The Disposition Effect: Why Investors Sell Winners Too Early

The disposition effect is one of the most robust and costly behavioral biases in finance: investors systematically sell their winning stocks too early and hold their losing stocks too long. The pattern has been documented in tens of thousands of trading accounts across multiple countries and decades — and it actively destroys returns.

1. Defining the Disposition Effect

The disposition effect was named and formally described by Hersh Shefrin and Meir Statman in their 1985 paper “The Disposition to Sell Winners Too Early and Ride Losers Too Long: Theory and Evidence,” published in the Journal of Finance, Vol. 40, No. 3, pp. 777–790.

The disposition effect describes a two-part behavioral pattern observed in investors:

• Selling winners too early: When a stock rises above the purchase price, the investor is disproportionately likely to sell it, locking in a gain that may have grown larger had they continued to hold.
• Holding losers too long: When a stock falls below the purchase price, the investor is disproportionately likely to continue holding it, hoping it will recover back to the original purchase price, even when the fundamental outlook has deteriorated.

Shefrin and Statman grounded their theory in Kahneman and Tversky’s Prospect Theory (1979), which established that people evaluate outcomes relative to a reference point and that losses loom larger than gains. The purchase price serves as the natural reference point for investors, and the asymmetric treatment of gains and losses predicted by Prospect Theory generates exactly the pattern Shefrin and Statman observed: eagerness to realize gains and reluctance to realize losses.

2. Odean’s Landmark Empirical Evidence

The most influential empirical study of the disposition effect was conducted by Terrance Odean and published in 1998 as “Are Investors Reluctant to Realize Their Losses?” in the Journal of Finance, Vol. 53, No. 5, pp. 1775–1798.

Odean obtained trading records for 10,000 accounts at a large discount brokerage covering the period from 1987 to 1993. This was not a survey or a laboratory experiment — it was real money, real trades, real consequences. He analyzed every sale decision and compared it to the opportunities the investor had to sell other stocks in their portfolio at the same time.

His methodology was elegant. For each sale, he calculated the Proportion of Gains Realized (PGR) and the Proportion of Losses Realized (PLR). PGR measures the fraction of stocks trading at a gain that were actually sold, while PLR measures the fraction of stocks trading at a loss that were actually sold. If investors had no disposition effect, PGR and PLR would be roughly equal.

The results were stark: investors were approximately 1.5 times more likely to sell a stock that was trading at a gain than one trading at a loss. The PGR was 0.148 while the PLR was 0.098 — a highly statistically significant difference. Investors were systematically, reliably choosing to sell their winners and hold their losers.

The Winners Sold Outperformed the Losers Held

Perhaps the most damaging finding in Odean’s study was what happened after the sell decisions. He tracked the subsequent performance of both the winners that were sold and the losers that were held. On average, the winners that investors sold went on to outperform the losers they continued to hold over the following 12 months. The disposition effect was not merely a neutral psychological quirk — it was actively destroying returns. Investors were selling their best stocks and keeping their worst ones.

This finding is consistent with the existence of momentum in stock returns. Stocks that have performed well in the recent past tend to continue performing well, and stocks that have performed poorly tend to continue performing poorly. By selling winners and holding losers, disposition-effect-afflicted investors are fighting against one of the most robust anomalies in financial markets.

3. The Tax Irrationality of the Disposition Effect

The disposition effect is not merely psychologically irrational — it is tax-irrational. In most tax jurisdictions, including the United States, the tax-optimal strategy is the exact opposite of what disposition-effect investors do:

• Hold winners longer: Long-term capital gains (assets held more than one year in the U.S.) are taxed at preferential rates — 0%, 15%, or 20% depending on income, compared to ordinary income tax rates of up to 37% for short-term gains. By selling winners early, investors often convert what could be a long-term gain into a short-term gain, increasing their tax bill.
• Sell losers sooner: Realized capital losses can offset capital gains, reducing tax liability. If losses exceed gains, up to $3,000 per year can be deducted against ordinary income in the U.S. (with the remainder carried forward). By holding losers, investors forgo these tax benefits. This is why tax-loss harvesting — the deliberate realization of losses for tax purposes — is the rational strategy.

The tax-optimal behavior is precisely the reverse of the disposition effect: let winners run (to qualify for long-term capital gains rates) and sell losers (to harvest the tax deduction). Odean estimated that the tax cost of the disposition effect was substantial, with investors paying significantly more in taxes than they would have if they had followed a tax-efficient strategy.

4. Psychological Causes of the Disposition Effect

Shefrin and Statman (1985) identified several psychological mechanisms that contribute to the disposition effect. Subsequent research has refined and expanded this list:

Loss Aversion (Kahneman & Tversky, 1979)

The primary driver is loss aversion, the central finding of Prospect Theory. Losses feel approximately 2.0 to 2.5 times as painful as equivalent gains feel pleasurable. Selling a losing stock means converting an unrealized (“paper”) loss into a realized loss — making it psychologically real and permanent. The pain of this realization is so intense that investors delay it indefinitely, even when holding the losing position is economically irrational.

Conversely, selling a winning stock converts a paper gain into a realized gain — a pleasurable experience. The concave value function for gains (diminishing sensitivity) means that the incremental pleasure of watching a gain grow from, say, $1,000 to $2,000 is smaller than the initial pleasure of the first $1,000. Combined with the fear of watching the gain disappear entirely (loss aversion applied to unrealized gains), this creates a strong impulse to sell early and lock in the profit.

Mental Accounting (Thaler, 1985)

Richard Thaler’s concept of mental accounting, described in his 1985 paper “Mental Accounting and Consumer Choice,” published in Marketing Science, Vol. 4, No. 3, pp. 199–214, explains how people create separate mental accounts for each investment. Rather than evaluating their portfolio as a whole, investors track each stock as an individual “account” and evaluate it against its own purchase price. This creates the reference points that drive the disposition effect.

Mental accounting means that the decision to sell Stock A is evaluated in the context of Stock A’s gain or loss, not in the context of the overall portfolio. An investor might hold a losing stock in one mental account while simultaneously failing to recognize that the capital could be redeployed more productively elsewhere. Each position is its own psychological world.

Regret Avoidance

Regret is a powerful driver of the disposition effect in both directions. Selling a losing stock and watching it subsequently recover would produce intense regret. Holding the loser avoids the possibility of this regret. On the flip side, selling a winner that subsequently declines confirms the sell decision and avoids the regret of watching profits evaporate. The asymmetry of potential regrets reinforces the disposition pattern.

Mean Reversion Belief

Many investors hold an intuitive belief that stock prices mean-revert — that stocks which have gone up are “due” to come back down, and stocks which have gone down are “due” to recover. This belief has some basis in very long-horizon returns (where value effects operate), but is wrong at the intermediate horizons relevant to most trading decisions. At 3–12 month horizons, momentum is the dominant pattern: winners tend to keep winning and losers tend to keep losing. The false belief in mean reversion reinforces the disposition effect by providing a seemingly rational justification for selling winners and holding losers.

5. Institutional vs. Retail Investors

Is the disposition effect limited to unsophisticated retail investors, or does it also affect professionals? The evidence suggests that professional investors are less susceptible to the disposition effect, but not immune.

Peter Locke and Steven Mann investigated this question in their 2005 paper “Professional Trader Discipline and Trade Disposition,” published in the Journal of Financial Economics, Vol. 76, No. 2, pp. 401–444. They studied professional futures traders on the Chicago Mercantile Exchange (CME) and found that while the disposition effect was less pronounced among professionals than among retail traders, it was still present. Even traders whose livelihoods depended on rational decision-making exhibited some tendency to realize gains faster than losses.

Importantly, Locke and Mann found variation among professionals: the most successful traders exhibited the least disposition effect. Discipline in cutting losses was associated with better performance, consistent with the interpretation that the disposition effect is costly and that overcoming it provides an edge.

Other studies of institutional investors have found similar patterns. Mutual fund managers, hedge fund managers, and pension fund managers all show some residual disposition effect, though typically weaker than what is observed in retail accounts. The institutional setting provides some natural defenses — risk management systems, stop-loss protocols, and oversight from colleagues — but the underlying psychological bias persists.

6. The Global Evidence

The disposition effect is not an artifact of American markets or American psychology. It has been documented in virtually every market that has been studied:

• Finland: Grinblatt and Keloharju (2001) studied Finnish stock market data and found a strong disposition effect, published in the Journal of Finance.
• China: Chen, Kim, Nofsinger, and Rui (2007) found a strong disposition effect among Chinese investors.
• Israel: Shapira and Venezia (2001) found the effect in Israeli brokerage accounts.
• Australia: Brown, Chappel, da Silva Rosa, and Walter (2006) documented it in Australian markets.
• Taiwan: Barber, Lee, Liu, and Odean (2007) found a particularly strong disposition effect in Taiwanese stock data.

The universality of the effect reinforces the interpretation that it is rooted in fundamental features of human psychology (loss aversion, reference dependence) rather than in culture-specific factors or institutional details of particular markets.

7. How the Disposition Effect Creates Momentum

The disposition effect has implications that extend far beyond individual portfolio performance. At the aggregate level, when millions of investors simultaneously exhibit the disposition effect, it creates measurable patterns in asset prices — most notably, it contributes to price momentum.

The mechanism is straightforward: when a stock receives positive news and begins to rise, disposition-effect investors sell prematurely, suppressing the stock’s price below its fundamental value. The stock rises, but not as fast as it should, because there is a steady stream of selling from investors who are eager to lock in gains. Over time, as these sellers are exhausted and new information continues to be incorporated, the price gradually catches up to fundamentals. This gradual catch-up is momentum.

The reverse operates for negative news. When a stock receives bad news and begins to fall, disposition-effect investors refuse to sell, holding onto their losing positions and delaying the price decline. The stock falls, but not as fast as it should, because there is less selling pressure than fundamentals would warrant. Over time, reality catches up, and the stock continues to decline.

Harrison Hong and Jeremy Stein formalized this insight in their 1999 paper “A Unified Theory of Underreaction, Momentum Trading, and Overreaction in Asset Markets,” published in the Journal of Finance, Vol. 54, No. 6, pp. 2143–2184. Their model showed how gradual information diffusion and underreaction to news (which the disposition effect amplifies) can generate the short-term momentum and long-term reversal patterns observed in stock returns.

8. Measuring the Disposition Effect in Your Own Trading

You can measure your own disposition effect using the same methodology Odean used. For every sale you make, ask: was this stock trading at a gain or a loss relative to my purchase price? Then look at all the stocks you held but did not sell at the same time: what fraction of them were at a gain, and what fraction were at a loss?

Calculate your personal PGR and PLR:

PGR = (Number of gains realized) / (Number of gains realized + Number of paper gains not sold)

PLR = (Number of losses realized) / (Number of losses realized + Number of paper losses not sold)

Disposition Effect = PGR - PLR

If PGR > PLR, you exhibit the disposition effect. If PLR > PGR, you are doing the opposite — holding winners and cutting losers, which is generally the more profitable pattern. Most investors, if they are honest with their records, will find that PGR significantly exceeds PLR.

9. The Subsequent Performance Test

Beyond simply measuring whether you sell winners more readily than losers, you should track what happens after your sell decisions. For every stock you sell at a gain, record its performance over the subsequent 3, 6, and 12 months. Do the same for the losers you continue to hold. If the sold winners go on to outperform the held losers — as Odean found in his aggregate data — then the disposition effect is actively costing you money.

This exercise is psychologically difficult. Nobody wants to confront evidence that their decisions are systematically wrong. But it is essential. Without objective measurement, the disposition effect feels rational in the moment (“I’m prudently taking profits” / “It will come back”), and only reveals itself as destructive when viewed in aggregate across many trades.

10. Strategies to Overcome the Disposition Effect

Trailing Stop-Losses

A trailing stop-loss moves upward with the stock price but does not move downward. It allows winners to continue running while automatically cutting losers. For example, a 15% trailing stop on a stock purchased at $100 would trigger a sale if the stock fell to $85. If the stock first rose to $150, the trailing stop would move to $127.50 (15% below $150). This mechanically reverses the disposition effect: it lets winners run and cuts losers automatically.

Systematic Entry and Exit Rules

The most effective cure for the disposition effect is to remove human discretion from sell decisions entirely. A systematic strategy with predefined exit criteria — based on technicals, fundamentals, time limits, or any other objective metric — does not experience loss aversion. It does not care whether a position is above or below the purchase price. It follows rules, not emotions.

Tax-Loss Harvesting

Implementing a deliberate tax-loss harvesting discipline directly counteracts the disposition effect. At regular intervals (quarterly, or whenever a position is down by a predetermined threshold), systematically review positions trading at a loss and sell those where the tax benefit of realizing the loss exceeds the cost of the transaction and any wash-sale considerations. This converts the disposition effect’s most tax-inefficient behavior — holding losers — into its tax-optimal opposite.

The “Would I Buy It Today?” Test

For each position in your portfolio, ask: “If I did not already own this stock and had the cash instead, would I buy it today at the current price?” If the answer is no, the only reason you are holding it is the endowment effect and loss aversion — the fact that selling would force you to confront a loss. This mental exercise can help break through the psychological barriers that keep you holding losers.

Premortem Analysis

Before opening any position, write down the specific conditions under which you will sell — both for a profit (take-profit level) and for a loss (stop-loss level). Include a time stop: how long will you hold this position if it goes nowhere? By making these commitments when you are emotionally neutral (before the position is established and you have a reference point), you can counteract the disposition effect that will inevitably arise once you are invested.

11. The Disposition Effect and Market Microstructure

Research has shown that the disposition effect influences not just which stocks are sold, but how they are sold. Disposition-motivated selling tends to be concentrated at specific price levels — particularly at and just above the purchase price. This creates clustering of sell orders near cost-basis reference points, which can be identified in order flow data.

For quantitative traders and market makers, this pattern is informative. Stocks that have recently risen to price levels where many investors purchased shares (identifiable through volume analysis) will face disposition-motivated selling pressure at those levels. This selling pressure is not informationally motivated — it reveals nothing about the stock’s fundamental value — and therefore represents an opportunity for informed traders to absorb the supply at favorable prices.

12. Summary: Fighting Your Own Worst Instincts

The disposition effect, first named by Shefrin and Statman in 1985 and empirically demonstrated at scale by Odean in 1998, is one of the most robust and costly behavioral biases in finance. Investors are approximately 1.5 times more likely to sell a winning stock than a losing one. The winners they sell go on to outperform the losers they hold. The behavior is tax-irrational. It has been documented across retail and institutional investors, across dozens of countries, and across decades of data.

The causes are deeply rooted in human psychology: loss aversion makes realized losses intensely painful, mental accounting creates reference points tied to purchase prices, regret avoidance discourages actions whose outcomes might produce regret, and false beliefs about mean reversion provide a pseudo-rational justification for holding losers.

At the market level, the disposition effect contributes to price momentum by causing underreaction to both positive and negative information. Widespread early selling of winners suppresses their prices; widespread reluctance to sell losers props up their prices. Both effects create momentum as prices gradually adjust.

The most effective countermeasures are structural: trailing stop-losses, systematic trading rules, tax-loss harvesting disciplines, and pre-commitment to exit criteria before positions are opened. Awareness alone is not enough — you need systems that enforce discipline when your emotions are screaming at you to do the wrong thing.