Loss aversion: losses hurt more than equivalent gains please. Kahneman and Tversky built it into prospect theory as the steepest part of the value function at the reference point — the asymmetry that makes the S-curve bend more sharply below zero than above.
The typical loss aversion ratio is roughly 1.5–2.5: you need a gain about twice the size of a loss to feel indifferent. The ratio rises with stake size and can go infinite when ruin or lifestyle threat is on the table.
Mechanism
Outcomes are coded relative to a reference point (status quo, expectation, purchase price, high-water mark) — not absolute wealth. A $10,000 pay cut feels worse than a $10,000 raise feels good, even when the two jobs are otherwise equal. Adaptation plus loss aversion produces status quo bias: twins in hedonically identical environments both prefer to stay put once each treats his current state as reference.
Evolutionary logic: organisms that treat threats as more urgent than opportunities survive and reproduce. Professional traders who "think like a trader" show lower loss aversion in lab tasks — emotional reactivity to every tick is not universal.
Behavioral signatures
- Risk aversion for gains: take $900 sure over 90% chance of $1000.
- Risk seeking for losses: gamble to avoid a sure loss rather than accept it.
- Endowment effect: owners demand more to sell than buyers offer to buy — giving up what you have registers as loss.
- Disposition effect: sell winners, hold losers — closing a gain feels good; realizing a loss hurts (mental-accounting).
- Negotiation friction: concessions hurt; labor talks anchor on the existing contract.
- Sunk costs: money already spent lives in a closed account; walking away feels like loss even when future-only logic says quit.
Matthew Rabin showed mathematically that loss aversion cannot be explained by diminishing marginal utility of wealth alone — you need reference-dependent value.
In markets and judgment
Loss aversion makes unassisted memory a poor teacher: wins and losses are encoded asymmetrically, so a trade journal matters for process review (art-of-trading-with-light-su-zhu-and-hasu). Thesis drift and timeframe conflation are loss aversion working through rationalization — admitting the thesis failed means realizing the loss.
Munger's Deprival-Superreaction tendency is loss aversion amplified by social and identity stakes (psychology-of-human-misjudgment). Combined with availability and overconfidence, it feeds probability-blindness in real time even when the trader "knows" the math.
What to do
- Pre-commit invalidation levels and size before entry — decide when loss is acceptable while gain still feels abstract.
- Evaluate portfolios and processes in broad frames; ignore sunk costs in forward-looking decisions.
- Separate "pain of realizing" from "expected value of holding" — the feeling is data about psychology, not about edge.
- In organizations, reset reference points deliberately (new leadership, premortem, written pre-mortem journals) when sunk costs encumber judgment.
Sources
- thinking-fast-and-slow — prospect theory value function, loss aversion ratio, status quo bias, endowment effect, disposition effect, Rabin proof
- prospect-theory — parent model: reference dependence, diminishing sensitivity, probability weighting
- poor-charlies-almanack — Deprival-Superreaction as amplified loss aversion