What is the sunk cost fallacy?

A sunk cost is any money, time, or effort that has already been spent and cannot be recovered, no matter what you decide next. The sunk cost fallacy is letting that unrecoverable past cost influence a decision that should be based only on two things: what you expect to happen from here, and what else you could do with your remaining time or capital instead.

The fallacy shows up in three characteristic phrases:

• Backward framing. The justification references what has already been spent — "I'm already down 40%" — instead of what lies ahead.
• Escalation. Rather than cutting exposure, the instinct is to add more, on the logic that doing so improves the average or "proves" the original call.
• Deferred admission. Selling or stopping gets pushed to some future condition — breakeven, a specific date, a recovery — rather than being evaluated on today's facts.

In plain terms: the fallacy is not that holding is always wrong. It is that the reasoning behind holding is being pulled from the wrong ledger.

The classic framing comes from economics — sunk costs should be irrelevant to rational forward-looking decisions, since they cannot be recovered either way. But the fallacy is a psychological one, not a math error. People understand the logic perfectly well in the abstract and still act against it when it's their own money or their own prior decision on the line.

Researchers including Hal Arkes and Catherine Blumer demonstrated this experimentally in the 1980s: people persisted with a costly course of action specifically because they had already invested in it, even when a fresh option was objectively better — a pattern that held regardless of whether the original investment was made by the subject or simply described to them.

Daniel Kahneman and Amos Tversky's prospect theory offers the underlying mechanism: losses are felt roughly twice as intensely as equivalent gains, so realizing a loss (by selling, quitting, or writing something off) triggers a sharper emotional response than continuing to hold does — even when continuing to hold is the objectively worse choice.

Further reading: Arkes & Blumer's original sunk cost research.

The pattern is easiest to see in someone else's decision. These are the shapes it takes most often:

• Trading

  A position is down significantly and the plan becomes "wait until I'm back to breakeven, then reconsider." The exit condition is anchored to the entry price, not to any change in the underlying thesis.

• Investing

  An investor keeps adding to a losing position "to lower my average," treating the original entry price as a target to defend rather than a number with no bearing on future returns.

• Startups

  A founder keeps funding a product with years of sunk development time, even after user data clearly shows no traction, because stopping would mean the time already spent "meant nothing."

• Hiring

  A manager keeps an underperforming hire on the team well past the point of clear signal, reasoning that the onboarding investment and ramp-up time would be wasted by starting over.

• Relationships and commitments

  Someone stays in an arrangement — a lease, a partnership, a long project — specifically because of how much time has already gone into it, independent of whether it still makes sense today.

• Personal finance

  A subscription, course, or gym membership goes unused but stays paid for, because canceling would mean "admitting" the earlier purchase was a waste.

Three forces keep it in place. Loss realization: selling, quitting, or writing something off makes the loss concrete and final, which is more painful than an ongoing, unrealized loss — even if the ongoing loss is objectively larger. Identity: reversing a prior decision can feel like admitting the original judgment was wrong, and that admission carries a social and self-image cost that has nothing to do with the numbers. Narrative completion: humans are drawn to finishing what they started; walking away mid-story feels unresolved in a way that continuing, even badly, does not.

This is why "just be rational about it" rarely works as a fix. The pull operates on the discomfort of stopping, not on a miscalculation that better math alone can correct.

In a trading account, sunk cost fallacy rarely shows up as one catastrophic decision — it shows up as a losing position that keeps getting more room, more patience, and more capital, purely because of what has already gone into it. The account bleeds slowly rather than all at once, because each individual decision to keep holding feels smaller than the cumulative one.

It compounds with other biases in predictable ways. Sunk cost reasoning borrows loss aversion's emotional charge to feel like patience, and it borrows confirmation bias's selective evidence-gathering to feel like conviction. A position held for sunk cost reasons often looks, from the inside, exactly like a position held with genuine thesis — which is exactly what makes it dangerous.

You cannot remove the pull entirely, but you can build structure that catches it before it drives a decision:

1. 1

   Ask the zero-based question

   "If I didn't already hold this position, would I open it today, at today's price?" This strips out the entry price entirely and forces a fresh, forward-looking answer.

2. 2

   Separate the decision from the ledger

   Evaluate what to do next based only on current information and future expectations. What has already been spent is a historical fact, not an input to the decision in front of you.

3. 3

   Set the exit condition before you enter

   Decide in advance — before any emotional stake builds up — what price, date, or change in facts would make you exit. A plan made in advance is much harder for sunk cost reasoning to override later.

4. 4

   Imagine advising someone else

   Ask what you would tell a friend holding the identical position with the identical reasoning, if it weren't your money on the line. People are consistently more rational giving advice than taking it.

5. 5

   Watch your own language

   Phrases like "already invested," "come this far," or "can't back out now" are markers that the ledger being consulted is the past, not the future.

6. 6

   Write the decision down before you act

   A decision journal exposes the shape of your reasoning in your own words. Sunk cost reasoning has a distinct fingerprint, and it becomes far easier to recognize once you've seen it written out once.

• Why day traders lose money

  The hub piece: why "wait until it comes back" and six other biases, trade by trade, are the real reason most day traders lose money.

• Loss aversion

  Losses feel roughly twice as heavy as equivalent gains, which is often the emotional engine behind sunk cost reasoning.

• Confirmation bias

  Selectively favoring evidence that supports a decision already made, which makes a sunk-cost position feel justified rather than stubborn.

• Anchoring bias

  Judging value against an irrelevant reference point — often the exact entry price sunk cost reasoning is defending.

• Overconfidence

  Excess certainty that shrinks the perceived need for an exit plan in the first place.

The Bias Checker screens for all of them at once, and the articles go deeper on each.