The sunk cost fallacy in investing is the tendency to keep putting money into a stock, a fund, or a business idea because of how much you've already spent on it, rather than because of what it's actually worth going forward. The dollars you've already committed are gone either way, whether you buy more shares or walk away completely, so they shouldn't carry any weight in your next decision. Investors let them anyway, buying more of a falling stock to "lower the average" or avoid admitting the first purchase was a mistake, and that single habit is one of the most expensive mental shortcuts in a portfolio.
What Is the Sunk Cost Fallacy, Exactly?
It's the bias of letting a past, unrecoverable cost influence a decision that should only depend on future outcomes. In economics, a sunk cost is any expense you can't get back no matter what you choose next, so a rational decision-maker is supposed to ignore it entirely and weigh only what happens from here. Investors routinely do the opposite: a trader who's down $4,000 on a stock asks "how do I get my $4,000 back" instead of "is this stock, at today's price, still a good place to put new money." Those are different questions with different answers, and only the second one is actually answerable by looking at the company or asset itself.
Is Averaging Down the Same Thing as the Sunk Cost Fallacy?
No, not inherently, and the difference comes down to why you're buying. Averaging down, buying more shares of something you own after its price drops, is a legitimate move when it's based on a fresh look at the asset's current price and prospects, the same way a new investor with no prior position would evaluate it today. It becomes the sunk cost fallacy the moment the reason shifts to "I already have so much invested, I need to lower my cost basis" or "I can't sell at a loss now." A simple test cuts through the confusion: would you buy this position today, from zero, at today's price, knowing what you know right now? If yes, buying more is a normal investment decision. If the honest answer is only "because of what I've already put in," that's the fallacy talking, not your analysis.
How Much More Do You Actually Need the Stock to Gain After Averaging Down?
Here's the real math, worked out. Say you buy 100 shares at $100, for $10,000. The stock drops 30% to $70, and your position is now worth $7,000, a $3,000 paper loss. If you do nothing, that stock has to climb back from $70 to $100, a 42.9% gain, just to get you to breakeven. If instead you average down and buy 100 more shares at $70 for another $7,000, your total cost is $17,000 for 200 shares, an average cost of $85. Now the stock only needs to climb from $70 to $85, a 21.4% gain, to get the whole position back to even. That's the real pull of averaging down: the breakeven math genuinely gets easier.
But look at what you actually did: you doubled your dollar exposure to a stock that, by definition, has been moving against you, and your total capital at risk went from $10,000 to $17,000. If the stock keeps falling instead of recovering, say it drops another 30% to $49, the original 100-share position would be worth $4,900, a 51% loss on your initial $10,000. The averaged-down position is worth $9,800 against a $17,000 cost basis, a 42.4% loss in percentage terms, but a $7,200 dollar loss instead of $5,100. You bought more of something that kept losing, and the fallacy is exactly what made that purchase feel like progress instead of what it was: a bigger bet on a thesis the market had already started rejecting twice.
What Does the Research Actually Say About Sunk Costs and Investors?
The strongest real-money evidence comes from corporate acquisitions, not retail trading, but the mechanism is the same one that shows up in a brokerage account. A 2025 Journal of Finance study by Marius Guenzel, "In Too Deep," examined U.S. public company stock acquisitions from 1980 to 2016 and used market-driven price swings in fixed-exchange-ratio stock mergers as a natural experiment for cost that had nothing to do with the deal's actual quality. An interquartile increase in that quasi-random acquisition cost reduced the odds of the company later divesting the acquisition by 8 to 9%, and the effect was strongest when the same CEO who approved the original deal was still in charge, consistent with a personal attachment to a cost already paid rather than a purely institutional one. Guenzel's own suggested fix for corporate boards, having someone other than the person who made the original purchase decide whether to sell, is a useful standard for a personal portfolio too.
The picture gets more nuanced in controlled lab settings. A 2020 CESifo working paper by Negrini, Riedl, and Wibral found that studies asking people to react to hypothetical scenarios reliably produce the sunk cost bias, but their own incentivized experiment, where real money was on the line instead of a hypothetical survey question, found no sunk cost bias at all, and even a reverse effect: people in a loss position actually invested less, not more. That's a real and useful caveat. It suggests the sunk cost fallacy isn't an iron law of human behavior, it's a bias that gets stronger when a decision feels emotional, social, or ego-driven (a trade everyone can see, a deal with your name on it) and weaker when the stakes are cold and immediate. A separate 2019 study by Tait and Miller, testing 168 participants, specifically examined whether loss aversion is what drives the effect, reinforcing that the two biases are tangled together rather than separate forces.
How Do You Avoid Falling for the Sunk Cost Fallacy With Your Portfolio?
Run every "should I add more" decision through a short checklist instead of your gut. First, ask whether you'd buy this position today, from scratch, at today's price, with zero shares already owned, and answer honestly instead of defensively. Second, keep scheduled investing separate from reactive investing: contributing to your 401(k) or buying a broad index fund on a fixed schedule was decided before any single stock's price drop and shouldn't change because of one, while a purchase triggered by "I need to lower my average" was decided by the price drop itself. Third, write your investment thesis down before you buy the first share, and when you're deciding whether to add more, reread that thesis rather than your current loss. Fourth, set position-size limits while you're calm, not after you're underwater, since a limit decided in the moment tends to move to fit whatever you already did. And if you genuinely can't tell your own motive, ask someone with no stake in your being right, the same logic Guenzel found actually works at the corporate level.
This bias runs on the same fuel as a couple of others worth knowing. The price you originally paid becomes a reference point that distorts how you see the stock today, which is exactly what we cover in our piece on anchoring bias in investing. And the broader pattern of holding losers too long while selling winners too early, which sunk cost thinking feeds directly, is the subject of our breakdown of the disposition effect.
—
Spicy Investing
