OptiNod Academy
Prospect theory and loss aversion — Why losses feel more than twice as strong as gains
People feel gains and losses relative to a reference price, and losses have a steeper psychological slope. Average winning and losing R and holding time reveal the asymmetry.
Even at the same size, the pain of a loss can exceed the pleasure of a gain by more than twofold. This asymmetry cannot simply be willed away; it appears in average winning and losing R and in holding times.
Prospect theory is a theory of decision making proposed by Daniel Kahneman and Amos Tversky in 1979. Its core is that people evaluate gains and losses relative to a reference point, often their current state or purchase price. The absolute account balance is not the immediate basis for judgment. What matters is how far profit and loss have moved from the point where they were zero.
The researchers also showed that the value function describing this evaluation is steeper for losses. In cumulative prospect theory in 1992, they estimated a loss-aversion coefficient of about 2.25. In that model, a loss of a given amount feels about 2.25 times as strong as a gain of the same amount. Losing $100 can hurt more than twice as much as gaining $100 feels good.
It is often reduced to being “too afraid to take a stop.” Then the proposed remedy is simply to “be tougher,” and resolve fails again at the next loss. Loss aversion is part of the structure of human judgment and does not disappear with a better attitude. Its asymmetry appears in two measurable places: choices about stop and profit-taking distances, and the arithmetic of recovering drawdowns. A 20% loss requires a 25% gain to recover; a 50% loss requires a 100% gain. The deeper the loss, the faster the required recovery grows relative to it.
Profit and loss are felt relative to a reference point
The first pillar of prospect theory is reference dependence. People do not experience gains and losses only as an absolute account balance; they feel movement from a chosen level. The same $5,000 balance can feel like a loss after falling from $6,000, or a gain after rising from $4,000. The amount is identical, but a different reference point reverses the experience.
For traders, the reference is usually their entry price. That creates a psychological line called break-even. Above entry is the gain region; below it is the loss region, and people handle risk differently on either side. The market does not know an individual's entry price. It moves independently of that personal reference. If entry remains the anchor, decisions about profit and loss can ignore market structure in favor of the price paid.
The value function is roughly twice as steep on the loss side
The second pillar is the shape of the value function. Its shape encourages risk aversion in gains and risk seeking in losses. The loss side is steeper than the gain side. A loss-aversion coefficient around 2.25 describes that difference in estimated sensitivity.
At an exit, the shape produces two opposing habits. In profit, the chance that a certain $100 gain disappears can feel more urgent than the chance it grows to $200, so a small gain is taken quickly. In loss, realizing a $100 loss feels disproportionately painful, so the trader holds on for a small chance of getting back to even. Each decision feels reasonable at the moment, though both arise from the same value function acting on opposite sides of the reference.
The asymmetry pushes stop and target distances in opposite directions
The value function produces two live behaviors that a backtest may not encode. Risk aversion in gains brings profit-taking forward, making realized winning R smaller than planned. Risk seeking in losses delays stops, making realized losing R larger. Together they can leave average losing R greater than average winning R.
This exit pattern is the disposition effect. The exit-focused articles in this series grow out of these two sides of loss aversion. Even when entry rules have positive expectancy, cutting the winning tail and stretching the losing tail can push live expectancy below the backtest's assumption. The tails of the outcome distribution change regardless of whether the entry signal was correct.
Loss asymmetry appears again in recovery arithmetic
The asymmetry is not only psychological. A loss always requires a larger percentage gain to return to the starting balance, with the gap widening as the loss deepens. A 20% drawdown needs 25% to recover; 50% needs 100%. Drawdown recovery math shows the calculation.
Real declines illustrate it. Bitcoin fell about 36% from $126,200 on October 6, 2025 to $80,600 by November 21. Recovering to the starting point from that low would require roughly 57%. In early 2026, it fell roughly 39% from about $98,000 in mid-January to $60,000 by February 6; recovery then required about 63%. As losses deepen, the gap between the decline and the needed rise expands. Psychological asymmetry (about 2.25 in the cited model) and arithmetic asymmetry act in the same direction. The longer a loss is held, the larger the rebound required to recover. That is one reason maximum drawdown limits what a strategy can withstand.
Two numbers in the trade record reveal loss aversion
If loss aversion is treated only as a willpower problem, there is no clear place to intervene. Compare winners and losers in R and in time held.
First, compare average winning R and average losing R. Convert realized profit or loss on every trade into units of planned entry risk (1R), then average each group. When loss aversion shapes exits, average losing R can exceed average winning R. For example, a 0.8R average win and 1.4R average loss give a reward-to-risk ratio of about 0.57. Combine that ratio with win rate to calculate expectancy. Even with many winners, a large imbalance can make it negative.
Second, compare average holding time for winners and losers. Closing gains quickly and keeping losses longer appears directly as a gap in duration. With loss aversion, losers are held longer. When losing R exceeds winning R and losers are held longer than winners, the two measures together support a diagnosis of loss aversion affecting exits.
Rules move the reference away from entry price
Loss aversion operates against a reference such as entry price. Just before a trade, profit and loss is zero and there is less reason to lean toward either side. Shift the decision basis to market structure and decide exits at entry. Set the stop and target before price moves away from the entry to reduce the room for later reference-driven decisions.
- Use structure rather than entry as the basis: Choose stop and target from volatility such as ATR and prior swing structure. Do not use the entry price itself to justify an exit level.
- Place exits with the entry: Submit a 1R stop and first target as OCO orders together with the entry, and avoid changing them after the fill.
- Move stops only toward profit: A stop may move toward break-even or gains, not farther into loss.
- Record in R: Note realized R and holding time at every exit.
- Compare weekly: Calculate average winning and losing R and holding time each week; if the gaps grow, revisit exit rules.
The purpose is to base decisions on market structure rather than a personal purchase price so live trading retains the distribution assumed in the backtest.
Two pitfalls
“I will sell at break-even.” Holding a loser until it returns to entry turns reference dependence into a supposed rule. Break-even comes from your personal entry; the market does not know it. Set the stop from structure and volatility instead.
A healthy-looking win rate is an illusion. Wide stops and narrow profit targets can lift win rate while average losing R overtakes average winning R and expectancy turns negative. Win rate alone hides the asymmetry. Always look at average gains and losses in R together.
Rules can work around loss aversion even if they cannot remove it
Loss aversion is a common feature of human judgment, not something willpower can simply erase. People feel outcomes relative to a reference and weight the loss side more sharply. What can be changed is whether that asymmetry enters an exit decision. Put stop and target on market structure rather than your entry price before the trade begins. Then compare average winning and losing R and the holding times of winners and losers. “I struggle with losses” becomes a measurable exit pattern. Moving the reference to market structure and recording outcomes in R are practical ways to work around the asymmetry.