The disposition effect

If loss aversion is the engine, the disposition effect is what it does to your account. It has a name, a date, and a body of evidence from real brokerage records, and once you have seen it you will not be able to stop seeing it in your own log.

Naming it

In 1985 Hersh Shefrin and Meir Statman published "The Disposition to Sell Winners Too Early and Ride Losers Too Long." They pulled together prospect theory, mental accounting, regret avoidance and self-control into a single prediction: investors will realise gains readily and hold onto losses, because closing a loss makes it real and closing a gain feels like a small victory. They called this the disposition effect.

The prediction was clean, but at the time it rested on theory and a limited data set. The decisive evidence came thirteen years later.

Odean's brokerage data

In 1998 Terrance Odean published "Are Investors Reluctant to Realize Their Losses?" using the trading records of roughly ten thousand accounts at a large US discount broker from 1987 to 1993. His method was simple and worth understanding, because you can apply it to yourself.

On every day an account sold something, Odean looked at every position in that account. Each position was either a paper gain or a paper loss relative to its purchase price. He then counted how often gains were realised (sold) compared with how often they could have been, and the same for losses. He called these the proportion of gains realised and the proportion of losses realised.

The result: investors realised gains at a substantially higher rate than losses. Given a winner and a loser in the same account on the same day, they were clearly more likely to sell the winner. The only month in which this reversed was December, when tax-loss selling gave people an external reason to close losers.

The part that should get your attention came next. Odean tracked what happened after the sales. The winners investors sold went on to outperform the losers they kept. Holding the losers was not just emotionally costly; it was financially costly, and selling the winners meant giving up returns they would otherwise have had. This was not a tax story and not a rebalancing story. It looked like exactly what Shefrin and Statman predicted.

Key idea: The disposition effect is loss aversion applied to open positions: gains get closed because closing feels good, losses get held because closing makes them real. In Odean's data, this was measurable across thousands of accounts and it cost money.

The same effect in every market

Odean's data was stock investors, but the effect has since been documented in futures traders, in currency markets, in real estate, and among professionals as well as retail. Later work on day traders found the same asymmetry on intraday timescales: a losing position gets held past its stop while a winning one is scalped for a few ticks.

Crypto traders have a particularly vivid version. The word "bagholder" exists because the disposition effect is so common that a community named it.

Prop-firm traders see it as the trade that "just needs to come back" while the daily drawdown limit creeps closer.

Why it feels rational in the moment

The disposition effect survives because each instance comes with a good-sounding story:

• "It's a good company, it'll recover." (But you did not buy it as an investment; you bought it as a trade with a stop.)
• "I'm not selling at the low." (You do not know it is the low. You know it is below your stop.)
• "If I close it, I've locked in the loss." (The loss already exists. Closing changes only whether it is written down.)
• "Take the profit while it's there." (Your plan had a target. The target has not been hit.)

Notice that none of these stories is about the trade's expected value from here. They are all about how the trader will feel after the click. Shefrin and Statman's original paper pointed at exactly this: regret avoidance and the pain of admitting a mistake do the driving, and the market analysis comes after, as decoration.

Measuring your own

The reason Odean's method is worth knowing is that you can run it on your own log in ten minutes.

1. List every closed trade for the last three months.
2. For each, record the planned stop and target in R (Risk Management course, Module 3) and the actual exit in R.
3. Compute the average actual loss and the average actual win.
4. Count how many losers were closed beyond the stop, and how many winners were closed before the target.

A trader with no disposition effect would have an average loss close to −1R and winners distributed around the target. A trader with a strong one has average losses of −1.4R or worse and a pile of +0.3R and +0.5R winners that were supposed to be +2R. That second pattern is extremely common. It is also very fixable, because the fix is mechanical: a stop placed with the entry order, a target placed with it, and a rule that neither is touched.

Try it: Do the four steps above on your last 30 trades. Write the two numbers, average actual loss and average actual win, at the top of your journal. Then compare them to what your plan says they should be. The gap is the disposition effect in your own hand.

Recap

• Shefrin and Statman (1985) predicted that investors sell winners too early and hold losers too long, and named it the disposition effect.
• Odean (1998) confirmed it in thousands of real brokerage accounts: gains were realised at a much higher rate than losses, and the winners sold went on to outperform the losers held.
• The effect appears in every market and timeframe, from long-term stock holdings to intraday futures.
• Each instance comes with a story about the market that is really a story about how the click will feel.
• Measure your own by comparing average actual loss and win to planned stop and target, then remove the decision from real time with bracket orders and a no-touch rule.