Sunk Cost And The Position You Cannot Let Go
Money already lost on a position is gone regardless of what happens next, yet it routinely shapes what a trader does going forward — the sunk cost fallacy is treating a past, unrecoverable loss as a reason to keep committing more capital or attention to the same trade.
A position is down significantly, and the reasoning for staying in — or worse, adding to it — starts to sound like "I've already put so much into this, I can't walk away now." That reasoning is the sunk cost fallacy, and it has a specific, identifiable flaw: the money already lost is gone no matter what happens from this point forward, so it has no legitimate bearing on the decision of what to do next. The only thing that should matter going forward is the position's expected future performance from here, evaluated fresh, and that has nothing to do with how much has already been lost getting to this point.
Why past losses feel relevant when they aren't
The trap works because "I've put so much in already" feels like a reason connected to the trade, when it's actually a reason connected only to the trader's own history with the trade — the market doesn't award better odds to a position because more capital has already been committed to it, and a fresh dollar allocated to that position competes on exactly the same footing as a fresh dollar allocated anywhere else. Averaging down into a losing position is the classic version: buying more of something at a worse price specifically because a lot has already been spent on it, framed as "improving the average cost," when the improved average cost changes nothing about whether the position itself is still a good idea from here.
A trader who had put on a large position in a merger-arbitrage spread watched the deal's completion probability deteriorate as regulatory opposition mounted, with the position now down substantially. Rather than reassessing the deal's odds fresh, the trader added to the position at the wider spread, reasoning that "there's too much on the line to walk away without seeing this through." The deal broke a month later and the additional capital added at the wider spread lost more than the original position alone would have. A fresh assessment of the deal's odds at the time of averaging down — ignoring the amount already committed — would have shown the same deteriorating picture that made the position a poor one to be adding to, sunk cost or not.
The test that cuts through sunk cost is the same forward-looking question used elsewhere: given only what's true now, is this still a position worth holding or adding to, completely independent of how much has already been lost getting here? If the honest answer is no, the size already committed doesn't change it.
Sunk cost is money already spent that cannot be recovered regardless of what happens next, and it has no legitimate bearing on a forward-looking decision — yet "I've already put so much in" routinely gets treated as a reason to hold or add to a losing position. Averaging down specifically to "fix the average price" is the classic version, and the test against it is asking whether the position is worth holding based purely on what's true from here.
Further reading
- Arkes and Blumer, The Psychology of Sunk Cost