The Sunk Cost Fallacy Is Why You Can't Sell — Here's How to Actually Fix It

“I’ve already put in $40,000, I can’t sell now” is one of the most common sentences in trading and investing, and it’s almost never followed by a good decision. It sounds like conviction. It sounds like patience. What it actually is, most of the time, is the sunk cost fallacy — a well-documented reasoning error where money, time, or effort you’ve already spent keeps influencing a decision that should only depend on what happens from here.

The fallacy has a name because it’s universal, not because it’s rare or a sign of being a bad investor. It shows up in academic research on decision-making, in corporate boardrooms deciding whether to keep funding a failing project, and in everyone who has ever finished a bad meal because they’d already paid for it. It shows up in trading accounts constantly, because trading provides a number — your entry price — that makes the sunk cost impossible to ignore even when it should be irrelevant.

The mechanic, stripped down

A cost is “sunk” once it’s spent and can’t be recovered, regardless of what you do next. The $40,000 in the example above is sunk the moment it’s invested — selling doesn’t spend it again, and holding doesn’t get it back. Rational decision-making says the only things that should matter to your next move are the future costs and future benefits of each option available to you now. What you already spent isn’t one of those things, because it’s identical no matter which option you pick.

The fallacy is weighing that already-spent amount anyway — usually as a reason to keep going, on the logic that quitting now would make the original expenditure “wasted.” It wouldn’t. The money is already spent, or already lost, in every branch of the decision tree. Selling doesn’t create the loss; it was created the moment the price dropped. All selling does is stop deferring a decision about what to do with what’s left.

Why it feels like conviction instead of a bias

The sunk cost fallacy is uncomfortable to recognize in the moment because it borrows the language of legitimate reasons to hold. “I believe in this” and “I’ve put in too much to quit now” produce the same felt sense of resolve, even though only one of them is actually about the asset’s future. A genuine thesis is about what happens next: adoption, earnings, a catalyst, a valuation gap. A sunk-cost justification is about what already happened: your entry price, your paper loss, how long you’ve held.

The giveaway is what the reasoning references. If you can explain why you’re holding without mentioning what you paid, how long you’ve held, or how much you’ve lost — using only information about the asset’s current prospects — you likely have a real thesis. If the explanation leans on the amount already committed, you’re rationalizing a decision that’s actually being driven by the discomfort of locking in a loss, not by an assessment of what’s likely to happen next.

A worked example

Take two versions of the same position, to make the asymmetry concrete.

Version A: You bought an asset at $100. It’s now at $60. You still believe the underlying thesis — the reasons you bought it — are intact, and if you didn’t already own it, you’d buy it today at $60 based purely on where you think it’s headed from here. That’s a coherent reason to hold or even add. The prior loss isn’t doing any work in that reasoning; the current price and forward view are.

Version B: Same asset, same price history. But if you ask yourself honestly whether you’d buy it today, brand new, at $60 with no prior position — the answer is no. You wouldn’t initiate this position at today’s price on today’s information. The only reason you’re holding is that selling would mean realizing a $40 loss you’ve been avoiding. That’s the fallacy. The position isn’t being held because it’s a good idea now; it’s being held because closing it would force you to admit it wasn’t.

The test that separates them — “would I buy this today, at this price, if I didn’t already own it?” — is the single most useful question in this entire framework, because it forces your reasoning to run forward instead of backward. It’s the same underlying test used in Down 50%? Here’s the Actual Plan to decide between holding, averaging down, and cutting a loss, because sunk cost is usually the hidden variable making that decision feel harder than it is.

Where it quietly does the most damage

A few specific patterns are worth watching for, because they’re where sunk cost operates without announcing itself:

Averaging down to “lower the average” rather than because the thesis improved. Adding to a losing position can be rational if the setup genuinely got better at the new price. It’s sunk cost in disguise when the actual motivation is making the loss on paper feel smaller by changing the reference point, not because you’d have sized in at this level from scratch.
Refusing to sell a specific lot because it “hasn’t recovered yet.” Tax lots are fungible for decision-making purposes even though they matter for accounting. A lot bought at a worse price isn’t owed a recovery before you’re allowed to sell it — that’s treating the purchase price as a debt the market owes you, which it doesn’t.
Sticking with a strategy or system after it stops working, because of time already invested building it. This is the non-financial version. Months spent developing an approach can make it harder to abandon even after the evidence says it isn’t working, for exactly the same reason a losing position is hard to sell.
Holding a “story” position past the point the story changed, because you’ve already defended it publicly or to yourself for a long time. Identity investment is a sunk cost too — abandoning a position you’ve argued for feels like abandoning being right, which adds social and ego cost on top of the financial one.

A decision process that removes the trap

State the current thesis in one sentence, using only present-tense, forward-looking language. No mention of entry price, holding period, or amount invested allowed. If you can’t do this, you likely don’t have a live thesis.
Ask the buy-today test. Would you initiate this exact position, at today’s price, with today’s information, if you didn’t already hold it? A genuine “yes” is a reason to hold or add. A “no” followed by holding anyway is the fallacy operating in real time.
Separate the decision from the emotion by writing it down before you act, the same way a stop-loss rule works by being set before you’re under pressure. Decisions made in the moment, while a paper loss is fresh, are the ones most vulnerable to sunk cost reasoning; decisions made against a pre-written rule are much less so.
If you decide to exit, treat it as closing a decision, not admitting a mistake. The mistake, if there was one, already happened at the point of entry or at the point the thesis broke. Selling is just the point where you stop deferring the consequence — it doesn’t add a second mistake on top of the first.

None of this requires being right about where the asset goes next. It just requires making sure the decision is actually about that, and not about money that’s already gone regardless of what you choose.

FAQ

Is the sunk cost fallacy the same thing as loss aversion? They’re related but distinct. Loss aversion is about how a loss feels worse than an equivalent gain feels good, which drives urgency to avoid realizing one. The sunk cost fallacy is specifically about letting money, time, or effort already spent — which you can’t get back either way — influence a decision that should only depend on what happens next. Loss aversion explains why the sunk cost feels so relevant; the fallacy is treating it as relevant anyway.

Isn’t it reasonable to give a position more time if I’ve researched it thoroughly? Yes — genuine conviction based on a thesis that’s still intact is a legitimate reason to hold. The test isn’t whether you’ve done research; it’s whether that research still supports the position today, independent of what you’ve already spent or lost. If your reasoning for holding keeps citing your entry price or how much you’ve put in rather than what the asset is likely to do from here, that’s the fallacy talking, not the research.

Does the sunk cost fallacy only apply to money? No — it applies just as strongly to time and effort. Someone who has spent months building a trading strategy that isn’t working, or years following a project they were early to, faces the same pull to keep going specifically because of what they’ve already invested, not because of what continuing is actually likely to produce.

How is this different from just having a long time horizon? A long time horizon is a plan you set in advance, independent of how the position has performed since. The sunk cost fallacy shows up after the fact, as a reason invented to justify not selling something that’s already lost value. A genuine long-term holder would size and hold the same position the same way whether it were up or down; if your holding period seems to stretch out specifically because you’re down, that’s the tell.