Psychology › Thinking, memory & biases › Behavioral economics
Loss aversion
1979· Finding or experiment
Well replicatedStrong, repeated evidence.
Losing something hurts roughly twice as much as gaining the same amount feels good.
See it on the Psychology map →What it says
Offered a coin flip that loses $100 on tails, most people want a win of about $200 on heads before they will take it. That ratio, near 2 to 1 in many studies, is the heart of prospect theory. It makes people cling to the status quo, refuse fair bets and dread paper losses. For investors, a 20% drop feels far worse than a 20% gain feels good, one reason many sell in panics or sit in cash.
Why it matters
Perhaps the most money-relevant bias: it fuels panic selling, holding losing positions, overpaying for small insurance and resisting any change framed as a loss.
Think of it like
Like a vacancy: losing a tenant stings more than signing a new one at the same rent pleases.
The evidence
Well replicated in core choice experiments with real and hypothetical money. A 2024 meta-analysis of hundreds of estimates by Alexander Brown, Colin Camerer and colleagues put the typical ratio near 2.
Open questions
Why it exists is unclear. David Gal and Derek Rucker (2018) argue many loss-aversion results reflect status-quo inertia, and it may be weak for small stakes.
Inside this (3)
Connections (5)
▸Loss aversion may partly depend on The amygdala: threat detection and fearMixed evidence
The connection
The amygdala, the brain's threat detector, is one candidate for the extra sting of losses. In a 2010 study, two women with rare damage to both amygdalae showed no loss aversion on coin-flip gambles, while matched volunteers did. Other imaging studies point more to reduced activity in reward areas, so its role is unsettled.
Who showed it and when
Caltech, 2010: Benedetto De Martino, Colin Camerer and Ralph Adolphs; earlier brain-imaging work at UCLA by Sabrina Tom and colleagues (2007).
How solid is it?
Mixed evidence: one small lesion study of two patients and imaging studies with differing results. Brain correlations alone do not show the amygdala causes the effect.
Open questions
Whether loss aversion comes mainly from fear-like emotion, weaker reward signals or extra attention to losses; larger patient and brain-stimulation studies are needed.
▸Loss aversion is the heart of Prospect theoryWell replicated
The connection
In prospect theory's value curve, the line is steeper for losses than for gains, about twice as steep in typical estimates. That one asymmetry drives many of the theory's predictions: refusing fair bets, sticking with the status quo, and the gap between buying and selling prices.
Who showed it and when
Kahneman and Tversky, 1979; their 1992 cumulative prospect theory paper estimated the ratio at about 2.25.
The evidence
Coin-flip acceptance studies and a 2024 meta-analysis of hundreds of estimates put the ratio near 2.
Open questions
Some researchers argue the asymmetry depends on stakes and context, and that other parts of the theory, like probability weighting, matter as much.
▸Disposition effect: selling winners, holding losers is usually traced to Loss aversionWell replicated
The connection
Investors measure each stock against their purchase price. Below it, selling makes a loss final, and prospect theory says people take risks to avoid locking in losses, so they hold and hope. Above it, they lock in the gain. Shefrin and Statman also pointed to mental accounting and regret.
Who showed it and when
Hersh Shefrin and Meir Statman, Journal of Finance, 1985; tested on brokerage records by Terrance Odean in 1998.
The evidence
Brokerage records worldwide match the pattern; in a 1998 lab study the effect shrank when holdings were automatically sold and reset, removing the purchase price.
Open questions
Models show loss aversion alone does not reliably predict it (Barberis and Xiong, 2009); a pleasure from realizing gains may explain it better.
▸Sunk cost fallacy is fueled by Loss aversionWell replicated
The connection
Walking away from money already spent means accepting a certain loss, and losses hurt roughly twice as much as equal gains please. So people keep investing to avoid 'wasting' the earlier outlay, hoping to break even. Prospect theory predicts this: after a loss, people become more willing to gamble to recover it.
Who showed it and when
Hal Arkes and Catherine Blumer at Ohio University (1985) tied the sunk cost effect to Kahneman and Tversky's 1979 prospect theory.
The evidence
Experiments show people hate waste and take bigger risks after losses; the link to loss aversion is a strong theoretical fit but rests partly on interpretation.
Open questions
Whether a dislike of appearing wasteful, or protecting one's reputation, explains sunk-cost behavior better than loss aversion alone.
▸Endowment effect: valuing what we own more is largely explained by Loss aversionWell replicated
The connection
Kahneman, Knetsch and Thaler argued that giving up an owned item is coded as a loss, while acquiring one is a gain. Because losses weigh about twice as much, sellers should want roughly twice what buyers offer, close to the mug study's gap of a bit over two to one.
Who showed it and when
Thaler proposed the link in 1980; the 1990 mug experiments at Cornell and Simon Fraser University by Kahneman, Knetsch and Thaler tested it.
The evidence
The gap appears for ordinary goods and shrinks for goods held for trade, as loss aversion predicts.
Open questions
Rival explanations include uncertainty about value, attachment to one's things and experiment design; Plott and Zeiler's 2005 procedures shrank the gap.