Disposition effect
Selling winners too soon and holding losers too long.
What it means
The disposition effect is the tendency to sell assets that have risen in value while clinging to those that have fallen — selling winners too early and riding losers too long. The phenomenon itself is well documented across individual investors and even some professionals. It is most often attributed to prospect theory — realizing a gain locks in a sure win in the domain of gains where people are risk-averse, whereas realizing a loss forces acknowledgment of a loss, and loss aversion plus risk-seeking in the loss domain make holding-and-hoping feel preferable to crystallizing the pain — but this account is contested: formal models show prospect-theory preferences often fail to predict the effect, and realization utility (pleasure booked at the moment of sale), mental accounting, and belief-based speculation are also implicated.
What the evidence shows
The effect was named by Shefrin and Statman in 1985, but the decisive evidence is Odean's 1998 study of roughly 10,000 discount-brokerage accounts. He compared the proportion of gains realized against the proportion of losses realized and found investors were about one and a half times more likely to sell a winner than a loser, a gap that survived controls for taxes, rebalancing, and transaction costs. Laboratory markets reproduce it: Weber and Camerer had subjects trade six risky assets and watched them sell winners and hold losers against Bayesian logic, with the tendency largely vanishing when shares were sold automatically each period. The pattern recurs across countries, asset classes, and, more weakly, some professionals.
Why the standard story is contested
The prospect-theory account is intuitive but formally shaky. Barberis and Xiong showed that when gains and losses are defined over an annual horizon, prospect-theory preferences often fail to predict a disposition effect at all; only a model of realization utility, where value is booked at the moment of sale, reliably generates it. Ben-David and Hirshleifer went further, mapping selling probability against profit and finding it V-shaped rather than a clean jump at zero: investors sell big losers readily, not just big winners, and there is little upward step exactly at breakeven. That pattern points less to pure loss aversion than to speculation and belief revision, meaning the effect may be a symptom of several mechanisms rather than one.
Realization utility and the brain
The realization-utility idea, that people derive a burst of pleasure or pain from the act of closing a position rather than only from final wealth, has neural support. Frydman, Barberis, Camerer, Bossaerts, and Rangel scanned traders in an experimental market and found activity in reward-related regions tracked the realized gain at the moment a stock was sold, matching the model's predictions, while subjects showed the usual disposition effect. This reframes the behavior: selling a winner delivers an immediate hedonic payout, and selling a loser inflicts an immediate hit, so holding a loser is a way of deferring pain. It also explains why automatic or rule-based selling, which removes the discrete moment of realization, blunts the bias.
Who escapes it, and how
The effect is not uniform. Feng and Seasholes, studying Chinese brokerage accounts, found that investor sophistication and accumulated trading experience together roughly eliminate the reluctance to realize losses, with individuals showing progressively less disposition as they trade more. This matters in practice: the countermeasures that work are structural rather than motivational. Pre-committed stop-loss rules, scheduled rebalancing, and tax-loss-harvesting routines strip out the discretionary moment where the bias operates. Framing helps too, since tracking a whole portfolio rather than each position's purchase price removes the account-level 'in the red' that mental accounting fixates on. Advice tools and robo-advisors lean on exactly these mechanics, automating the sells that a human, left to feel each realization, tends to avoid.
Examples
A trader cashes out a stock up 10% but clings to one down 30%, hoping it 'comes back.'
In a falling market, homeowners list at the price they paid and refuse to move, waiting years for a sale rather than booking a loss on the account they opened at purchase.
A collector sells the print that doubled within a month but keeps the one that halved for a decade, because selling would turn a paper loss into a real one.
A crypto holder sells a token up 25% within a week yet keeps one down 60% for years, since selling would turn an on-screen paper loss into a booked, tax-reported one.
An in-play bettor cashes out the moment a wager edges into profit but lets a losing bet ride to the final whistle, hoping the scoreline turns rather than settling for the loss.
First described in Shefrin & Statman (1985).
Key references
- Frydman, C., Barberis, N., Camerer, C., Bossaerts, P., & Rangel, A. (2014). Using neural data to test a theory of investor behavior: An application to realization utility. The Journal of Finance, 69(2), 907-946. doi.org/10.1111/jofi.12126
- Ben-David, I., & Hirshleifer, D. (2012). Are investors really reluctant to realize their losses? Trading responses to past returns and the disposition effect. The Review of Financial Studies, 25(8), 2485-2532. doi.org/10.1093/rfs/hhs077
- Feng, L., & Seasholes, M. S. (2005). Do investor sophistication and trading experience eliminate behavioral biases in financial markets? Review of Finance, 9(3), 305-351. doi.org/10.1007/s10679-005-2262-0
- Odean, T. (1998). Are investors reluctant to realize their losses? The Journal of Finance, 53(5), 1775-1798. doi.org/10.1111/0022-1082.00072
- Weber, M., & Camerer, C. F. (1998). The disposition effect in securities trading: An experimental analysis. Journal of Economic Behavior & Organization, 33(2), 167-184. doi.org/10.1016/S0167-2681(97)00089-9
- Shefrin, H., & Statman, M. (1985). The disposition to sell winners too early and ride losers too long: Theory and evidence. The Journal of Finance, 40(3), 777-790. doi.org/10.1111/j.1540-6261.1985.tb05002.x