The Sunk Cost Fallacy in Investing: 5 Patterns Quietly Costing the Average Portfolio 1.5% a Year

The average investor holds losing stocks roughly 20% longer than winning stocks — about 124 days versus 104 days, according to Terrance Odean’s landmark 1998 study of 10,000 brokerage accounts. The kicker: the stocks they refused to sell went on to underperform the ones they did sell by about 3.4 percentage points over the following year. That’s not a rounding error. That’s the sunk cost fallacy in investing quietly draining returns from millions of portfolios every year.

The reason it happens isn’t ignorance. Most investors who hold a losing stock too long can recite the textbook rule perfectly: past purchase price is irrelevant; only future expected return matters. They still hold. The sunk cost fallacy in investing is one of the most thoroughly documented behavioral patterns in finance, and one of the most expensive to ignore.

What the Sunk Cost Fallacy in Investing Actually Looks Like

The classical sunk cost fallacy says past, unrecoverable costs should not influence future decisions. In a brokerage account, that translates very simply: the price you paid for a stock is irrelevant to whether you should hold it tomorrow. The only question that matters is the same one you’d ask if someone handed you cash equal to today’s market value — would you buy this same position with that cash right now?

If the honest answer is no, you are holding because of what you paid, not because of what the position offers from here. That’s the disposition effect — the academic name for the sunk cost fallacy in investing — and it shows up in three predictable patterns:

• Holding losing positions waiting to “break even.” The price you need for break-even is a fact about your past, not the company’s future.
• Selling winners too early. The same bias works in reverse — locking in small gains so the position can’t “go back down” and erase what you already “made.”
• Refusing to rebalance away from a concentrated legacy holding. Inherited shares, employer stock, or an early winner that grew to 30% of net worth — kept because you’ve “owned it forever” rather than because it still fits the allocation.

Vanguard’s Advisor’s Alpha research attributes roughly 150 basis points (1.5%) per year of incremental return to “behavioral coaching” — the part of an advisor’s job that involves stopping clients from doing things exactly like this. Over a 30-year horizon, 1.5% per year on a $200,000 starting balance is the difference between roughly $1.5 million and $2.4 million at a 6% real return. The fallacy isn’t a small leak.

The Real-Money Cost: What 1.5% per Year Compounds Into

Behavioral mistakes feel small in the moment — one position held a few months longer than it should have been, one rebalance skipped, one concentrated holding never trimmed. The compounding math is what makes them expensive. Here’s what the gap looks like across three realistic starting balances and a 30-year horizon, comparing a disciplined portfolio earning 6% real to one losing 1.5% per year to behavioral drag:

Those numbers assume only the Vanguard-estimated 1.5% behavioral drag. DALBAR’s annual Quantitative Analysis of Investor Behavior has historically reported even wider gaps between average equity-fund investor returns and the S&P 500 itself — often 2 to 3 percentage points per year over rolling 20- and 30-year windows, depending on methodology. The methodology is contested (some academics argue DALBAR overstates the gap), but the directional finding — that investor behavior subtracts from index returns — has been replicated across multiple datasets.

Why Your Brain Won’t Let You Sell a Loser

The disposition effect is one of behavioral finance’s best-documented results because it shows up across cultures, account sizes, and asset classes. There are three reinforcing mechanisms at work, and any one of them is strong enough to override the rational case for selling.

1. Loss aversion. Kahneman and Tversky’s prospect theory work demonstrated that losses feel roughly twice as painful as equivalent gains feel pleasurable. Selling a loser realizes the loss — converts a paper loss into a real, accounted-for, irrevocable one. Holding lets you keep the loss in a quantum state where it might still resolve in your favor. The math of expected return doesn’t care about that distinction. Your nervous system absolutely does. (We’ve covered the broader pattern in our piece on how loss aversion affects budgeting — the investing version is the same wiring applied to a different account.)

2. Identity defense. If you bought a stock based on a thesis you researched and posted about, selling it concedes you were wrong. For many investors — especially the ones who are most active and most confident — that concession is more uncomfortable than the financial loss. The position is no longer just a position; it’s a referendum on the investor’s judgment.

3. The “house money” inversion. Mental accounting separates “principal” from “gains.” A losing stock has eaten into principal — selling is psychologically more painful than selling a winner, even when the dollar amount and the future expected return are identical. The endowment effect makes this worse: once you own the share, you implicitly value it above what you’d pay to buy the same share today. Our deep dive on the endowment effect goes through the experimental evidence on roughly how much — typically 30 to 100 percent above fair market value.

Five Sunk Cost Fallacy in Investing Patterns That Quietly Cost the Most

Some sunk-cost-driven holdings are small enough to be rounding errors. A few patterns reliably cost real money. These are the ones most worth auditing first:

Pattern 1: The “It Was Up 40% Once” Concentrated Position. A single stock you bought early, watched run, never trimmed, and now refuses to sell because it briefly traded much higher. The peak price has become the anchor. The fact that you could buy the same shares today at the current price and choose not to is the relevant data; the peak is sunk. Single-stock positions concentrated above 5-10% of a portfolio carry idiosyncratic risk that index-fund investing was specifically designed to eliminate. (See our three-fund portfolio guide for the alternative.)

Pattern 2: The “Just Need to Get to Break-Even” Stock. A position down 30-40%. The investor will not sell at the current price but says they’ll “definitely sell when it gets back to my cost basis.” Break-even isn’t a financial event — it’s a psychological one. The market does not know what you paid. The stock’s expected return from today’s price is identical whether you paid double or half.

Pattern 3: The Inherited Stock. Often the worst offender, because the cost basis stepped up at death — meaning the tax cost of selling is much lower than people assume. Despite this, inherited concentrated positions get held for decades out of a vague sense that selling would be disrespectful. The relevant question is whether the deceased person would have bought that same allocation for the heir’s circumstances today. Usually no.

Pattern 4: The Employer Stock Overweight. Common when an employer match comes in shares or when ESPP/RSU income accumulates faster than the investor diversifies it. The sunk cost feel is “I earned this stock, selling it feels like undoing my work.” Financially, holding more than 5-10% of net worth in employer stock concentrates two risks — your paycheck and your portfolio — in the same company. The 2001 collapse of Enron’s employee 401(k) plans, where company stock was over 60% of average plan assets, is the canonical cautionary tale.

Pattern 5: The Active Fund You’ve Held for a Decade. The expense ratio is 1.1%. The fund has underperformed its benchmark for the last seven years. The investor has held it since the early 2010s and “doesn’t want to give up after waiting this long.” S&P’s SPIVA scorecard consistently shows that over 15-year periods, roughly 85-90% of US large-cap active funds underperform their benchmark. Waiting longer doesn’t change the underlying odds. The years you’ve already paid the fee are sunk.

The Replacement Test: A Decision Rule That Defeats the Bias

The technique that actually works against the sunk cost fallacy in investing is mechanical, not motivational. The goal is to design a process that makes the decision without consulting your feelings about the position. Here’s the one I use in my own portfolio.

For every position above a meaningful threshold (I use 5% of portfolio value), I write down — in advance, in a notes file — the answer to a single question: If this position were liquidated to cash today, would I buy it back at the current price with the proceeds?

• Yes: Hold. The position still earns its place.
• No, but I’d buy something similar in the same category: Sell and rotate into the better option. The bias is brand-loyalty, not investment thesis.
• No, I’d reduce my equity exposure here: Sell to whatever target weight passes the replacement test. Possibly zero.

I started running the replacement test on my own holdings a few years back, mostly out of curiosity about whether the much-praised approach actually moved the needle. The honest answer: yes, but less than personal finance Twitter implies. The biggest single benefit wasn’t the trades I made — it was the trades I didn’t make. Forcing myself to ask the question every quarter killed three or four “I’ll just hold it until it recovers” rationalizations that, in hindsight, would have cost me the difference between flat returns and small losses on multi-thousand-dollar positions.

As a software engineer with a DIY approach to my own investing — index funds, tax-advantaged accounts, no advisor — I lean naturally toward systems that take judgment out of single-instance decisions. The replacement test is one of the few behavioral tools I’ve found that survives the test of actually doing it for years rather than reading about it once.

When the Sunk Cost Instinct Is Actually Right

Worth a caveat: not every instinct to hold is sunk cost fallacy in disguise. There are real situations where the “feels like sunk cost” instinct is actually pointing at something legitimate.

Tax cost of selling can dwarf behavioral cost. A position with a very large embedded long-term capital gain — say, an early Apple or Amazon purchase up 800% — may rationally be worth holding even if the position is overweight, because realizing the gain triggers a 15-23.8% federal long-term capital gains tax (plus state). For an investor planning to use a step-up in basis at death or to donate appreciated shares to charity, holding has a clear after-tax case that has nothing to do with sunk cost.

The pattern-seeking instinct can also misfire in the other direction. Selling losers is sometimes the right call and sometimes a separate cognitive trap — the gambler’s fallacy, where investors abandon a temporarily underperforming position assuming the streak “must end soon.” We covered that mirror-image bias in our piece on the gambler’s fallacy in investing. The replacement test cuts through both. It doesn’t ask whether the position will recover. It asks whether you’d buy it today.

Transaction costs and bid/ask spreads on illiquid holdings. For thinly traded securities, the spread itself can be 1-3% of the position. If the expected difference in future return is smaller than the round-trip transaction cost, holding may be rational even when the position fails the replacement test on pure fundamentals.

None of these exceptions change the central claim. The default human pattern is to hold losers too long and sell winners too early, and the cost compounds. The exceptions are real but narrow — and notably, none of them are “I just need to wait until I’m back to break-even.” The complementary piece, our shorter sunk cost trap guide, walks through the same fallacy applied to subscriptions and recurring services rather than portfolio holdings — a useful companion read for the non-investment side of the same bias.

Key Takeaways

Photo by Maxim Hopman on
Unsplash

Written by Chris Steve

Chris Steve is a software engineer with a deep interest in personal finance, behavioral economics, and AI. He started Money & Planet to share clear, research-backed money guides — the kind that explain the math instead of pushing products. His writing focuses on long-term wealth building, the psychology behind spending and investing decisions, and the practical tools regular people can use to make smarter financial choices.