Skip to main content
The Investor’s MindLesson 1 of 6 · The classic traps

Judgment & behaviour

The disposition effect — selling winners, keeping losers

Individual investors sell their gains and hold their losses, at a rate too large to be chance and in the direction that costs them money. It is the best-measured mistake in retail investing.

7 min readBeginnerLast checked against the product on

The finding

Terrance Odean took the trading records of about 10,000 accounts at a discount brokerage and asked a simple question: when an investor sells, are they more likely to be selling something that has gone up or something that has gone down?

The answer was up, by a wide margin. Investors realised their gains at noticeably higher rates than their losses, all year except December, when tax selling briefly reverses it.

That alone is only a preference. The part that matters is what happened next: the winners they sold went on to outperform the losers they kept. The behaviour was not merely arbitrary — it was backwards.

Why it happens

The mechanism is prospect theory. A loss is felt roughly twice as intensely as a gain of the same size, and a loss on paper is not yet a loss you have admitted to. Selling converts it into one.

So the account holds two positions that feel completely different despite being the same decision:

  • The winner offers a small, certain pleasure now. Take it.
  • The loser offers a certain pain now, or the possibility of getting back to even later. Wait.

Neither of those sentences contains a single fact about the business.

Why the purchase price is the wrong anchor

The clean way to see the error is to notice what the decision is anchored on. Your purchase price is a fact about your history, not about the company. The market cannot see it, the business does not know it, and nothing about the next five years depends on it.

The only question that carries information is: given what this company is worth now, is this the best place for this money? An investor who was handed the same portfolio today, with no knowledge of what was paid, would frequently make a different decision — and that difference is the disposition effect, measured.

"Getting back to even" is the phrase to watch for. It has no referent in the business.

How this connects to Pythia

Almost everything on a company page is deliberately independent of what you paid. A PAS, a guru pass count, a valuation band — none of them can see your cost basis, and that is not an oversight.

The portfolio surfaces are the place to be careful. Cost basis is shown there because you need it for tax and for record-keeping, not because it belongs in the decision. If you find yourself reading a red number as a reason to wait, the number is doing work it cannot support.

A written thesis helps here for a specific reason: it fixes, in advance and in your own words, what would make you sell. A sell rule written before the position moved is the one piece of evidence the disposition effect cannot argue with.

The limits of this idea

It is a tendency measured across many accounts, not a law about any one trade. Selling a winner is often exactly right — the thesis played out, the position outgrew its slot, you need the cash. Holding a loser is often right too, because a price fall with the business intact is the ordinary case for buying more.

The effect says nothing about which of those you are doing. What the evidence supports is narrower and more useful: if your sell decisions correlate strongly with your purchase price rather than with your current view of the business, that correlation is costing you, and it is measurable in your own records.

Nor is the finding universal. It is strongest in individual accounts, weaker among professionals, and reverses in December when tax rules give the loss a use.

Check yourself

4 questions. Nothing is recorded unless you are signed in, and nothing here affects anything else.

1. An investor holds two stocks and has the same current view of both businesses. One is up 30% and one is down 30%. They sell the winner to lock in the gain and keep the loser until it gets back to even. Which documented pattern is this?
2. Odean found that investors realised gains at noticeably higher rates than losses. What turned that from a preference into a measurable mistake?
3. What is wrong with the sentence: I will sell this position once it gets back to even?
4. After reading about the disposition effect, an investor resolves never to sell a winner and to dump every loser at once. What does the evidence actually support?

4 questions left. An unanswered question counts as a miss, so the check waits for all of them.

Sources