Disposition Effect

The disposition effect is the tendency to sell an investment that has risen in value sooner than one should, and to keep an investment that has fallen in value longer than one should. It was named and documented by Hersh Shefrin and Meir Statman in a 1985 study, "The Disposition to Sell Winners Too Early and Ride Losers Too Long," and it has since been confirmed in large samples of actual brokerage records. The behavior is driven by how gains and losses feel: selling a winner delivers the satisfaction of a realized gain, while selling a loser forces the investor to admit a loss, so the instinct is to take the gain and postpone the loss. That ordering conflicts with tax-efficient investing, which generally favors the reverse.

The disposition effect is a downstream consequence of prospect theory, and two of that framework's findings push in the same direction. The first is loss aversion, the finding that a loss of a given size is felt more intensely than an equal gain: because closing out a losing position makes the loss feel final, investors avoid it and hold on, while closing a winning position feels good, so they do it readily. The second is the reflection effect, the finding that people become comparatively willing to accept risk once they are below their reference point, so continuing to hold a loser can feel like the less painful choice rather than the riskier one. The distinctive feature of the disposition effect is that it turns those general findings into one specific, countable trading pattern. The result is a portfolio that sheds its winners and accumulates its losers, which is usually the wrong direction. Why the losers are held is not the same question as why they hurt, which is loss aversion's territory, nor the same as continuing a losing course because of money already committed, which is the sunk-cost fallacy, nor the same as dumping a falling holding in a broad decline, which is panic selling. The disposition effect is the narrower, mirror-image tendency about winners and losers held side by side.

The concrete cost is a tax cost. Selling an appreciated holding realizes a capital gain and, in a taxable account, brings the tax on that gain forward in time. Holding a depreciated holding means the loss is never harvested, so it cannot be used to offset gains elsewhere. Tax-efficient practice generally runs the other way: let unrealized gains keep compounding so the tax is deferred, and realize losses deliberately to offset gains, subject to rules like the wash-sale rule that bar rebuying the same security within a set window. The disposition effect pushes an investor toward the tax-inefficient ordering, paying tax on gains early and leaving losses on the table. None of this means a winner should never be sold or a loser never held; rebalancing, changed fundamentals, and cash needs are all valid reasons to trade. The effect is the systematic bias that operates when those reasons are absent, and the value of naming it is that an investor can check a sell decision against it.

The pattern operates position by position. Faced with a holding that is up, the investor feels the pull to realize the gain; faced with one that is down, the investor feels the pull to wait for a recovery. Left unchecked across many decisions, this sells the winners and keeps the losers.

A hypothetical example, with invented figures, of the tax cost in a taxable account. Suppose an investor holds two positions: Stock A, which is up $5,000, and Stock B, which is down $5,000, and both would be long-term sales. Acting on the disposition effect, the investor sells A and holds B. Selling A realizes a $5,000 gain, which at a 15 percent long-term capital gains rate costs $750 in tax now. Meanwhile the $5,000 loss on B goes unharvested, so it offsets nothing. The tax-efficient order is the reverse: sell B to harvest the $5,000 loss, which could offset $5,000 of gains and save about $750 in tax, and hold A to keep deferring the gain. The difference between the two orderings is roughly $1,500 in this scenario, the $750 of tax paid early plus the $750 of tax benefit forgone.

Pros

• Naming the effect gives an investor a specific bias to test a sell decision against, which is more actionable than a vague warning about emotions.
• It appears in real trading data, so it is a measured phenomenon rather than a folk belief.
• Understanding it points directly at the tax-efficient habit of deferring gains and harvesting losses.

Cons

• The effect describes a bias, not a rule; there are legitimate reasons to sell a winner or hold a loser, and treating the pattern as an absolute would be its own mistake.
• Because it feels like prudence ("lock in the gain," "give it time to recover"), it is easy to rationalize and hard to notice in oneself.
• Acting to counter it in a taxable account requires attention to related rules, such as the wash-sale rule, so it is not simply a matter of flipping the order.