Practice ・ 6 min
Selling winners, holding losers: the disposition effect
In 10,000 US brokerage accounts (1987–93), rising stocks were over 1.5 times as likely to be sold as falling ones, yet the winners sold later beat the losers kept.

The evidence behind this article
- Randomised trial1
- Adjusted observational2
Includes studies built to test cause and effect (randomised trials or natural experiments).
3 sources・Countries studied: United States, South Korea, Japan
How to read the kinds of evidence →What to do with this
When you buy a stock, write one line on why and what would make you review it. If unsure about selling, wait a night and ask whether you would buy it today.
When a stock goes up, it is tempting to sell and lock in the gain. When it goes down, it is tempting to hold on until it “comes back”. This pattern is called the disposition effect, and it showed up clearly in the trading records of 10,000 US individual investors. This article covers that study and research from Korea and Japan. It does not recommend any stock or any timing for buying or selling.
Key points
- In trading records from 10,000 accounts at a US discount brokerage (a low-fee broker) for 1987–1993, 14.8% of gains were realized when investors sold, against 9.8% of losses. Stocks that were up were over 1.5 times as likely to be sold. It is an analysis of observational data.
- In the same study, the winners investors sold went on to beat the losers they kept by 3.4 percentage points over the next year. Holding on to losers in the hope that they would recover did not pay off on average.
- Trading data from Korea’s stock index futures market showed the same pattern, which was stronger among individual investors than among institutional or foreign investors. In a Japanese randomised trial in a simulated market, teaching the importance of cutting losses weakened the pattern.
How did 10,000 accounts treat stocks that were up and stocks that were down?
Terrance Odean obtained records from a large US discount brokerage for 10,000 accounts, chosen at random from accounts that traded in 1987. The data cover about 163,000 trades from 1987 to 1993.
On any day an investor sold a stock, every stock in that portfolio was sorted into “up” (a paper gain) or “down” (a paper loss). The study compared the share of gains actually sold with the share of losses sold.
Over the whole year, 14.8% of gains were sold, against 9.8% of losses. The ratio is a little over 1.5, so a stock that was up was more than 50% more likely to be sold than one that was down. The direction was the same when the measures were computed account by account, or in dollars.
December was the exception. In the US, selling a losing stock can lower the tax bill, and in December alone the share of losses sold (12.8%) was higher than the share of gains sold (10.8%). Some investors clearly do trade with taxes in mind.
Did the losers people kept come back?
Holding a loser could be sensible if it was likely to recover. So the study compared what happened next to the winners sold and the losers kept, against the overall market index.
It was the other way round. Over the following year, the winners that were sold did 3.4 percentage points better than the losers that were kept. The pattern remained after accounting for selling to rebalance a portfolio and for the higher trading costs of low-priced stocks. Odean concluded that, for taxable investments, the behaviour is costly and leads to lower after-tax returns.
Other countries show something similar. A study using trading data from Korea’s KOSPI 200 stock index futures found strong evidence of the disposition effect, and individual investors were far more prone to it than institutional and foreign investors. More sophisticated and experienced traders showed less of it, and investors with a stronger disposition effect tended to perform worse.
In Japan, there is a randomised controlled trial in a simulated market. People who were taught that “you should cut losses according to a rule” showed a weaker disposition effect than those who were not.
What does disposition-effect research tell us?
This evidence supports one statement: in US and Korean trading records, the tendency to sell winners first and hold on to losers was clear, and on average it was bad for performance.
- The US and Korean studies are observational. They record what people did, not why.
- Times and rules differ. The US data come from discount brokerage customers in 1987–1993, under the US tax rules of that time. The results may not carry over directly to other tax systems or today’s trading costs.
- The Japanese result is an experiment. It shows that teaching can reduce the pattern, but it was run in a simulated market and did not measure gains or losses in real accounts. We did not find a Japanese study that measures this pattern in real individual accounts.
A small step for today
Whether to sell or hold is your decision. These steps make it harder for feelings to decide for you.
- When you buy a stock, write one line on why you bought it and what would make you review it. It works like the app habit “Tracked spending”: a record you can read later.
- If you are unsure about selling, decide tomorrow instead, in the same spirit as “Waited a night before buying”. Then ask yourself whether you would buy the stock today at this price.
- Use the same note whether you sell a winner or keep a loser, so you judge on today’s reasons rather than the price you paid.
FAQ
What is the disposition effect?
It is the tendency to sell investments that have gone up too soon and to hold investments that have gone down too long. In the study of 10,000 US accounts, 14.8% of gains were sold against 9.8% of losses, so stocks that were up were over 1.5 times as likely to be sold.
If I hold a losing stock, will it come back?
This research cannot say that. In the US study, the winners that were sold beat the losers that were kept by 3.4 percentage points over the next year. That is an average and does not predict how any one stock will move.
Does setting a rule for cutting losses reduce this tendency?
In a Japanese randomised trial in a simulated market, teaching people to cut losses according to a rule weakened the disposition effect. Because it was a simulated market, it has not been shown that this improves results in real accounts.
Sources
- [X-odean-1998] Odean (1998) Are Investors Reluctant to Realize Their Losses? The Journal of Finance.
- [X-choe-eom-2009] Choe & Eom (2009) The disposition effect and investment performance in the futures market. Journal of Futures Markets.
- [X-ando-2021] Ando (2021) Does educational guidance reduce the disposition effect? Evidence from an RCT in a simulated market (in Japanese). Gendai Finance.
Studies describe average tendencies. This article is not a promise that a habit will lift your income, and it is not medical or investment advice. About these articles ・ Search & sources
Updated:
← All articles