What a sunk cost actually is
A sunk cost is any resource (money, time, effort) you've already spent and can't recover, whatever you decide from here on out. Classical economic theory is blunt about it: a sunk cost is irrelevant to any future decision, because that decision should only depend on the costs and benefits still ahead, not the ones already behind you.
The problem is that human beings, systematically and predictably, don't reason that way. Psychologists Hal Arkes and Catherine Blumer formally demonstrated this bias in a 1985 study: in one of their experiments, people who had paid more for a non-refundable ski trip ticket were more likely to go, even when told that an alternative (cheaper) trip would objectively be more enjoyable given the weather conditions.
The Concorde effect: when a cognitive bias gives its name to an airplane
The Concorde was a supersonic passenger jet jointly developed by France and the United Kingdom, whose development cost far exceeded every initial projection. Even in the project's early phases, back in the 1960s, internal reports seriously questioned its commercial viability: the plane burned fuel with extraordinary inefficiency and its passenger capacity was too small to cover its operating costs.
Despite those early warning signs, both governments kept investing billions for years, largely because canceling the project would have meant publicly admitting that all the money already spent was lost for good. The Concorde finally entered service in 1976 and operated for 27 years, never achieving sustained profitability, until its final retirement in 2003. The case is so widely cited in behavioral economics today that the sunk cost fallacy itself is also known as the 'Concorde effect.'
An everyday example: the stock you 'can't' sell
imagine you bought shares of a company for $10,000 and they're now worth $4,000. Analyzing the company today, with current information, you conclude its prospects are poor and that money would do better elsewhere. The purely rational decision would be to sell and reinvest that $4,000 wherever it performs best, with the original $10,000 playing no role whatsoever in that decision: that money is already spent, whether you 'lost' it in the trade or not.
The sunk cost fallacy, however, pushes you to reason backward: 'if I sell now, I'll have lost $6,000 for good; if I hold on, there's still a chance of getting it back.' That reasoning conflates two completely different questions: how much you've already lost (irrelevant to the future) and what's the best thing to do with the money you have left (the only question that matters).
Why this bias is so hard to beat
Part of the sunk cost fallacy's power comes from how it intertwines with the loss aversion we covered in the previous chapter: selling an investment that has fallen forces you to convert a loss 'on paper' (which your brain can keep postponing as long as you don't sell) into a realized, final loss. There's also a social dimension: admitting a past decision was a mistake can feel like admitting you personally got it wrong, something many people instinctively avoid even when no one else is watching.
The larger the cost already invested, the stronger this effect tends to be, exactly the opposite of what economic logic would suggest (the more money at stake, the more careful the analysis should be, not less).
The question that neutralizes this bias
There's a relatively simple mental question to neutralize this bias at the moment of deciding: 'if I didn't already own this investment and had this money in cash in my hand today, would I buy it again with the information I have now?' If the answer is no, the fact that you already own it changes nothing: you should sell it just the same, exactly as you wouldn't buy it from scratch.
This same question applies beyond pure investing: to a subscription you no longer use, a business that keeps losing money, or any decision where 'I've already invested too much to stop now' is the main argument for continuing. Almost always, that argument is the sunk cost fallacy talking, not a real analysis of what makes sense to do from today forward.
Apply This Knowledge to Your Real Finances
eXcenda combines net worth, expenses, simulations, and macro context into a single app so you can move from theory to action.
Frequently Asked Questions
It's any resource (money, time, effort) that has already been spent and can't be recovered, regardless of the decision made from here on. Economic theory holds that a sunk cost shouldn't influence any future decision.
Because the French and British governments kept funding the Concorde supersonic jet for years despite early signs of its commercial unviability, mainly because they had already invested colossal sums and canceling it would have meant admitting that loss.
A clear sign is justifying continuing with something by saying 'I've already invested too much to stop now' instead of evaluating whether, with current information, you'd make that same decision from scratch.
They're related but not identical: loss aversion is the psychological pain of losing, while the sunk cost fallacy is specifically the flawed reasoning of letting what's already been invested in the past shape a decision that should only depend on the future.
No, it shows up in any area: relationships, professional projects, subscriptions, or businesses. The common denominator is always the same: continuing to invest resources in something because of what's already been invested, not because of what's expected in the future.