You can spot a revenge trade in almost anyone's trading history without knowing a single thing about the trader. It sits a few minutes after a big red trade, usually in the same market, at noticeably larger size, and the stop is either far wider than anything else in the log or missing entirely. I have several of these in my own history and they follow the pattern so faithfully it is almost funny. Big loss, then within minutes a re-entry that looks nothing like the trades that came before it.
Part of why it happens is that a large loss registers, in the moment, as an error that can still be corrected rather than as a sunk cost. The money feels temporarily misplaced, retrievable if you move fast enough. So the next order goes in as an attempt at a refund, and refunds run on different rules than trades. A refund needs to be big enough to cover the loss, which explains the doubled size. A refund cannot be allowed to fail, which explains the widened stop. And a refund is urgent, which explains why the whole thing happens inside five minutes instead of waiting for the next real setup.
The three moves, and why each one feels reasonable
The immediate re-entry feels reasonable because you were just in this market and you have a view. Nothing new has happened since your stop was hit, but sitting flat feels like conceding, so back in you go, usually in the same direction, occasionally in the opposite direction out of spite, which is somehow worse.
The doubled size feels reasonable because of the recovery math. If the loss was two units of risk and you double your size, one ordinary winner makes you whole. Written out like that it looks like arithmetic, and arithmetic feels calm and objective, so the sizing decision borrows a credibility it has not earned. The math only works if the next trade wins, and there is no reason to think your hit rate improved in the last four minutes. If anything it got worse.
The widened stop feels reasonable because, in the story you are telling yourself, the stop is what caused the loss. The trade was right and the stop was wrong, so this time you give it room. What you have actually done is remove the one control that caps how bad the next mistake can be, on the largest position you have taken all week, at the moment your judgment is most impaired.
Put the three together and you get the structural problem. Whatever edge you have lives in planned setups, taken at planned size, with planned exits. The revenge trade is by definition unplanned, oversized, and loosely exited, so your expectancy is at its lowest at exactly the moment your exposure is at its highest, and that inversion is what turns a bad day into a bad month. The second loss is bigger than the first, which produces a third trade sized to recover both, and account-ending drawdowns are usually built from this staircase rather than from one bad idea.
A cooldown protocol that actually holds
The fix has to be mechanical, because the entire problem is that your judgment is offline right when you need it most. Rules that live in your head fail under tilt. Rules written down when you were calm have a chance. Mine has three parts.
First, a mandatory time-out. Define a threshold in advance, for example a single loss bigger than roughly twice your average loss, or hitting your daily loss limit. Cross it and you are flat and away from the screen for a fixed period, no exceptions and no just watching. For a moderate breach I use somewhere between thirty and sixty minutes. For a big one, the rest of the session. The point of the time-out is physiological rather than analytical. The stress response after a large loss does not clear in two minutes, and you cannot reason your way out of it while staring at the chart that caused it. Close the platform and go walk somewhere, because the market will still be there when your pulse settles.
Second, a size reset. Your first trade back after a time-out is at half your normal size, or your minimum unit, regardless of how good the setup looks. This is temporary, one or two trades, then normal sizing resumes. The reset does two jobs. It caps the damage if you are still tilted and misjudging your own state, which is common, since the trader who most needs the cooldown is the one most convinced he does not. And it forces you to accept that the loss gets made back slowly or not at all. If half size feels unbearable, that feeling is the diagnostic, because it means you wanted the refund and not the trade.
Third, a re-entry checklist that has to be satisfied, in writing, before the next order goes in. Mine looks roughly like this:
- This setup would qualify under my rules if today's loss had never happened, and I can write the thesis, the entry, and the invalidation in two sentences.
- The stop is at my normal technical distance, and the position size is computed from that stop, not from the amount I want back.
- I can name specific new information since my losing trade. If nothing has changed, I am re-entering the same trade for emotional reasons and it fails the checklist.
- The size is at or below my reset size.
- The full time-out has elapsed, all of it, not most of it.
None of these items is clever, and that is deliberate. A checklist for tilted traders has to be answerable by a tilted trader, which means yes or no questions with no room for interpretation. The moment an item requires judgment, tilt will happily supply the judgment.
Making it stick
Enforcement is where most versions of this fail, so add friction on the violation side. Log out of the platform when the time-out starts instead of minimizing the window. If your venue supports a daily max-loss setting or a trading pause, turn it on, since a rule the software enforces beats a rule you enforce. Tell one other person about the protocol so that breaking it carries a small social cost.
Then audit yourself, because the argument with your own brain is easier to win with your own data. Tag every trade placed within, say, thirty minutes of a loss and check how that cohort performed after a quarter. I ran this on my own fills in Blockcircle and the fast re-entries were my worst group by a margin wide enough that I stopped debating the rule with myself. Most traders who try this exercise find something similar, which makes sense once you accept what those trades actually were.
The protocol will feel like overkill the first several times it triggers, because most of the time the trade it blocks would have been fine, and that is the cost of the insurance. The one time it stops you from doubling into your worst hour of the year pays for every fine trade it blocked, and you will probably never know in advance which time that will be.