A trade is down 3R and has blown through the stop-loss level twice already, but closing it now would mean "admitting" the first bad decision, the hours spent watching it, and the stress already absorbed were all for nothing — so it stays open, hoping to get back to even. A strategy has eaten six months of development time and a paid data subscription, and even though the live results say it doesn't work, shutting it down feels like throwing away everything already invested. Neither decision is about the trade or the strategy anymore. Both are about refusing to let go of a cost that is already gone and can never be recovered — the sunk cost fallacy.
What Is the Sunk Cost Fallacy?
The sunk cost fallacy is the tendency to keep investing time, money, or effort into a decision because of what has already been committed to it, rather than because of what makes sense going forward. A "sunk cost" is any resource that has already been spent and cannot be recovered regardless of what happens next — a loss already sitting on an open trade, months already spent building a strategy, a subscription already paid for. Rational decision-making says sunk costs should be irrelevant to what happens from this point forward. The fallacy is acting as if they aren't.
Sunk Cost Fallacy: continuing a losing course of action because of the resources already invested in it, even when those resources can never be recovered and are irrelevant to the best decision going forward.
In trading, the fallacy shows up in two places at once: inside a single trade, where the "cost" is the loss already showing on the screen, and across a whole strategy or account, where the "cost" is the time, money, or identity already poured into an approach that isn't working. Both versions produce the same behavior — holding on longer than the setup or the evidence justifies, purely because stopping now would mean accepting that the sunk cost bought nothing.
How the Sunk Cost Fallacy Actually Shows Up at the Chart
- Holding a losing trade well past its stop-loss level because closing it "locks in" a loss that otherwise still feels avoidable
- Widening a stop or removing it entirely mid-trade to avoid realizing a loss that was already, functionally, decided by the setup failing
- Continuing to fund or trade a strategy that has stopped working because of the months spent building, backtesting, or learning it
- Averaging down into a losing position specifically to lower the average cost and "rescue" the money already committed, rather than because the setup still qualifies for entry
- Staying in a trade because of the hours already spent watching it, as if screen time were a resource that could still be recovered by waiting longer
Why the Brain Is Wired This Way
The sunk cost fallacy comes from loss aversion combined with a need for past decisions to feel justified. Admitting a trade or a strategy isn't working means admitting the original decision to enter it was wrong, and the brain treats that admission as a loss in itself — on top of the financial one. Continuing to hold or fund the position lets the story stay "not decided yet" instead of "wrong," even though the market has already made the decision either way.
It overlaps with loss aversion but isn't identical to it. Loss aversion is the general tendency to feel losses more intensely than equivalent gains. The sunk cost fallacy is more specific: it is letting a past, unrecoverable cost influence a forward-looking decision that should be made purely on current information. A trade already down 3R should be evaluated the same way whether it's down 3R or freshly entered — what matters is whether the setup still has an edge from here, not how much has already been lost getting to this point.
The Math: Why Money Already Lost Should Never Decide the Next Move
A trade currently down 3R has, from this exact moment forward, only two possible paths: it can be closed now for a known 3R loss, or it can be held and either recover or lose more. The 3R already lost is identical in both paths — it cannot be changed by holding, and it cannot be "recovered" by anything other than the trade's future price action, which has no idea how much has already been lost getting here. The only question that should decide the next move is whether, starting from the current price with no prior history attached, this is still a position worth holding. If the answer is no, the size of the existing loss is irrelevant to that answer.
$0
the amount a sunk cost is worth toward improving the next decision — it is already spent regardless of what happens next
2x
roughly how much larger the eventual loss tends to be when a losing trade is held past its stop specifically to avoid "wasting" the loss already showing
1
the number of questions that should decide whether to hold a position: does it still qualify for entry from here, not how much has already gone into it
What the Sunk Cost Fallacy Looks Like in a Trading Journal
The sunk cost fallacy hides behind language that sounds like patience or conviction, so it is easiest to catch by comparing what actually happened at the stop-loss level against what happened after it.
| Metric | Sunk Cost Fallacy Present | Healthy Process |
|---|---|---|
| Behavior at the original stop-loss | Stop moved, widened, or ignored to avoid locking in the loss | Exit taken at the pre-defined level regardless of how it feels |
| Reason given for holding | "I've already lost so much, may as well see it through" | A specific, current reason the setup still holds, independent of the loss so far |
| Strategy funding decisions | Continued because of time and money already spent building it | Based on live, out-of-sample results measured against a pre-set threshold |
| Averaging down | Used to lower cost basis and rescue the position already taken | Used only when the plan defines it in advance, with a hard size limit |
This is the same pattern that turns a single bad trade into a blown account — see loss aversion for how the fear of realizing a loss compounds with the reluctance to walk away from one already taken.
How to Spot It in Your Own Data
- 1Pull every trade where the exit came significantly later than the original stop-loss level, and check whether the delay was justified by new information or just by reluctance to lock in the loss
- 2Compare the average size of losses that were held past their stop to the average size of losses taken exactly at the stop — a large gap points to sunk cost thinking
- 3Review journal notes for phrases like "already down this much," "might as well wait it out," or "too far in to quit now" around exit decisions
- 4Check whether any strategy has kept getting capital or screen time specifically because of how long it's been worked on, rather than because of its live results
- 5Look at averaging-down entries and see how many happened outside the original trading plan, purely to lower the cost basis on a losing position
Fixing It: Make the Exit Decision Before the Cost Exists
The sunk cost fallacy is hardest to fight in the moment, because by the time it activates, the cost already feels too large to walk away from. The fix, as with loss aversion and the gambler's fallacy, is to remove the in-the-moment decision entirely by setting the exit before any cost has accumulated.
- Set the stop-loss at entry and treat moving it further away as against the rules, not as a judgment call to revisit once the trade is losing
- Use hard stop orders where the platform allows it, so the exit doesn't depend on willpower once the position is underwater
- Evaluate every open position with a simple test: "If I had no position right now, would I enter this trade at the current price?" — if not, the existing loss is not a reason to stay
- Set a pre-defined, measurable threshold for when a strategy gets shut down (e.g., max drawdown or a losing-trade count), decided before live trading starts, not while sunk costs are mounting
- Log every instance of a stop being moved or widened mid-trade as its own tag, so the pattern is visible in the data instead of buried in one-off excuses
TradeEdge flags trades that closed well past their original stop-loss level, so a pattern of chasing sunk costs shows up in the data instead of staying hidden in each trade's excuse.
The sunk cost fallacy is expensive precisely because it feels like commitment rather than error — staying in "to see it through" sounds like discipline, not a mistake. But a loss already taken has no bearing on whether the trade is still worth holding, and the traders who protect their accounts are the ones who decided, before any cost existed, exactly when they would walk away.