Loss Aversion

Loss aversion is the tendency for losses to weigh more heavily than gains of equal size. It was identified by Daniel Kahneman and Amos Tversky in prospect theory, published in Econometrica in 1979, and it is one of the central pieces of the work cited when Kahneman received the 2002 prize in economic sciences. The Royal Swedish Academy's scientific background for the 2002 prize in economic sciences summarises both the effect and its measured size: people "appear to be more adverse to losses, relative to their reference level, than attracted by gains of the same size," and a later estimate by Tversky and Kahneman put "the value attached to a moderate loss" at about twice the value attached to an equally large gain. The comparison is always to a reference point, most often the current state of affairs, rather than to any absolute level of wealth.

The most useful thing to be precise about is that loss aversion is not risk aversion, because the two get used as synonyms and they predict opposite behaviour in one important case. Risk aversion is a dislike of uncertainty as such, and someone who is risk averse dislikes it whether they are ahead or behind. Loss aversion depends on which side of a reference point an outcome falls on. In prospect theory the value function is steeper just below the reference point than just above it, producing the asymmetry, and it also flattens as outcomes get further from the reference point in either direction. The consequence is counterintuitive and well documented: people tend to be cautious about gains and comparatively bold about losses. The 2002 scientific background notes that this apparent risk-loving behaviour over large losses "is inconsistent with the traditional assumption of risk aversion." Someone already down 30% may reach for the riskier option precisely because they are down, which is the opposite of what a simple dislike of risk would produce.

The reference point does most of the work, and it is more arbitrary than it feels. For an investment it is usually the purchase price, occasionally the highest value the account ever reached, and sometimes a number the investor hoped to hit. None of those is information about the asset. A fund bought at $40 and now worth $32 is the same fund, with the same prospects, as one bought at $24 and now worth $32, yet the two owners will describe their positions in opposite language and often act differently. How often the reference point is reset matters as much as where it sits. The Royal Swedish Academy's scientific background for Thaler's 2017 prize describes a model by Shlomo Benartzi and Thaler in which "the impact of loss aversion depends on how often investors reset their reference point," a combination of loss aversion with narrow bracketing that the Academy calls myopic loss aversion. Checking a portfolio daily registers far more losses than checking it twice a year does, without changing the outcome at the end.

In practical financial life the effect shows up as a small number of recurring patterns. Favourable gambles get declined when the downside is described as a loss. Positions that have fallen get held past the point where the owner would buy them again, because selling converts a paper loss into a settled one. Rebalancing feels wrong at exactly the moment it does the most work, since it requires buying more of whatever has just fallen. And a sharp market decline invites selling, which is where the largest measurable damage is usually done. Each of those has its own name in the literature and its own mechanics; loss aversion is the common ingredient rather than the full explanation of any one of them.

One honest caveat about the number. The roughly two-to-one figure comes from experiments, and the Academy's own summary attaches it to a "moderate" loss and describes the phenomenon as local. Estimates move with the size of the stake, the way the choice is framed, and the population studied. Treat it as a reliable direction and a rough order of magnitude, not a constant to compute with.

Start from a reference point, usually where things stand now. Score each possible outcome as a gain or a loss relative to that point rather than as a final level of wealth. Weight the losses roughly twice as heavily as the gains. Then choose. That sequence reproduces a large amount of observed behaviour that a straightforward expected-value calculation gets wrong.

A worked example, using an illustration from the Royal Swedish Academy's 2002 scientific background. You are offered a coin flip: heads you win $12, tails you lose $10. The expected value is favourable, at (0.5 × $12) − (0.5 × $10) = $6 − $5 = +$1, and yet most people decline it. Apply a loss weight of two and the refusal becomes arithmetic rather than a mystery: (0.5 × $12) − (0.5 × $10 × 2) = $6 − $10 = −$4. On the felt scale the bet is a loser, so declining it is a correct decision about the wrong quantity.

The same weighting scales up, which is where it starts to cost real money. A hypothetical portfolio falls from $400,000 to $340,000, a $60,000 decline. Weighted at two to one, that registers with about the force of a $120,000 gain, and it is competing for attention against a plan built around a horizon of decades. Nothing about the household's goals changed when the balance did. The felt magnitude of the move is simply out of proportion to its significance, which is why the decision most likely to be regretted is the one made while looking at the number.

Pros

• The instinct is protective in the situations it evolved for. Weighting losses heavily discourages leverage, speculation, and betting money you actually need.
• It is a reasonable default when a loss would be unrecoverable, which is the case for an emergency fund or money needed next year.
• Because the effect is systematic rather than random, it can be planned around in advance instead of merely regretted afterwards.

Cons

• It leads people to decline risks that are favourable on the numbers, including the long-horizon investment risk most retirement plans depend on.
• It makes selling a losing position feel like causing the loss rather than recognising one, so bad holdings get kept.
• It makes rebalancing feel wrong at the moment it matters most, because restoring a target allocation means buying whatever just fell.
• It attaches decisions to the purchase price, a number that carries no information about what the investment will do next.
• Applied to each holding and each short period on its own, it produces worse decisions than the same person would reach looking at the whole portfolio over its full horizon. That narrow framing is the other half of what researchers call myopic loss aversion.