Loss Aversion and the Disposition Effect

The asymmetry, and what it costs

Prospect theory describes how outcomes are experienced rather than how they are summed, and its central finding is an asymmetry: a loss of a given size is felt at roughly twice the intensity of an equivalent gain. That ratio changes nothing about the arithmetic of a portfolio, and it changes a great deal about the decisions made inside one. An investor looking at two open positions and deciding which to close is not choosing between two equal actions, because one of them feels considerably worse — and the position that is kept is the one that would produce the worse feeling to end. The consequence has a name. The disposition effect is the systematic tendency to sell winners too early and hold losers too long, and it is visible in brokerage records rather than inferred from a laboratory. Studies of individual accounts consistently find that the winners people sell go on to outperform the losers they keep, which means the bias does not merely feel bad — it moves money in the wrong direction. The mechanism is worth stating precisely, because it makes the failure easier to spot in a real account: the unpleasant half of a portfolio stays open as long as the loss is unrealised, so the tidier the account, the worse it feels. The cost also has a second, quieter form. Avoiding a realised loss delays the point at which capital is redeployed, so the position that should have been recycled into something better instead sits in the account doing nothing while it is being avoided. That is why the discipline is not about being brave enough to take a loss: it is about making the decision at a time when the asymmetry is not working on you. A day, priced two ways — Realised gain: +$1,200 · Unrealised loss: −$1,500 · Economic result: −$300 ← · Felt result, weighting the loss at 2.25×: −$2,175 ← · Gain that would feel neutral against that loss: $3,375 Note what the last row describes: to a investor feeling that loss, a $3,375 gain is the equivalent of flat. That is the size of the gap between the economic account and the experienced one, and it is the reason advice about discipline sounds reasonable and changes nothing.

The two questions that break it

Because the bias is arithmetic rather than a character flaw, it can be argued with numbers. The first question is the one that survives every portfolio review: would I buy this position today, at this price, with no history? If the answer is no, the position is being held for the accounting rather than for the thesis. The entry price is a tax record rather than a valuation, and it does not appear in anyone else’s estimate of what the business is worth. The second question reframes the whole decision. Instead of asking what the position might recover, ask what it would be worth to be flat — the figure in the table is $3,375 of avoided feeling — and then compare it with the cost of actually being flat, which in that example is $300. Framed that way, the investor who would pay thousands in felt value to avoid the loss should prefer to pay hundreds to be done with it. That is the move the lesson is built on: it converts a preference about how the day should feel into a decision about what the position is worth. Both questions need a moment when they can be asked honestly, which is why the process wraps them in structure rather than leaving them to a bad morning. A written stop, a target and a size decided at entry mean the exit was designed before the position could feel like anything; a scheduled review means the decision happens on a calendar rather than in a rush; and reviewing decisions rather than results means a good process that lost money is not punished. None of that removes the asymmetry, and all of it removes the situations in which the asymmetry decides. • “Would I buy it today, at this price, with no history?” • “What would I pay to be flat?” — then compare with the cost of being flat. • Exits written at entry, so the decision predates the feeling. • Reviews on a calendar, so the moment is chosen rather than arrived at. • Decisions reviewed on process, results read in samples rather than in single trades. Holding a loser because the thesis is intact and holding one because selling would admit the loss look identical from outside. The test that separates them is the first question, and it has to be asked against the facts as they are now rather than as they were when the position was opened.

The reference point is chosen — and the tax bill doubles the cost

Loss aversion is not a property of money; it is a property of a comparison. The same thirty thousand dollars feels like a loss against yesterday and a gain against a year ago, which means the intensity of the whole effect depends on the **reference point** you are using — and reference points are the most manipulable part of the apparatus. Your entry price, your month-to-date, your high-water mark and your position versus a benchmark are all available, and the one you happen to be looking at decides whether you feel ahead or behind. That is why the practical exercise in this lesson is to name the reference explicitly: a feeling about a position is not actionable until you can say what it is being compared with. The disposition effect has an empirical size worth knowing. The classic study of retail accounts found that investors were roughly **half again as likely to sell a winner as a loser** in a given period, which is the pattern you would predict from loss aversion plus the desire to feel vindicated. The cost is not only the winners sold early; it is the losers held past the point where the original thesis had clearly failed, because realising a loss turns a paper mistake into a recorded one. Both halves of that trade show up in the same account, and they compound. There is a hard financial reason the bias is more expensive in a taxable account than it looks, and it runs in the direction most people assume wrongly. **Realising a loss produces a tax benefit** — it can offset gains and, within limits, ordinary income — while an unrealised loss produces nothing. So the investor who cannot bear to sell a loser is simultaneously avoiding a benefit and holding an asset whose thesis they no longer believe. The disposition effect is not merely a psychological discomfort with a modest cost; in a taxable account it is a systematic transfer of value upward, to the tax authority and to whoever takes the other side of the trades. • The reference point is chosen — entry price, month-to-date, high-water mark — and it decides whether you feel behind. • Classic retail evidence: winners are sold about half again as readily as losers. • The cost is twofold: gains cut short, losses held past the point of disproof. • In a taxable account, realising a loss has a tax benefit that an unrealised one does not — so the bias costs more than it feels like.

The shape of the curve you carry

Behind the disposition effect and the reluctance to take a loss sits a description of how people actually weigh outcomes, and it is worth knowing because it predicts behaviour in places where the intuition alone does not. The description has three parts, and each one explains a different trading mistake. The first is that outcomes are evaluated as gains and losses against a **reference point**, not as levels of wealth — which is why the same balance feels like a win or a loss depending on what it was last week. This is the origin of the whole subject: the reference point is chosen, and a chosen reference point can be reset. Moving the reference to your entry price makes a position that is down feel like a deficit; moving it to the value of the business makes the same position a question about whether the business is worth today’s price. The second is that losses weigh more than equivalent gains — the commonly cited ratio is around two to one — which produces loss aversion. Its consequences are the ones this lesson has already covered: holding losers, selling winners, and the reluctance to realise a loss even when the tax code makes realising it cheap. The third is the part that is least familiar and most predictive: **the fourfold pattern**. People are risk-averse for moderate-to-high probability gains (they take the sure thing), risk-seeking for moderate-to-high probability losses (they gamble to avoid the certain loss), risk-seeking for small probability gains (they buy the lottery ticket), and risk-averse for small probability losses (they pay for insurance against the unlikely). Read as a description of a portfolio, that is a striking list: the same person will hold a diversified position, refuse to take a 20% loss, buy a long-shot option, and over-insure the house — and every one of those four decisions follows from the same shape. Two implications are worth carrying. The first is that the fourfold pattern predicts where selling discipline will fail, and it is not uniformly: the failure is strongest in the quadrant where a moderate probability loss is on the table, which is the ordinary losing position rather than the catastrophic one. The second is that probability weighting means small probabilities are overweighted — a one-in-a-hundred event feels more likely than it is — which is the mechanism behind the demand for far out-of-the-money options and lottery-like positions, and it is why those instruments are systematically expensive. The practical use is not to attempt to feel differently, which does not work. It is to recognise the shape and write rules that anticipate it: exit rules set before the position is entered, a realised-loss review on a schedule, and a deliberate check on any position whose appeal rests on a small chance of a large payoff. The curve is not going anywhere; the rules are what stop it from making the decisions. • Outcomes are weighed against a reference point you choose, and the choice can be moved. • Losses weigh roughly twice equivalent gains, which produces the disposition effect. • The fourfold pattern explains four ordinary behaviours with one shape. • Small probabilities are overweighted, which is why long-shot positions are expensive. The fourfold pattern is the best available explanation of why a portfolio can contain a disciplined core and a lottery ticket at the same time without the holder noticing a contradiction. Both decisions came from the same curve.

Sources

• Prospect theory and the asymmetry of valueKahneman & Tversky (1979), “Prospect Theory”
• The disposition effect in brokerage recordsOdean (1998), “Are Investors Reluctant to Realize Their Losses?”
• Process over prediction in decision-makingKahneman, “Thinking, Fast and Slow”