TL;DR
A revenge-trading chain rarely costs what the first trade cost. It costs what the whole sequence costs — usually two or three trades, each one bigger than the last, because the trader is solving for "get back to breakeven" instead of following their normal size. Measuring loss per trade hides this pattern. Measuring loss per episode is what shows the real damage, and where to actually intervene.
A revenge trade is a re-entry right after a loss, at a size nobody planned for. That part is well covered elsewhere. What matters more is what happens next: that first re-entry is rarely the end of it, and trade two and trade three in the sequence aren't just "more of the same loss." They're worse-positioned, worse-sized, and cost disproportionately more than the trade that started the chain.
The chain, with numbers
Take a trader with a normal setup: 3 lots, a defined stop, a loss of ₹6,000 when the stop is hit. Nothing unusual — this is what a bad trade inside a sound process looks like.
Eight minutes later, the same trader is back in, same instrument. The size is now 4.5 lots — 1.5x the original — not because the setup improved, but because the entry was chased and the stop was given "more room to work." That trade also fails. Loss: ₹12,000. Not 1.5x the first loss — 2x it, because the worse entry and the wider stop did more damage per lot than the original position would have.
Six minutes after that, a third trade. The deficit now sitting at ₹18,000 is the number in the trader's head, and the size needed to erase it in one shot works out to roughly 7 lots — 2.3x the original size. From here the chain splits two ways:
- It loses again. Down ₹20,000 on trade three. Total cost of the episode: ₹38,000. Trade one alone cost ₹6,000 — the chain cost 6.3x that.
- It partially recovers. Up ₹9,000 on trade three. Net cost of the episode: ₹9,000 — better than the losing branch, but still 1.5x what trade one alone would have cost, and it came from putting 14.5 lots of cumulative exposure at risk to defend a position that was originally sized at 3.
Look at the shape of that, not just the total. The loss roughly doubles from trade one to trade two, and the size needed to "solve" trade three is almost 2.5x where the trader started, all inside about fifteen minutes. None of that was a plan. It was arithmetic done under pressure, three times in a row.
See the real cost of your chains
Auraxon groups re-entries after a loss into a single episode automatically, so you see the total cost — not just the first trade.
Why size creeps up without anyone deciding it should
Nobody sits down after a loss and consciously decides to risk more. What actually happens is smaller and more mechanical: the trader anchors to the rupee number they're down, not to their normal risk percentage, and then works backward to a lot size that could plausibly erase that number in one trade. At a given stop distance, recovering a bigger deficit requires a bigger position — that's just the math of it. The size increase isn't a risk decision at all. It's the output of a different, unstated question: "how many lots do I need to get back to zero," instead of "what does my plan call for here."
Each failed attempt makes this worse, because the anchor keeps moving. After trade two, the deficit isn't ₹6,000 anymore — it's ₹18,000, and the lot size that answers that new number is bigger again. Time compresses too: the eight minutes between trade one and trade two shrinks to six between trade two and three, which is exactly the window in which a trader would normally step back and reassess. There isn't one bad decision to point to in this sequence. There's a sizing formula running silently in the background, recalculating upward every time it's fed a bigger number.
Cost per episode, not cost per trade
This is where most trade review gets the measurement wrong. Logged trade-by-trade, this sequence looks like three separate, unremarkable events: a loss within normal range, a somewhat larger loss, and — in the recovery branch — a win. Reviewed that way, trade three even shows up as a positive entry in the ledger. Nothing about the individual rows flags a problem.
Grouped as one episode, the picture changes completely. The relevant number isn't any single trade's P&L — it's the total capital put at risk across the sequence (14.5 lots deployed against a plan that called for 3) and the net cost of the whole thing, which is higher than trade one's cost even in the branch where the trader "won back" money. A partial recovery doesn't erase the episode's cost; it just makes the episode look survivable in a ledger that isn't tracking risk exposure.
This reframing changes where you'd actually intervene. If you're measuring per trade, you're looking for a bad decision inside trade two or three to correct. If you're measuring per episode, the intervention point is obvious and it's earlier than that — it's the gap between trade one and trade two, before the first re-anchoring has happened and before size has moved at all. Catch the chain there, and there's no trade two to escalate from, no trade three to size up further. The entire multiplier — 6.3x in the example above — never gets applied.
Why does a revenge-trading chain cost more than the sum of its individual losses suggests?
Because position size increases at each re-entry, later trades in the chain lose disproportionately more than the first one did — a 1.5x size increase on trade two produced roughly a 2x loss increase in the example above, not a proportional one.
Why does position size increase across a revenge-trading chain, if the trader isn't deciding to risk more?
The trader is solving for the rupee amount needed to erase the running deficit at a given stop distance, not for their normal risk percentage — a bigger deficit mechanically requires a bigger position to 'fix' in one trade.
If the third trade in a chain partially recovers the loss, did the episode still cost something?
Yes — the net P&L may look small, but the capital exposure across all three trades was several times the original planned size, and the net cost is still higher than the first trade would have cost on its own.
How many trades does a typical revenge-trading chain involve?
Most chains run two to three trades before the trader either stops, hits a hard limit, or the account absorbs a loss large enough to force a pause.
Should trading cost be measured per trade or per episode?
Per episode. Per-trade review can make the second and third trades in a chain look like ordinary results — sometimes even a win — which hides the total capital exposure and net cost of the sequence as a whole.
What's the best point to interrupt a revenge-trading chain?
Between the first loss and the first re-entry, before any re-anchoring has happened — once trade two is open, size has already moved once and is likely to move again.