Revenge Trading Is Real. The Story You Were Told About It Is Backwards.

Every trading course has the revenge-trading chapter. You take a painful loss, your judgment is hijacked, and you immediately fire off a larger, worse trade to win it back. It is a vivid story and most traders recognise it, which is why nobody checks it.

It has been checked. The research on how prior outcomes change subsequent risk-taking is decades old and largely points the other way.

What the data actually shows

The cleanest finding comes from a 2016 American Economic Review paper by Alex Imas, which resolved a long-standing contradiction in the field. The direction of post-loss risk-taking depends on one variable almost nobody controls for: whether the loss has been realised.

• After a realised loss — the position closed, the money gone, the number booked — risk-taking falls.
• After an unrealised loss — the position still open, still hurting, still theoretically recoverable — risk-taking rises.

That reconciles two field studies that had appeared to disagree for years. A study of professional futures locals at the Chicago Board of Trade found that traders who lost money in the morning were about 16% more likely than traders who were up to take above-average risk in the afternoon. A study of Taiwan Futures Exchange participants found the opposite — more afternoon risk after morning gains. Both are true. They were measuring different states.

Why the correction matters more than the diagnosis

If the danger were a bigger trade after a closed loss, the fix would be trivial: a cooling-off rule. Close a loser, stand up, no new position for twenty minutes. Every trading guide recommends it, and against the actual mechanism it is close to useless — because the dangerous state is not the one after the position is closed. It is the one where the position is still open.

Read your own worst days honestly and the pattern is usually this: the loss was never realised. It was managed. The stop was widened because the level “was not really invalidated.” The size was added to because “it is a better price now.” The target became breakeven, and then just getting out somewhere sensible, and then the market decided. The escalation happened inside a single trade, not between two of them.

This is why averaging down and revenge trading are the same disorder wearing different clothes. And it is why every institutional catastrophe in the stories section has the same shape: Leeson hiding a loss in account 88888 for four years, the London Whale desk selling ninety percent of a day's volume to defend a mark, the Hunts financing more silver. Not one of them was chasing a realised loss. Every one of them was defending an open one.

The three sentences

The internal monologue of a defended position is remarkably consistent, and worth learning to hear.

“It hasn't hit my stop yet.” True, and irrelevant, if you are watching the stop rather than the level that defined it. The question is not whether price has reached a number. It is whether the reason you took the trade is still on the chart.

“It's a better price now.” A better price for a position you have not yet analysed at this price. The test is the one from the sunk-cost problem: if I were flat right now, would I open this? If the honest answer is no, then adding is not conviction, it is accounting.

“I just need it to get back to breakeven.” The moment your objective becomes a number from your own past rather than something the market is offering, you have stopped trading the market and started trading your entry price. The market does not know your entry price and has no opinion about it.

What actually helps

Make the loss real early. The single most effective intervention available is a hard, pre-placed stop, because it converts the state where risk-taking rises into the state where risk-taking falls. Not a mental stop. A resting order — something that acts while you are in the state that stops you from acting.

There is a useful piece of evidence here from an unexpected direction: a 2022 Review of Financial Studies study found that regulatory leverage caps on retail FX traders reduced the disposition effect and improved timing, apparently by raising the cost of putting off a loss. A constraint you did not choose improved the behaviour you could not change. Choosing your own constraints in advance is the version of that available to you.

Decide the invalidation before the entry. Not the stop price — the condition. “This idea is dead if price accepts back inside the range” is a thought you can only have honestly while flat. Written down, it becomes something you consult rather than something you renegotiate.

Cap the day, not the trade. A daily loss limit is the only rule that works on the state you are actually in, because it does not require you to correctly diagnose your own condition in the moment. Two full stops and the platform closes is a rule that a compromised person can still follow. “Trade carefully when tilted” is not.

Journal the open position, not just the closed one. Most journals record what happened. The entry that changes behaviour is the one written while the trade is live and red: what is the reason it is still open? If the honest answer is “it might come back,” you have found the real leak, and it has a name and a timestamp.