Why sunk cost fallacy matters
It can lead investors to add to weak positions, retain failing managers, or fund projects after the original thesis has broken.
How it is applied
Decision-makers exclude unrecoverable costs, update forecasts, compare alternatives, use stop criteria, and assign independent review. The sunk cost fallacy is continuing a decision because of resources already spent even though those costs cannot be recovered and should not affect the best choice now. Investment processes counter it by reassessing expected return, downside, alternatives, tax, and transaction cost from the current point forward.
Portfolio example
A fund provides more capital to a failing project mainly because it has already invested heavily. An investor bought a stock at 100 and it now trades at 60 after the business deteriorated. Refusing to sell until it returns to 100 anchors the decision to an irrecoverable purchase price. The relevant question is whether the stock offers the best prospective use of 60 today.
How to interpret it
The correct question is whether incremental expected value is attractive today. Prior cost matters only if it changes future cash flows or information. Past expenditure can explain emotions and constraints but should not create incremental economic value. Rational decisions include future switching costs and taxes, which are not sunk. A losing investment can still be worth holding if forward prospects justify it, but not merely because the investor has already lost money.
Limitations and common misconceptions
Persistence can be rational when turnaround value exists. Mechanical abandonment can destroy options and relationships. Labeling every persistent commitment a fallacy is too simplistic. New information may support the original plan, and abandoning projects has reputational or contractual consequences. Outcome bias can make a sound decision look irrational after bad luck. The diagnosis requires reconstructing information available at each decision date. Practical controls include written sell criteria, independent review, base-rate comparison, and asking whether the position would be purchased at its current price. Portfolio systems should display opportunity cost and thesis changes, not only gain or loss since purchase. Related terms include anchoring, loss aversion, disposition effect, and confirmation bias.
Sources and further reading
- The Behavioral Biases of IndividualsCFA Institute
- Behavioral Patterns of U.S. InvestorsU.S. Securities and Exchange Commission