You know when to sell a stock when a pre-committed condition fires: a thesis-breaking fundamental change, a valuation extreme relative to your own return target, a position size that exceeds your risk budget, or a liquidity and tax event that makes holding irrational. The trigger should be written before you buy, quantified in dollars or percentages, and tied to evidence you can check. Selling is a decision process, not a feeling, and the process has to survive the moment when the position is emotionally loudest.

The Tax Asymmetry That Governs Every Exit

The reason selling feels harder than buying is structural, not psychological. In the United States, a long-term capital gain held over one year is taxed at preferential rates, while short-term gains are taxed at ordinary income rates. The gap between those two regimes can be 15 to 20 percentage points or more for a high earner. That means the same economic gain is worth materially different after-tax dollars depending on the calendar.

This asymmetry creates a mechanical bias toward holding. A position that has appreciated 40% and is 11 months old carries a hidden liability: selling now converts a lower-taxed gain into a higher-taxed one. The rational comparison is not "do I still like the stock" but "does the expected after-tax return of holding exceed the after-tax return of the best alternative, including the tax cost of switching." Most investors never compute that comparison, which is why they default to inertia.

There is a second layer. Unrealized gains are not a loan from the government you can defer forever. They represent future tax that will eventually be paid, either by you or by your heirs. For taxable accounts above the estate-tax exemption threshold, the step-up in basis at death means the embedded gain may never be taxed at all. That is a legitimate reason an elderly holder might rationally never sell a concentrated position, and it is the single most common professional caveat in wealth management that retail content ignores entirely.

The practical translation: know your holding period clock and your marginal rate before you decide anything. A sale decision made on a Wednesday at month 11 is a different financial event than the same decision at month 13, and the difference is often larger than the return you expect to earn by switching.

How the Thesis-Break Mechanism Actually Works

Every stock purchase is an implicit bet on a small number of causal claims. The company will grow revenue because a specific product is adopted. Margins will hold because a cost advantage is durable. The balance sheet can fund the plan without dilution. When you sell, you are typically not selling because the price moved. You are selling because one of those causal claims failed.

The mechanism matters because price declines are ambiguous. A 30% drawdown can mean the market has repriced a still-intact thesis, or it can mean the thesis was wrong and the market figured out first. The way to distinguish them is to pre-specify the observable that would falsify each claim before you invest. If the claim is "adoption is accelerating," the falsifier is quarterly unit growth turning negative for two consecutive periods. If the claim is "margin is defensible," the falsifier is gross margin compression beyond a stated threshold for a stated number of quarters.

Professionals use this discipline because it converts a diffuse anxiety into a binary check. Write three to five falsifiers per position at entry. When one fires, you have an answer that does not depend on how you feel that morning. When none has fired, a drawdown is noise you can choose to hold through, or a position-sizing question rather than a thesis question.

The counterintuitive detail: most thesis breaks are visible in the numbers before they are visible in the price. Earnings revisions and guidance cuts lead price declines more often than they follow them. The sell decision that protects capital is usually made on the filing, not the chart.

Valuation, Position Size, and the Two Independent Triggers

There are two different reasons to sell that get conflated constantly. The first is that the business no longer deserves your capital. The second is that your portfolio can no longer tolerate the concentration. These are independent, and either alone is sufficient.

Valuation-based exit: if you bought at 15 times earnings expecting a 10% annual return, and the multiple is now 35 times earnings with no change in the underlying growth rate, the forward return implied by that price is below your target. The stock did not become a worse company. It became a worse investment at that price. The trigger is quantitative: a stated multiple, a stated implied return, a stated margin of safety you require to keep holding.

Position-size-based exit: a reasonable concentration limit for most individual investors is a single position at 5% to 10% of the portfolio, and a sector at 20% to 25%. When a winner grows past those bands, the portfolio's risk is now dominated by one outcome you cannot control. Rebalancing trims the position back to target, and it is a sell even when the thesis is intact. The rule is mechanical: at a stated weight, trim to a stated weight, on a stated schedule.

The reconciliation between these thresholds is where most frameworks fail. Valuation says hold because the implied return is still above your floor. Concentration says trim because the weight breached the band. The resolution is not to pick one. It is to trim to the band and hold the remainder, which satisfies both constraints. State this in advance so the argument with yourself does not happen at the moment of action.

Exit Triggers Compared

Trigger type What fires it Typical threshold What it protects Common failure mode
Thesis break A pre-written falsifier is observed Company-specific, often 2 consecutive quarters Capital from permanent loss Falsifier written too vaguely to observe
Valuation extreme Implied forward return falls below floor Stated multiple or expected return Future returns Multiple shifts on tax or rate regime, not fundamentals
Position size Weight exceeds band 5% to 10% single, 20% to 25% sector Portfolio risk Selling winners only, never trimming losers
Tax event Holding period or loss harvesting One year for long-term rates After-tax return Ignoring embedded gain entirely
Liquidity need Cash required within stated horizon 2 to 3 years of expenses Forced sale at bad prices Selling the liquid asset, holding the illiquid one
Risk tolerance shift Horizon or obligations change Life-event defined Plan alignment Pretending nothing changed

The Time Course of a Position

Phase Typical duration What is happening What the exit rule should do
Entry and thesis formation 0 to 1 month Falsifiers written, size set Record triggers in writing
Early holding 1 to 12 months Short-term tax regime applies Avoid switching for noise
Long-term regime begins After 12 months Preferential rates apply Re-evaluate tax cost of any sale
Rebalancing Ongoing Weight drifts with returns Trim at band, no discretion
Review cycle Quarterly to annually Falsifiers checked against filings Act on fired triggers only
Deterioration Company-specific Estimates cut before price falls Sell on evidence, not on drawdown
Forced liquidity Event-driven Cash needed on a date Match asset duration to horizon

Red Flags That Force a Re-Review

Accounting restatement, auditor resignation, or a material weakness disclosure. Insider selling that is broad, sustained, and not explained by diversification. Guidance withdrawn or repeatedly cut within a short window. Debt covenants approaching breach, or refinancing at materially higher rates. A key customer or product line loss that the thesis assumed would persist. Any position weight that crossed your band while you were not looking.

None of these is an automatic sell. Each is an automatic re-review, meaning the falsifier list gets checked immediately rather than at the next scheduled review.

Stop-Loss Orders and Mental Accounting

The most persistent misconception is that a stop-loss order is a sell discipline. For a concentrated, fundamentally driven position, a fixed percentage stop is a mechanism that guarantees you sell low after a thesis-intact drawdown and does not protect you from the slow fundamental decay that actually destroys capital. Volatility is not the same as risk, and a stop converts volatility into a realized loss.

A subtler error is mental accounting around the purchase price. The price you paid is irrelevant to whether the stock is a good holding today, except for its tax consequences. Investors routinely refuse to sell a loser because selling "locks in" the loss, while the market has already marked the position down. The loss is real whether or not you transact. What transacting changes is the tax treatment and the opportunity cost.

Practical Actions With Concrete Thresholds

Write three to five falsifiers per position at entry, each observable in a filing or earnings release. Set a single-position cap and a sector cap, and commit to trimming to the cap when breached. Compute your holding-period clock and marginal capital gains rate before any sale. Schedule a quarterly review and check falsifiers against the actual reported numbers. Keep two to three years of expenses in liquid assets so market timing never becomes forced. For any position above 10% of the portfolio, run an explicit after-tax comparison of holding versus switching.

Limitations and What This Framework Cannot Do

This framework cannot tell you the future price, and it cannot eliminate the possibility that a thesis-intact position falls further. It also cannot substitute for tax advice specific to your bracket and state, because the capital gains treatment described here is a general US federal mechanism and state rules, thresholds, and estate considerations vary. The thresholds above are starting bands, not universal constants, and they should be adjusted to your horizon, obligations, and risk capacity. Finally, the framework assumes you can observe falsifiers honestly; if you cannot, the discipline collapses regardless of how well the rules are written.

This article is general educational information about decision processes, not a diagnosis, legal opinion, or investment recommendation. It does not promise any outcome or return, and it cannot account for your personal tax position, obligations, or risk tolerance. Consult a qualified financial, tax, or legal professional in your jurisdiction before acting on any sale decision.