation disclosures, and stable, named segment management.
• Source: USC Gould School of Law commencement, 2007 — "The highest form that civilization can reach is a seamless web of deserved trust." (quote-backed)

M59. Ideology and promotion flag

• Asks: Does management communication read like dispassionate ownership or like ideology and promotion — new-era language, mission-talk substituting for economics, refusal to name mistakes?
• Good: The last three shareholder letters each admit at least one specific, quantified mistake; call answers cite numbers rather than adjectives; no "this time is different" framing of deteriorating figure-table trends.
• Source: USC Gould commencement, 2007 — intense ideology "cabbages up one's mind." (quote-backed)

M60. Bad-news candor audit

• Asks: Comparing management's discussion of each major problem across consecutive years, are big troubles named, quantified, and faced — and does bad news reach shareholders promptly, or do messengers get shot?
• Good: A promise-vs-outcome side-by-side for two or three past problems, with credit for pre-announced problems, autopsied failures, and conservative guidance; a history of surprises that outsiders discovered first is evidence the internal channel kills Persian messengers.
• Source: Investing Principles Checklist (Focus — face your big troubles); Psychology of Human Misjudgment, tendency #10 — "Always tell us the bad news promptly. It is only the good news that can wait." (quote-backed)

Psychology of misjudgment

This section walks the tendencies from "The Psychology of Human Misjudgment" as sources of error — in management, in the market, and in the analyst. Where a tendency IS the moat (habit, association, reciprocation, social proof, lollapalooza on the customer side), the check lives in the moat section (M27–M31).

M61. Whose psychology is the story built on?

• Asks: Read the IR narrative and mid-term plan for the misjudgment tendencies — social proof (strategy justified because peers are doing it), commitment-and-consistency (escalating a segment ten years of figures say is failing), authority (a plan wrapped in consultant or government-initiative endorsement instead of unit economics) — and run the same track-two audit on the market's pricing and on your own attraction to the idea.
• Good: Each tendency named with its location and the specific figure that contradicts the story (the multiple, the guidance record, your own dated notes). Management that has visibly exited a failed commitment scores higher than management with an unbroken record of self-praise.
• Source: Talk Four (two-track analysis — subconscious influences); Psychology of Human Misjudgment (Social-Proof, Inconsistency-Avoidance, Authority-Misinfluence tendencies). (teaching-derived)

M62. Liking-blindness self-check [PRACTITIONER]

• Asks: What do I like or admire about this company, its founder, or its product — and would each admired "fact" survive verification by someone who feels nothing for it?
• Good: Writes the affection list explicitly (love the product, admire the CEO's story) and re-sources each load-bearing claim from documents rather than from the admired person, because liking makes one ignore faults and distort facts in favor of the loved object.
• Source: Psychology of Human Misjudgment, tendency #2 (Liking/Loving). (teaching-derived)

M63. Hatred-distortion check [PRACTITIONER]

• Asks: Is this company or industry widely hated, or personally repellent to me — and is that dislike distorting the market's price (opportunity) or my own weighing of facts (risk)?
• Good: Separates moral discomfort from economic analysis, checks whether the hated-sector discount is the actual source of expected return, and re-weighs cold the facts one is tempted to dismiss merely because they favor a disliked object.
• Source: Psychology of Human Misjudgment, tendency #3 (Disliking/Hating). (teaching-derived)

M64. Doubt-avoidance speed check [PRACTITIONER]

• Asks: What is actually forcing a decision deadline on this idea — and if nothing is, why am I in a hurry to conclude?
• Good: States the real cost of waiting two more weeks (usually near zero) and imposes a deliberate cooling-off delay on any idea found under excitement or time pressure, since puzzlement and stress are exactly what make the mind grab a fast, bad conclusion.
• Source: Psychology of Human Misjudgment, tendency #4 — "The brain of man is programmed with a tendency to quickly remove doubt by reaching some decision." (quote-backed)

M65. Management status-quo lock-in

• Asks: Where is management maintaining a strategy mainly because they publicly committed to it or have always done it — dying divisions kept alive, legacy pricing defended, refusal to cannibalize the core product?
• Good: Finds at least one instance of management reversing a prior public commitment when facts changed (strong signal), or conversely a textile-mill situation fed capital long past economic sense; treats loud public strategy pledges as consistency traps that raise the cost of future correction.
• Source: Psychology of Human Misjudgment, tendency #5 (Inconsistency-Avoidance). (teaching-derived)

M66. Fairness of the business model

• Asks: Would a customer who fully understood the pricing, terms, and fine print feel fairly treated — or does profit depend on hidden fees, gotcha renewals, and exploited inattention?
• Good: Reads the actual customer-facing contract and fee schedule; splits the profit pool into "paid happily for value received" versus "extracted from confusion," modeling the second bucket as melting ice that attracts regulators — while noting that conspicuous fair dealing compounds into a trust moat.
• Source: Psychology of Human Misjudgment, tendency #7 (Kantian Fairness). (teaching-derived)

M67. Luck-versus-skill association check

• Asks: Is management's (or my own) confidence built on real causation, or on success merely associated with a bull market, one lucky product, or a predecessor's groundwork?
• Good: Decomposes the track record — organic versus acquired growth, cycle tailwind versus share gain, results in the only downturn the CEO has faced — examining each past success for accidental, non-causative factors before extrapolating it.
• Source: Psychology of Human Misjudgment, tendency #10 (Mere-Association — the prescribed antidote). (teaching-derived)

M68. Denial-of-decline test

• Asks: Which fact about this business is so painful — secular decline, a dead flagship product, broken unit economics — that management, or I, might be distorting it to make it bearable?
• Good: States the most painful plausible fact out loud, then checks what management does rather than says about it: reinvestment, capex, and headcount decisions reveal whether the decline is accepted or denied. The same test applies to one's own thesis after a drawdown.
• Source: Psychology of Human Misjudgment, tendency #11 — "reality is too painful to bear, so one distorts the facts until they become bearable." (quote-backed)

M69. Endowment re-underwrite [PRACTITIONER]

• Asks: If I did not already own or follow this stock and encountered it fresh today at today's price, would I buy it — or am I over-appraising it merely because it is mine and the idea was mine?
• Good: Re-underwrites each holding periodically from a blank page at the current price and sells when the fresh case would not justify buying; recognizes that self-generated conclusions get the same irrational premium as self-picked lottery numbers.
• Source: Psychology of Human Misjudgment, tendency #12 (Excessive Self-Regard — the endowment effect). (teaching-derived)
• Agent proxy: re-derive intrinsic value from the figure table without referencing any prior valuation in the study, and verify the fresh derivation matches the one already stated.

M70. Excuse-culture (Tolstoy) test

• Asks: When this company misses — a failed product, a lost quarter, a botched integration — does the culture fix the failure or manufacture reasons why it wasn't really a failure?
• Good: Reads several years of letters and calls for the ratio of ownership ("we misjudged X, here is what changed") to weather-and-currency excuses; people don't fix what they have excused, so a fluent excuse culture predicts repeat failure.
• Source: Psychology of Human Misjudgment, tendency #12 (the Tolstoy effect). (teaching-derived)

M71. Projection-optimism haircut

• Asks: Which numbers in this thesis are wishes wearing spreadsheets — management guidance, TAM slides, hockey-stick margin inflections — and what do base rates and simple probability math say instead?
• Good: Replaces guidance with the historical base rate for comparable companies (how many ever sustained the assumed growth for a decade) and runs the thesis in expected-value terms across scenarios rather than pricing only the happy one.
• Source: Psychology of Human Misjudgment, tendency #13 — "What a man wishes, that also will he believe." (quote-backed)

M72. Loss-chasing escalation check [PRACTITIONER]

• Asks: Is anyone here doubling down to avoid crystallizing a loss — management pouring good money into a failing project because so much is already sunk, or me averaging down mainly to get back to even?
• Good: Distinguishes averaging down on verified-unchanged fundamentals from deprival-driven chasing; flags "we've invested too much to stop now" language and credits managements that kill projects and take writedowns early.
• Source: Psychology of Human Misjudgment, tendency #14 (Deprival-Superreaction). (teaching-derived)

M73. Take-away and scarcity dynamics

• Asks: Does this business risk deprival-superreaction from its own customers — removed features, sharp price hikes, New-Coke moves — or deliberately exploit the tendency through artificial scarcity, drops, and near-miss mechanics?
• Good: Reviews the price-increase history and the customer reaction each time; where scarcity and loss-triggering are the core engine, asks explicitly what happens when regulators or customer fatigue arrive, and prices that revenue on a shorter duration.
• Source: Psychology of Human Misjudgment, tendency #14 (the New Coke episode). (teaching-derived)

M74. Copycat capital allocation

• Asks: Are management's recent big moves — buybacks at highs, trendy acquisitions, me-too capex waves, entries into hot segments — independently justified by their own economics, or do they track what every peer did at the same moment, driven by imitation and envy?
• Good: Dates each major capital decision against the industry fashion cycle and quantifies whether the me-too moves earned their cost of capital; credit goes only to allocators who acted against the crowd — buying back when hated, sitting out bidding wars, expanding when peers retrenched.
• Source: Psychology of Human Misjudgment, tendency #15 — "it is not at all uncommon to find leaders who display followership akin to that of teenagers"; Daily Journal AGM 2022 — "The world is not driven by greed. It's driven by envy." (quote-backed)

M75. Independent-thesis entry check [PRACTITIONER]

• Asks: Was the thesis built from your own reconciled figures before reading any analyst target, 13F, or admired investor's write-up — or is "everyone I respect owns it and someone no smarter than me just got rich on it" doing the work?
• Good: The write-up derives value from the figure table alone, with consensus cited only as prevailing opinion to be tested; 13F cloning plus envy of a neighbor's gains is a null thesis requiring a restart from primary documents.
• Source: Psychology of Human Misjudgment, tendency #15 — "Learn how to ignore the examples from others when they are wrong, because few skills are more worth having." (quote-backed)

M76. Sector-mood context

• Asks: Is this company or sector currently adored or shunned — measured by its own valuation history against its ten-year range and by equity-issuance volume in the sector?
• Good: Places today's multiple within its historical distribution built from filed financials; heavy issuance waves mean insiders think shares are dear. The buy case is strongest when the mood is fear but the filings show the business intact.
• Source: Investing Principles Checklist (Decisiveness — fearful when others are greedy, greedy when others are fearful). (teaching-derived)

M77. Contrast-anchor valuation check

• Asks: Is "cheap" here an absolute judgment against cash flows, or a contrast effect — cheap versus the 52-week high, versus a crazier peer, versus its own bubble multiple?
• Good: Values the business standalone before consulting any comparative anchor, and recognizes the real-estate-broker trick (show two awful houses first) when bankers present comps or IPO ranges — a stock down 80% is not cheap merely because it contrasts with its old price.
• Source: Psychology of Human Misjudgment, tendency #16 (Contrast-Misreaction). (teaching-derived)

M78. Boiling-frog trend check

• Asks: What has deteriorated in small, low-contrast steps — gross margin, market share, product quality, culture — that a ten-year comparison would make obvious even though no single quarter alarmed anyone?
• Good: Builds a ten-years-ago versus today table for the key operating metrics rather than reading sequential quarters; treats a slow monotonic slide as destiny unless a specific, funded reversal mechanism exists.
• Source: Psychology of Human Misjudgment, tendency #16 (low-contrast trends as destiny — the boiled frog). (teaching-derived)

M79. Stress on the decision-makers [PRACTITIONER]

• Asks: Who in this situation is deciding under heavy stress — a forced seller, a levered management in a liquidity crunch, a founder mid-crisis, or me during a drawdown — and how does that change what prices and actions mean?
• Good: Identifies stress-driven selling as the likely source of the bargain (or stressed management fire-sales as the risk); as a personal rule, refuses to make buy/sell decisions during acute personal stress, since heavy stress degrades judgment even as light pressure sharpens it.
• Source: Psychology of Human Misjudgment, tendency #17 (Stress-Influence — Pavlov's flood-stressed dogs). (teaching-derived)

M80. Availability audit [PRACTITIONER]

• Asks: Is this thesis built on the evidence that matters or the evidence that was easy to get — the vivid anecdote, the latest quarter, the product I personally use, the CEO's charismatic interview?
• Good: Lists the three facts driving the most conviction and asks of each: load-bearing, or merely available and vivid? Then deliberately hunts the unavailable data — private competitors, foreign filings, churned customers, ex-employees — before concluding.
• Source: Psychology of Human Misjudgment, tendency #18 — "An idea or a fact is not worth more merely because it's easily available to you." (quote-backed)
• Agent proxy: explicitly list the three load-bearing facts of the answer with their sourcing, flagging any anecdote not backed by a figure-table row or archived source.

M81. Moat-skill atrophy

• Asks: Which specific skills does this company's advantage depend on, and is the organization still practicing them — or have they been outsourced, financialized, or cost-cut away?
• Good: Traces where the critical craft physically lives (in-house teams versus vendors), what happened to the people who carried it, and whether efficiency programs removed the practice repetitions the moat quietly requires.
• Source: Psychology of Human Misjudgment, tendency #19 — "All skills attenuate with disuse." (quote-backed)

M82. Senescence and succession

• Asks: How dependent is this thesis on aging key people, and is there evidence of maintained sharpness plus a practiced succession plan — operators already running real segments — rather than a name on a slide?
• Good: States key-person ages and exactly what breaks on their exit (deal judgment, cost culture, creative taste); looks for intensely practiced skill and continued joyful learning in the aging leader, and for successors with observable track records inside the company.
• Source: Psychology of Human Misjudgment, tendency #21 (Senescence-Misinfluence). (teaching-derived)

M83. Who can tell the CEO no

• Asks: If the admired founder-CEO steers toward a cliff, who in this company has the standing to grab the controls — or is the board a crew of copilots who would let the plane crash rather than override the captain?
• Good: Genuine independents with relevant expertise and reputations of their own, plus at least one historical instance of the CEO being overruled, or of a dissenting senior executive surviving and being promoted.
• Source: Psychology of Human Misjudgment, tendency #22 (Authority-Misinfluence — the copilot simulator experiments). (teaching-derived)

M84. Twaddle density in communications

• Asks: Strip the buzzwords from the shareholder letter and earnings call — what fraction remains as falsifiable statements, and does the organization keep its prattlers away from the serious work?
• Good: Rewrites a paragraph of the letter as plain claims checkable within a year: a candid letter survives the translation, a twaddle letter evaporates. Buzzword density and non-GAAP adjustment counts against sober peers serve as a proxy for internal honesty.
• Source: Psychology of Human Misjudgment, tendency #23 (Twaddle). (teaching-derived)

M85. The reason test — real whys, three layers deep

• Asks: For the company's headline economics, can you answer "why?" three layers deep using only the filings — and do management's stated reasons carry causal content, or are they reason-shaped noise that merely restates the conclusion?
• Good: Every "because" in the chain cites a filing fact down to causal bedrock ("margins 30% — why? pricing power — why? switching costs — why? disclosed ten-year customer relationships"); "we will win because we are the leader" is flagged as the Xerox-queue trick, and letters explain why capital went where it went — Braun's who, what, where, when, and why.
• Source: Investing Principles Checklist (Preparation — ask why, why, why); Psychology of Human Misjudgment, tendency #24 (Reason-Respecting — Carl Braun's rule). (teaching-derived)

M86. Lollapalooza test — one force or many?

• Asks: Are the good results driven by a single factor or by several independent forces compounding — moat plus aligned incentives plus secular tailwind plus net-cash balance sheet? And is a negative lollapalooza forming: several separately-survivable problems (shrinking end market, high fixed costs, financing pressure) whose combination is a spiral?
• Good: Two or three mutually reinforcing advantages, each independently evidenced in the figure table — not one advantage restated three ways (the Coke pattern: brand conditioning + distribution scale + social proof + habit). Symmetrically, co-occurring negative forces (leverage + customer concentration + secular decline) are analyzed jointly, never dismissed one at a time.
• Source: Psychology of Human Misjudgment, tendency #25; Talks Four and Eleven (really big effects come from large combinations of factors operating in the same direction). (teaching-derived)

M87. Lollapalooza aimed at the buyer

• Asks: Is the buying situation itself engineered to combine tendencies against me — an IPO roadshow (authority plus social proof plus scarcity), a hot competitive auction, a FOMO-laden narrative stock — and is the right move simply not to enter the room?
• Good: Recognizes structured-persuasion settings — roadshows, competitive deal processes, open-outcry-style bidding — as lollapalooza machines, and either abstains entirely or pre-commits to a maximum price set alone in a cold room beforehand. For the open-outcry auction, the antidote is Buffett's: don't go.
• Source: Psychology of Human Misjudgment, tendency #25 (the auction as stacked tendencies). (teaching-derived)

Process & discipline

M88. Margin-of-safety gap

• Asks: After reconciling the figures, what conservative intrinsic-value range does the evidence support, and how large is the gap between the LOW end of that range and the current price?
• Good: A value range built under conservative assumptions with the current price meaningfully below the low case, and the discount stated explicitly ("price is ~60% of low-case value") rather than compared against a single optimistic point estimate.
• Source: Investing Principles Checklist (Risk — incorporate a margin of safety in all valuation). (teaching-derived)

M89. Circle-of-competence plain statement

• Asks: Can you state in one plain-language paragraph exactly how this company turns inputs into cash, and list the parts of its filings you do NOT understand?
• Good: The paragraph survives a skeptic's reading without jargon; the don't-understand list is explicit and honest ("I cannot independently assess the actuarial reserve adequacy"), with a stated response: get expert help, size down, or pass.
• Source: Investing Principles Checklist (Intellectual humility — stay within a well-defined circle of competence). (teaching-derived)

M90. Decisive-variables declaration

• Asks: Can the entire thesis be stated in one paragraph with at most three load-bearing variables — each named in advance, traceable to the figure table, and shown to be judgeable from the assembled sources?
• Good: The paragraph exists and names its variables ("renewal rate stays ≥90%, pricing +3%/yr, share count flat"); everything else is supporting detail. For any variable that is genuinely unknowable (regulatory whim, commodity price, scientific outcome), the study shows the thesis is insensitive to it or routes to too-hard.
• Source: Investing Principles Checklist (Focus — keep things simple); Daily Journal AGM Q&As on knowing the edge of your own competency (2014–2017). (teaching-derived)

M91. False-precision guard

• Asks: Does the valuation conclusion survive replacing every point estimate with a plausible range — does any part of the decision hinge on second-decimal precision?
• Good: Value presented as a range with price below the low case, no terminal-value knife edge, and a sensitivity table showing the thesis works at the conservative bounds of each input.
• Source: Investing Principles Checklist (Intellectual humility) — "Resist the craving for false precision, false certainties, etc." (quote-backed)

M92. Macro-independence of the thesis

• Asks: Strip every macro forecast (rates, GDP, elections, commodity calls) out of the write-up — does the thesis still stand on business-level facts from the filings?
• Good: What remains rests on unit economics, competitive position, and the balance sheet; unavoidable macro sensitivity is handled as a stated range, never as a prediction the thesis depends on.
• Source: Investing Principles Checklist (Analytic rigor — be a business analyst, not a market or macroeconomic analyst). (teaching-derived)

M93. Obvious-over-esoteric test

• Asks: Does the thesis rest on remembering obvious facts (dominant share, decades of stable margins, net cash) or on grasping something esoteric — an intricate sum-of-parts, an accounting subtlety the market allegedly missed?
• Good: The three load-bearing facts are each verifiable in five minutes from the filings; if the edge claimed is esoteric insight, that is flagged as fragility to be discounted, not as advantage.
• Source: Investing Principles Checklist (Analytic rigor — better to remember the obvious than to grasp the esoteric). (teaching-derived)

M94. Napkin-math no-brainer

• Asks: Does simple arithmetic — units × price × margin × years, done on one page from the reconciled figures — get you most of the way to the conclusion before any model is built?
• Good: The one-page version exists and roughly agrees with the full valuation, with the big no-brainer questions answered first (how big can the market be, what share and margin are plausible); if elementary math can't carry the thesis, it is reclassified as speculation.
• Source: Talk Four (the two-trillion-dollar Coca-Cola question, solved with big no-brainer sub-questions and high-school math). (teaching-derived)

M95. Latticework pass

• Asks: Run the company deliberately through at least four disciplines — mathematics (compounding, unit economics), psychology (customer and market behavior), microeconomics (moats, cost curves), biology/ecology (niche and competitive ecosystem): what does each model reveal that the others miss?
• Good: One paragraph per model, each anchored to a filing fact; the exercise exposes the man-with-a-hammer failure of judging the company through a single metric like the P/E ratio.
• Source: Talk Two (elementary worldly wisdom — a latticework of mental models). (teaching-derived)

M96. Cancer-surgery test

• Asks: When parts of this business went bad, did management cut them out cleanly and refocus on the healthy core — and if the thesis is a turnaround, is there an identifiable, historically wonderful core visible in past filings, so the fix is cutting away folly rather than inventing a new business?
• Good: A ten-year list of exits with proceeds and rationale, the GEICO pattern — strip away the folly and return to the wonderful business at the core. The failing answer: chronic "turnaround" segments absorbing the good segments' capital, or a plan pivoting into markets where the company has no record.
• Source: Talk Two / "The Art of Stock Picking" (the cancer-surgery formula — GEICO). (teaching-derived)

M97. Kill conditions before purchase

• Asks: What specific, filing-observable events would falsify this thesis — written down before purchase — and what have you concretely done to seek disconfirming evidence before the first conclusion hardened?
• Good: Three to five kill conditions tied to reportable numbers ("gross margin below X for two straight years," "share count grows >2%/yr," "founder departs"), dated before the buy decision and reviewed at every filing; Darwin's asymmetry applied — extra suspicion for the hypotheses that please you most.
• Source: Investing Principles Checklist (Allocation — don't fall in love with an investment); Psychology of Human Misjudgment, tendency #5 (Darwin's disconfirmation discipline). (teaching-derived)

M98. Bear case a short-seller would endorse

• Asks: Has the study produced a bear case a committed short-seller would endorse — the strongest facts against the thesis findable in the company's own filings, with exact locations — before the buy conclusion was written?
• Good: At least three quantified negatives drawn from the reconciled figures (a deteriorating segment, risk-factor language hardening year over year, rising receivables, customer concentration), each rebutted with other figures or explicitly accepted and priced. If it reads weaker than a real short report on the name, the item fails.
• Source: Investing Principles Checklist (Intellectual humility — reconcile disconfirming evidence); Munger's standing rule — know the other side's argument better than they do. (teaching-derived)

M99. Bet-heavily conviction test

• Asks: Is this one of the rare good ideas that deserves a heavy weighting on the evidence assembled — and if the honest answer is "only a small position feels safe," why is it a position at all?
• Good: Pre-set conditions of price and evidence, defined in advance, are now met — "I finished the work and the price fell below my low case," not "the stock fell and I got interested" — and conviction is expressed as sizing logic with two or three figure-backed reasons, not adjectives.
• Source: "The Art of Stock Picking" — "The wise ones bet heavily when the world offers them that opportunity… And the rest of the time, they don't. It's just that simple." (quote-backed)

M100. Drowning-in-cash patience check

• Asks: Is this purchase driven by price-versus-value today, or by accumulated cash and the itch to act — would you still buy if required to wait 90 days?
• Good: The trigger for acting now is a documented price/value gap or a dated catalyst from filings — not portfolio cash level, FOMO on a rising quote, or study-completion momentum; a "why now" section citing figure rows exists.
• Source: Wesco AGM Q&A — there are worse situations than drowning in cash and sitting, sitting, sitting. (teaching-derived)

M101. Too-hard basket honesty

• Asks: Can the analyst explain, in one paragraph with every claim tied to a figure-table row, how this company makes materially more money in ten years — and if not, has the study been routed to too-hard rather than padded?
• Good: The paragraph survives having each sentence footnoted to a figure row or filing citation; any sentence that can't be footnoted forces either more source work or a too-hard verdict, and "probably fine" language anywhere in it fails.
• Source: Wesco/Berkshire Q&As — three baskets: in, out, and too tough, with most things in the too-tough basket. (teaching-derived)

M102. Fish-where-the-fish-are pond check

• Asks: Why does this bargain exist — is the company in an under-fished pond (neglect, forced selling, index exclusion, small size, unfashionable geography), or a heavily-fished mega-cap where mispricing requires believing you out-see everyone?
• Good: A named, evidence-backed reason for the mispricing (no analyst coverage, recent index deletion, an uneconomic seller) documented in the write-up; "the market underappreciates the quality," alone, on a name with thirty covering analysts, fails.
• Source: Daily Journal AGM 2018 — "The first rule of fishing is 'fish where the fish are.' And the second rule of fishing is 'don't forget the first rule.'" (quote-backed)

M103. Standard-stupidities abstention

• Asks: What are the three standard ways companies in this industry destroy themselves — underwriting for volume, expanding at cycle tops, stuffing the channel, chasing yield — and does the figure table show this company abstaining from each?
• Good: The three industry-specific stupidities named from historical failures, each with a company-level metric across the cycle (reserve development, capacity additions vs. cycle position, receivables vs. revenue growth) demonstrating abstention.
• Source: Wesco-era remark — "trying to be consistently not stupid, instead of trying to be very intelligent." (quote-backed)

Verdict guidance

Invert first. Before scoring anything, write the obituary: if M1 — or any of the failure-mode items behind it (M2–M17) — finds a plausible death path already in motion, the verdict is too-hard or pass no matter the multiple. Quality and moat durability (M18, backed by M19–M34) outrank statistical cheapness — a mediocre business at a bargain loses to a great business at a fair price (M37), because the great business does the compounding for you. M46 (incentives) and M61 (narrative psychology) act as character gates: misaligned incentives or a story that fights the figure table caps the verdict at watch even when the economics clear, with the specific contradiction recorded as the re-check trigger. Where the book cannot be independently reconstructed or the decisive variables are unjudgeable (M12, M90, M101), route to too-hard rather than haircutting.

Buy-below-¥X is set by the opportunity-cost hurdle (M35), not an absolute yardstick: the price at which this business beats the named best alternative, with room to be wrong (M88, M91). Between pass and buy, the default is to sit on your hands (M100) — few names clear all the gates, and the discipline is to do nothing until one does rather than to lower the bar. A lollapalooza on the positive side (M86, and on the customer side M31) is what upgrades watch to buy-below; it should be rare.

Archive-depth rule. When the archived sources are shallower than an item requires (fewer years, missing competitor/customer filings, absent disclosures), the item is answered data-insufficient with the missing source named — never estimated from general knowledge without an explicit [general-knowledge: not-from-archive] flag. Load-bearing items stuck at data-insufficient push the verdict toward too-hard or watch, not toward a guess.

A too-hard verdict names the specific item ID(s) that triggered it. Item-level answers use pass / fail / data-insufficient / practitioner-pending; items after a too-hard early exit are marked not-evaluated, never silently skipped.