Disposition Effect: Why Traders Sell Winners Early and Sit on Losers

Open any retail demat account that has been running for a few years and you will usually see the same thing: a handful of stocks bought at much higher prices, still sitting there, and a trade history full of small, quick profits. That pattern has a name. The disposition effect is the habit of selling winning positions too early and holding losing positions too long, and it is one of the most studied behaviour patterns in finance.

We are going to look at where the idea came from, what the best-known study actually measured, why our brains push us this way, and what a trader can do about it with rules rather than willpower.

What the disposition effect means

The term was coined by Hersh Shefrin and Meir Statman in a 1985 paper in The Journal of Finance, titled “The Disposition to Sell Winners Too Early and Ride Losers Too Long”. The title is the definition. When a stock is above your buy price, you feel a pull to book the profit. When it is below, you feel a pull to wait until it “comes back”.

Notice what is driving both decisions: your purchase price. The market has no idea what you paid. The stock’s future depends on earnings, flows, sector news and a hundred other things, none of which care about your entry. Yet the entry price ends up deciding whether you sell. That is the core of the problem.

According to the Wikipedia summary of the research, Shefrin and Statman argued that four things combine to produce the effect:

• Loss aversion, from prospect theory: losses hurt more than equal gains feel good, and people avoid locking in a sure loss.
• Mental accounting: we track each stock as a separate account measured against its purchase price, instead of looking at the portfolio as a whole.
• Regret avoidance: selling a loser means admitting a mistake, and the regret only becomes real when the position is closed. Selling a winner gives a hit of pride.
• Self-control: many traders know they should cut a loser and still cannot make themselves do it.

Their point was that no single one of these explains the full behaviour. It is the mix.

What Odean’s study of 10,000 accounts found

The paper most people cite is Terrance Odean’s 1998 study, “Are Investors Reluctant to Realize Their Losses?“, also in The Journal of Finance. He used the trading records of 10,000 randomly selected accounts at a large US discount brokerage, covering January 1987 to December 1993.

His method is simple enough to use on your own trades, which is why it is worth understanding. On every day an investor sold something, he looked at each stock in the portfolio and sorted it into one of four boxes:

• Realised gain: sold above the purchase price.
• Paper gain: held, but currently above the purchase price.
• Realised loss: sold below the purchase price.
• Paper loss: held, but currently below the purchase price.

From these he built two ratios:

• Proportion of Gains Realised (PGR) = realised gains ÷ (realised gains + paper gains)
• Proportion of Losses Realised (PLR) = realised losses ÷ (realised losses + paper losses)

If investors treated winners and losers the same way, the two numbers would be about equal. They were not. Across the whole year, Odean’s Table I shows a PGR of 0.148 and a PLR of 0.098. In plain terms, on a day when an investor sold something, 14.8% of the winning positions available to sell were sold, against 9.8% of the losing ones. Odean notes that the ratio of PGR to PLR is a little over 1.5, meaning a stock that was up was more than 50% more likely to be sold on a given day than a stock that was down.

The market has no idea what you paid for a stock. When the purchase price decides whether you sell, the trade is being run by your memory, not by the chart.

December was the exception. In December alone, the same table shows PGR falling to 0.108 and PLR rising to 0.128, so losses were realised faster than gains. Odean links this to US tax-loss selling at the calendar year-end. For January to November the gap was even wider than the full-year figure: PGR of 0.152 against PLR of 0.094. So investors could sell losers when there was a concrete reason, a tax benefit, to do so. Most of the year, they did not.

Did holding the losers pay off?

One defence of the behaviour is that people hold losers because they expect a rebound, and sell winners because they expect a pullback. Odean tested that. Measured against a market index, the winners investors sold went on to earn an average excess return over the following year that was 3.4 percentage points higher than the losers they kept. The expectation of mean reversion was, on average, wrong in his data.

A worked example: measuring your own disposition effect

You do not need a research database to check yourself. A trading journal or your broker’s tax P&L report is enough. Here is a hypothetical example with made-up numbers to show the arithmetic.

Say over three months you had 10 days on which you sold at least one stock. On each of those days you note how many positions were in profit and how many in loss, and which ones you sold. Adding it all up:

Your PGR is 12 ÷ (12 + 28) = 12 ÷ 40 = 0.30. Your PLR is 4 ÷ (4 + 36) = 4 ÷ 40 = 0.10. The ratio is 0.30 ÷ 0.10 = 3. On a selling day, you were three times as likely to sell a winner as a loser. That is a strong tilt, and it tells you exactly where to look.

Now take a single position to see the cost in rupees. Again, hypothetical numbers.

• You buy 100 shares of Stock A at ₹500 and 100 shares of Stock B at ₹800. Total outlay: ₹50,000 + ₹80,000 = ₹1,30,000.
• A month later, A is at ₹560 (up ₹60 a share, a gain of ₹6,000) and B is at ₹720 (down ₹80 a share, a loss of ₹8,000).
• You need ₹50,000 for something else. The disposition-effect move is to sell A, book ₹6,000, and keep B “until it recovers”.

Look at what you now own. You have exited the stock the market was rewarding and kept ₹72,000 in the stock the market was marking down. Your portfolio is now more concentrated in your worst idea. If B slides another 10% to ₹648, that is a further ₹7,200 gone, larger than the whole profit you booked on A. None of this means B must fall or A must rise. It means the decision was made on the wrong question. The right question is: if I had ₹1,30,000 in cash today, would I buy these two positions at today’s prices, in these sizes?

Why it matters more for traders than for long-term investors

A long-term investor holding a business through a bad patch can have good reasons to sit tight, provided the reasons are about the business and not about the purchase price. A short-term trader has no such cover. A swing or positional trade is built on a setup, and once price breaks the level that defined the setup, the reason for the trade is gone.

This is where the disposition effect quietly wrecks a trading system. Most trading edges come from cutting losers small and letting some winners run. That is what makes a strategy with a 40% hit rate profitable. Flip the behaviour, and the same entries give you small wins and large losses. You can work through the maths in our guide to trading expectancy: an average loss that grows even modestly bigger than the average win can turn a positive expectancy negative.

The research also suggests professionals are not immune. The Wikipedia summary notes Andrea Frazzini’s 2006 study, which found mutual fund managers also showed the disposition effect, though to a lesser degree than individual investors. Experience helps. It does not switch the bias off.

A 40% hit rate can make money only if the losers stay small. The disposition effect does the opposite: it keeps the wins small and lets the losses grow.

The link with momentum

There is a market-wide angle too. Mark Grinblatt and Bing Han (2004), as summarised on Wikipedia, argued that the disposition effect can help explain short-term price momentum. If lots of holders rush to sell a stock after good news, that selling slows the price’s move up to its new fair value. If holders refuse to sell after bad news, the lack of selling slows the move down. Either way, prices adjust gradually, which shows up as momentum. For a trader, the practical takeaway is uncomfortable: the stocks you are most tempted to sell are, on average, not the ones you should be selling.

How to fight the disposition effect

Telling yourself to be rational does not work for long. What works is taking the decision away from the moment when emotions are high. Here are approaches we find useful, several of which come straight from the research.

1. Put the stop-loss in before you enter

Shefrin and Statman themselves described stop-loss orders as a self-nudge against the effect. Decide the exit level while you are still neutral, and place it with the broker. The research on the disposition effect is a good reason to treat a stop as a commitment rather than a suggestion. Our risk management in trading guide covers how to set stops and size positions around them.

2. Lean on automation

In Weber and Camerer’s 1995 experiment, again as summarised on Wikipedia, when shares were automatically sold after each period, the disposition effect was greatly reduced. You can borrow the idea: GTT orders, bracket orders and pre-set trailing stops all remove the moment of choice.

3. Use exits that don’t care about your entry

A trailing stop based on a moving average, the previous swing low, or a multiple of ATR (average true range) is anchored to price action, not to what you paid. The same goes for time stops: if a trade has not worked in a set number of sessions, it is closed. Both shift attention away from “am I in profit?” to “is the setup still valid?”.

4. Ask the fresh-cash question every week

Go through your holdings and ask, for each one: would I buy this today at this price and this size? If the answer is no, the only reason to hold it is your entry price. That is the disposition effect talking.

5. Scale out of winners, don’t dump them

If you feel the urge to book profit, sell a part and trail the rest. This scratches the itch for a realised gain while still leaving room for a winner to run, which is the logic behind anti-martingale position sizing as well.

6. Reframe the portfolio, not the position

The Wikipedia article mentions research showing investors behave differently when they think of their money as progress towards a specific goal rather than a set of separate bets. Looking at the total portfolio value, rather than a red or green number next to each stock, weakens the mental-accounting part of the bias.

7. Track your own PGR and PLR

Run the calculation from the worked example every quarter. If your ratio is well above 1, you have hard evidence of the habit. If it starts to come down, your rules are working.

Common mistakes when trying to fix it

• Overcorrecting into panic selling. The fix is not to dump every red position at the first dip. It is to exit when the reason for the trade is gone.
• Averaging down to “fix” a loser. Buying more of a falling stock to lower the average price is often the disposition effect in disguise: the goal becomes getting back to break-even rather than owning the best position.
• Moving stops further away. Shifting a stop lower once price gets close to it is the same bias with an extra step.
• Judging a sale by what happened next. A cut loser that later bounces doesn’t make the cut wrong. Judge exits by whether they followed your plan.

The bottom line

The disposition effect is simple to describe and hard to beat: we sell what is working because booking a profit feels good, and we keep what is failing because closing it feels like admitting a mistake. Odean’s data showed investors selling winners at roughly one and a half times the rate of losers, and the winners they sold went on to do better than the losers they kept. You won’t out-think this bias in the moment. Set exits before entry, automate what you can, and check your own selling numbers every few months.

Disclaimer: This article is for education only and is not investment advice. Trading in stocks involves risk, including loss of capital. The worked examples use hypothetical figures. Consult a SEBI-registered adviser before making investment decisions.