The Sunk Cost Fallacy: Why You Keep Making Bad Decisions

You're at a restaurant. The food is terrible. Your plate is half full. But you keep eating anyway — because you paid $28 for it and it feels wrong to waste it.

That $28 is gone whether you finish the meal or not. Eating more of something you don't enjoy doesn't get your money back. It just makes you feel worse. Yet we do this constantly — not just with food, but with investments, relationships, careers, and entire life paths.

This is the sunk cost fallacy, and it's probably costing you thousands of dollars a year.

What Actually Happens in Your Brain

Economists have a term for this: loss aversion. Research by Daniel Kahneman and Amos Tversky (who won a Nobel Prize for this work) found that losses feel about twice as painful as gains feel good. Losing $100 feels as bad as finding $200 feels good.

So when you've invested time, money, or effort into something, walking away feels like admitting that investment was lost. Your brain would rather keep going — even into a bad situation — than accept the loss. It's not rational. It's emotional. And it's incredibly hard to override.

The Financial Sunk Costs That Are Killing You

This isn't just about bad restaurant meals. The sunk cost fallacy shows up in much more expensive ways:

Holding losing stocks. You bought a stock at $50. It's now at $20. Every day you hold it is a day you could've invested that money somewhere else. But you keep hoping it'll bounce back because "I can't sell at a loss." The $30 drop already happened. Selling doesn't make it worse. Holding a bad investment for emotional reasons does.

Finishing a degree you hate. Two years into a career path you realize is wrong. "I've already spent so much time and money — I can't quit now." So you finish the degree, get the job, and spend the next 40 years doing something you despise because the first two years felt too expensive to waste.

Keeping a car that's falling apart. "I just put $2,000 into repairs, I can't sell it now." Then you put in another $1,500. Then another $800. The total repair cost exceeds the car's value, but you keep going because you've already "invested so much." A buddy of mine spent $9,000 over two years keeping a $4,000 car on the road. When he finally sold it, he got $800. Total loss: $12,200. A reliable used car would've cost $8,000 and zero headaches.

Staying in a bad subscription or membership. You signed up for a 12-month gym contract. You've gone twice in 4 months. But you keep paying because "I already paid for the year." Cancel the contract. The money's gone. Stop paying for the privilege of feeling guilty.

The Two-Year Rule

Here's a mental model that helps: Would you start this today, knowing what you know now?

If the answer is no, you should probably stop — regardless of what you've already invested. The past is the past. The only question that matters is whether this is worth continuing from this moment forward.

Think of it like this: if somebody handed you the current situation as a brand new opportunity, would you take it? If you own a stock at $20 that you bought at $50, ask yourself: "Would I buy this stock today at $20?" If the answer is no, sell it. The $30 loss is already real. Holding it doesn't change that.

How to Make Better Decisions

  • Separate past costs from future costs. When evaluating any decision, ignore what you've already spent. Ask only: "From today forward, is this worth continuing?"
  • Set exit criteria in advance. Before you start anything — an investment, a project, a relationship — define what failure looks like. "If this stock drops 20%, I sell." "If I'm still unhappy in 6 months, I quit." Having rules in place removes the emotion when the time comes.
  • Get outside perspective. Talk to someone who has no emotional investment in the decision. They'll see the situation clearly in 30 seconds because they don't carry the baggage of what you've already spent.
  • Reframe the loss. Instead of "I'm losing $30,000 by selling this stock," think "I'm freeing up $20,000 to invest in something better." The money isn't lost — it's being reassigned.

When Sunk Costs ARE Worth Honoring

There's one exception: when the investment is still paying off, just slower than expected.

Learning to play guitar sucks for the first 6 months. If you quit then, you waste those 6 months. But if you push through, you get decades of enjoyment. The key difference: guitar practice has a clear payoff curve. You get better, it gets more fun, and the investment compounds.

The test is simple: Is the fundamentals of this thing sound — am I just in a rough patch — or was the whole idea flawed from the start? Bad timing isn't the same as a bad idea. Sometimes the sunk cost is worth carrying because the future payoff is real.

🛒 Before You Buy...

Stuck on whether to keep spending on something? Try the "Can I Afford It" calculator. It strips away the emotion and gives you the raw numbers.

Try the Affordability Calculator →

FAQ

Is the sunk cost fallacy the same as throwing good money after bad?

Essentially yes. "Throwing good money after bad" is the colloquial version. The sunk cost fallacy is the psychological mechanism behind it — the irrational tendency to continue investing in something because of past investments, rather than future value.

How do I apply this to relationships?

The same principle applies, though with more nuance. "I've been with this person for 5 years" is a sunk cost. The right question is: "If I met this person today, would I start a relationship with them?" If the answer is no, the length of the relationship is keeping you there — not the relationship itself. That said, relationships involve genuine emotional bonds and shared history that stocks don't have. Use the framework, but weigh it carefully.

What about education? Isn't more education always worth it?

No. The average student loan debt in the US is $37,000. If you're two years into a degree that won't increase your earning potential enough to justify that debt, finishing it might be the sunk cost fallacy in action. Run the numbers. If the ROI doesn't work, the "but I've already spent so much" argument is costing you more by the day.