——
Level — · Streak —
— XP
Mastery—weighted
Quizzes 30d—last 30 days
Cert progress—active level
Rank—global
Loading...
The page behind this dialog is live. Create a free account or sign in and you'll land right back on it.
The hardest decision in investing — made systematic
In 1980, an investor in Walmart faced a hard choice. The stock had risen ~50% since IPO three years earlier. Conventional advice said 'take some off the table.' The investor sold half the position. By 2020, that decision cost them roughly \$2 million per \$10,000 sold — Walmart compounded for forty more years. The opposite case: an investor in Sears Holdings in 2007 watched the stock rise ~30% on a hedge-fund 'turnaround' thesis; conventional advice said 'let your winners run.' The investor held. By 2018, Sears was bankrupt and the position was worthless. Selling and holding both have catastrophic failure modes — but the failure modes are completely different. Selling too early is the lost-compounder failure; holding too long is the lost-thesis failure. This lesson is about systematizing the decision so the failure mode you avoid is the one you're actually in.
The decision to is statistically more bias-prone than the decision to buy because it triggers two distinct behavioral failures simultaneously: loss aversion (the pain of realizing a loss exceeds the pleasure of an equivalent gain by 2-2.5x — Kahneman-Tversky) and the disposition effect (selling winners early to 'lock in' gains and holding losers indefinitely to 'avoid realizing' the loss — Odean 1998 documented this in 10,000 retail accounts). The combination is uniquely destructive: most retail investors sell their best positions too early and hold their worst ones too long, producing the reverse-Pareto pattern where the largest dollar weights end up being the worst-performing theses. Pre-defined sell rules — written before any specific position is held — are the standard discipline because they replace emotional in-the-moment decisions with binding pre-commitments.
There are exactly five valid sell triggers; everything else is emotion. (1) THESIS BROKEN. The specific reason you bought the stock no longer applies. The KPIs you identified at purchase have deteriorated structurally; the moat has eroded; the addressable market has shrunk; the management team you trusted has been replaced. Verifiable on the KPIs and Ratios tabs. (2) EXTREME OVERVALUATION. The position has run far above your estimated fair value range — typically 50%+ premium to your DCF and multi-method triangulation (M6 framework). At extreme overvaluation, the expected forward return is poor regardless of business quality. Verifiable on the Valuation tab. (3) BETTER OPPORTUNITY. The /screener page surfaces a name with materially better risk/reward at the same conviction level. Selling the held name to fund the new one is a portfolio-improving trade. (4) POSITION SIZE EXCEEDS LIMIT. A position has grown above your IPS-imposed maximum (m8_l2 sizing framework). Trim back to the limit even if conviction is intact — the limit exists precisely because conviction sometimes turns out to be wrong. (5) MATERIAL CIRCUMSTANCE CHANGE. Your time horizon, risk tolerance, or financial needs have shifted (approaching retirement, large cash need, etc.). None of those triggers includes 'price went down.' A price decline absent any of the five is NOT a sell signal — and in many cases it's a buy signal if the thesis is intact.
For every position, periodically ask: 'If I had cash equal to this position's value, and I knew everything I now know about the business, would I open this position TODAY at the current price?' If yes, hold. If no, sell — regardless of what you paid, regardless of whether you're up or down on the position. The reframe works because it strips away the entire emotional tax of the position (your purchase price, your unrealized gain or loss, your sunk-cost attachment) and forces you to evaluate the position on the same basis you'd evaluate any new investment opportunity. The Howard Marks 2000 memo 'On the Couch' frames the same idea: the investor's purchase price and the stock's future returns are independent random variables. Charlie Munger's framing in his 'A Lesson on Elementary, Worldly Wisdom' (1994 USC commencement): 'A great business at a fair price is superior to a fair business at a great price.' Your purchase price tells you nothing about the future; the current business quality and current valuation tell you everything.
Sequoia Fund (founded 1970 by Bill Ruane, the Ben Graham disciple Buffett personally recommended to Berkshire shareholders when Buffett closed his original partnership) has been one of the most respected concentrated-portfolio funds in US investment history. By 2014-2015, Sequoia held a roughly 30% position in Valeant Pharmaceuticals — extraordinary concentration even by Sequoia's standards. Valeant's business model was acquisition-driven roll-up plus aggressive drug repricing. In 2015, short-seller research and Senate investigations exposed accounting irregularities (Philidor specialty pharmacy revenue recognition issues) and the unsustainability of the price-hike strategy. The stock fell from ~\$260 to under \$30 over roughly twelve months — a ~90% decline. Sequoia did not exit until late in the move. The fund's longtime manager, Robert Goldfarb (Bill Ruane had passed in 2005), departed in 2016. The fund subsequently rebuilt under a new team with much more diversified concentration. The lesson is two-fold. First: the SELL trigger that should have fired was 'thesis broken' — Valeant's acquisition-and-repricing model depended on continuation of a regulatory and political environment that broke in 2015 (M&A drying up, drug-pricing scrutiny intensifying, the Philidor revelations). The cheapness of the stock relative to its peak was a value trap; the underlying business had structurally weakened. Second: the SIZE was the underlying error. A 30% position in any name leaves no room for being wrong, and at sufficient concentration even a correct thesis can be sized into permanent capital loss if a single failure mode fires. The combination of unbroken loyalty to the position (no thesis-broken sell trigger) and extreme sizing (no size-cap discipline) produced a ~5-10pp/year drag on Sequoia's long-run returns. The fund recovered, but the cohort of Sequoia investors who sold during the 2015-2016 redemption window experienced permanent losses. Cite: Valeant 2015-2016 episode; Sequoia Fund 13F filings 2014-2017; the canonical retrospective is Charlie Gasparino's 2016 Fortune piece and the Sequoia 2015 annual report. The Sequoia 2015 annual report itself is unusually candid about the mistake — worth reading as a primary source on the discipline of acknowledging error.
The Stock Analysis page on every name is structured around the five sell triggers. The KPIs tab tracks the fundamentals you defined as the thesis at purchase — review quarterly to detect thesis-broken conditions before they become severe. The Ratios tab supports the same review with the M5 ratio framework (margins, ROIC, leverage, cash-flow conversion). The Valuation tab triangulates fair value across multiple methods (M6) — when current price exceeds fair value by 50%+, the extreme-overvaluation trigger fires. The /screener page lets you compare current holdings against alternatives at the same conviction level — surfacing the better-opportunity sell trigger. The Portfolio Hub displays each position as a percentage of total portfolio with target-vs-actual comparison against your IPS limits — surfacing the size-exceeds-limit trigger. The Insights tab on each holding flags significant management changes, regulatory shifts, and competitive position erosion — supplemental context for the thesis-broken trigger. For your own decision discipline, write your sell triggers into your purchase notes (decision journal from m7_l4) AT THE TIME OF PURCHASE — sell triggers written when you're calmest about the position are the only ones you can trust to fire correctly under stress.
Among retail investors, the single most common destructive sell pattern is selling at breakeven. The mechanism: you bought at \$50, the stock fell to \$35, you held through the decline (loss-aversion bias from m7_l1), the stock recovered to \$50, you sold to 'get back to even.' From a portfolio-management standpoint, this is one of the worst possible decisions. At the moment you sold, you had access to all the information that justified holding through the decline — you correctly maintained your conviction through a painful drawdown, the thesis re-validated, AND you sold at the exact moment the position's expected forward return was likely improving (because the thesis-validating recovery suggested fundamentals remained intact). The 'breakeven' reference point has no economic content; it's a behavioral relief response. Your purchase price tells you nothing about the stock's future return. The disciplined response: when a position recovers from drawdown, run the 'would-I-buy-it-today' reframe at the recovery price. If the answer is yes (thesis intact, valuation reasonable, conviction unchanged), HOLD or add — never sell to relieve the relief. If the answer is no (thesis weakened, valuation extended, conviction lower), sell — but on the basis of forward expected return, not the breakeven recovery. Three structural defenses. First: never set a sell target at your purchase price; targets should be based on fair value (M6 framework) or thesis triggers, not on the cost basis. Second: when a position recovers from drawdown, EXPLICITLY rerun the buy decision at the recovery price. Third: track 'breakeven sales' as a separate category in your decision journal — most retail investors who track this find that breakeven sales as a category have negative net forward returns (i.e., the positions they sold at breakeven outperformed the market afterward), which makes the pattern visible and easier to break.
All I want to know is where I'm going to die so I'll never go there. Invert, always invert. The best way to make a great sell decision isn't to ask 'how do I sell well?' — it's to ask 'how do I sell terribly?' and then refuse to do those things. Sell because the thesis is broken, sell because the valuation is absurd, sell because you found something genuinely better, sell because the position is too large for your discipline, sell because your circumstances changed. Don't sell because the price went down without your thesis going down with it. Don't sell at breakeven to feel something. Don't sell winners early to 'lock in gains' when the underlying business is still compounding. The whole game is identifying the failure modes and refusing to fail in those particular ways.