Loss Aversion: Why You Cut Winners Early and Hold Losers Too Long
The pattern of closing winners early and letting losers run is not a character trait but a measurable quantity. Three numbers from your journal reveal it: average holding time of winners against losers, average R of winners against losers, and the share of trades that kept moving your way after you exited. When all three point the same direction, you do not have a strategy problem.
What is loss aversion in trading?
The tendency to weight a possible loss more heavily than an equally sized possible gain. In trading it does not operate abstractly but at two very specific moments: when closing a gain and when not closing a loss.
Both decisions feel reasonable at the time. Taking the gain presents itself as securing something, holding the loss as patience. That is exactly why the pattern goes unnoticed: it disguises itself as prudence, and every individual decision can be justified.
It only becomes visible across many trades, and only if you look at winners and losers separately. In a combined analysis it disappears completely.
Why does an unrealised loss feel different from a realised one?
Because it can be treated as though it has not happened yet. While the position stays open, the possibility of the price coming back exists, and that possibility keeps the decision open.
Arithmetically that is wrong. Your account is already smaller whether you close the position or not, because an open position is worth its current value and not its intended one. Closing changes nothing about your capital, it only changes when you acknowledge it.
The same applies in reverse to gains. An open gain is as real as a realised one, and taking it secures no more than holding it risks. It simply feels different, and that feeling decides whenever no rule stands against it.
How do you measure the pattern?
Through three figures, all built from data your journal already holds.
Holding time of winners against losers
Calculate average holding time separately for winning and losing trades. If losers are held substantially longer than winners, that is the most direct evidence of the pattern.
What matters is the direction rather than the absolute value. A difference can have a strategic reason, for instance with an approach using tight stops and wide targets. In that case, though, the opposite should appear: short losers and long winners.
Average R of winners against losers
Calculate both averages in R, meaning relative to the initial risk on each trade. An average R on losers above minus one means stops were overrun. An average R on winners well below your planned target means you closed early.
This is the most informative of the three, because it expresses both halves of the pattern in one unit and is independent of position size.
Share of winners closed too early
For a period, record what the price did after your exit, measured either to your originally planned target or across a fixed window afterwards. The share of trades that kept going without you is the most uncomfortable and the clearest number of the three.
This recording costs effort, because it happens after the trade. It is worth running for a bounded period such as a quarter and does not have to continue indefinitely.
Why is a good hit rate a warning sign here?
Because a high hit rate alongside a weak result is almost the signature of this pattern. Taking gains early means winning more often, since a small target is reached more frequently than a large one. Letting losses run means losing less often but more expensively.
The result is a record that looks good at first glance: many wins, few losses, and an account that nonetheless does not grow. Hit rate on its own therefore says nothing about the quality of an approach.
The cross check is simple: multiply average gain by your hit rate and average loss by the complement. Only the difference tells you whether your approach carries. The groundwork sits in trading psychology and discipline.
How does this differ from tilt and FOMO?
Through the trigger. Tilt is a reaction to a loss, FOMO a reaction to a missed move. Both have an external cause, set in quickly and fade.
Loss aversion has no cause. It operates identically in every trade, permanently and inconspicuously, and it shows up not as conspicuous behaviour but as a slight shift in every single exit decision. That is why you never notice it in the moment, only in analysis.
Practically that means the measures against tilt and against FOMO do not apply. A mandatory break addresses a reaction, not a baseline pattern. And a cap on trades, which works against overtrading, changes nothing about how you close positions you already hold.
Why is the pattern so persistent?
Because it works often enough. A loss you sit on does come back in a share of cases, and those are exactly the cases you remember. They supply the evidence that patience was right, and they occur frequently enough to keep reinforcing the behaviour.
That is why arguments achieve so little here. You do not remember a statistic, you remember specific positions that recovered. That the cases without recovery cost more is arithmetic rather than experience.
Only one approach handles that: counting instead of remembering. Take every trade where you overran your original stop and put two numbers side by side. How many came back and with what result, and how many ended worse than the original stop and by how much.
That analysis needs one journal field, namely the stop you set first. Without it you cannot establish afterwards which trades were even affected, and that is precisely why the analysis is missing from most journals.
Which countermeasures actually work?
Only those that take the exit out of the moment. Resolutions do not work, because they get tested precisely while the price is moving.
- Define both exits before entry. Stop and target belong in the same working step as the entry decision, written down with prices.
- Treat the stop as immovable, except in one direction. Trailing in favour of the position is a rule, moving it to create room is the suspension of one.
- Automate whatever can be automated. An order resting in the market acts without you. It is the single most effective step against both halves of the pattern.
- Scale out by rule rather than by reaction. Anyone who knows the urge to secure early can define a fixed step in R. What matters is that it is set beforehand and always applies, not only when the price action turns uncomfortable.
What do you do with the stop you want to move?
Record the impulse instead of acting on it. That sounds like a way of dodging the question, but it is the single most effective practical measure, because it achieves two things at once.
First, it inserts an action between impulse and execution. Second, it generates data: after a quarter you can analyse how the trades would have ended in which you wanted to move the stop, and whether your impulse was right in those cases.
In most journals that analysis comes out one sided, and more convincingly than any argument could. That is the difference between an insight you read somewhere and one that comes out of your own numbers.
How do you check whether anything changed?
Through those same three figures compared against the prior period, not through your result. Four weeks suffice for the process figures and do not suffice for the result.
Expect an uncomfortable interim phase: your hit rate falls, because gains are no longer taken early, and that initially feels like a deterioration. Only the average R of your winners shows whether the trade off pays.
A properly kept trading journal supplies all three figures without extra effort, once timestamps, planned target and stop are recorded per trade. The only item that has to come from you is the planned exit, and it has to be there before the trade.