Sunk Cost Fallacy
The sunk cost fallacy is the cognitive bias in which individuals continue investing time, money, or resources into a failing endeavor because of the accumulated prior costs that have already been spent and cannot be recovered, rather than evaluating the decision based solely on future costs and benefits. In investing, this manifests as holding losing positions too long due to reluctance to realize losses, distorting rational portfolio management.
Key takeaways
- Sunk costs are by definition irretrievable and should have no bearing on forward-looking investment decisions; only future expected returns and risks are relevant.
- In portfolio management, sunk cost fallacy causes investors to hold deteriorating positions to 'get back to even,' allowing losses to compound rather than reallocating to better opportunities.
- The fallacy is closely related to loss aversion in prospect theory—the psychological pain of realizing a loss exceeds the pleasure of an equivalent gain, creating reluctance to exit losing positions.
- Institutional investors combat the sunk cost fallacy through systematic review processes, stop-loss rules, and investment committee structures that force reconsideration of investment theses independent of historical cost basis.
- The fallacy affects corporate capital allocation decisions as well: managers continue funding failing projects because of prior R&D expenditures rather than evaluating marginal returns on incremental capital.
Explanation
The sunk cost fallacy represents one of the most practically significant cognitive errors in financial decision-making because it systematically distorts portfolio management behavior in ways that compound losses and delay reallocation to productive opportunities. The term 'sunk cost' refers to expenditures that have already been made and cannot be recovered, regardless of future actions. Sound economic theory dictates that sunk costs should be irrelevant to forward-looking decisions, which should be evaluated solely on the basis of future incremental costs and benefits. The fallacy occurs when decision-makers allow past costs to influence future choices despite this irrelevance.
In investment management, the sunk cost fallacy most commonly manifests as the 'anchoring to cost basis' phenomenon. An investor who purchased shares at $100 and has watched them fall to $60 faces a psychological barrier against selling because doing so would 'lock in' a 40% loss—even if the forward-looking investment case has deteriorated significantly and the $60 currently invested could generate better risk-adjusted returns elsewhere. The loss aversion literature (Kahneman and Tversky) suggests that losses loom approximately 2.25 times larger than equivalent gains in psychological terms, creating a powerful emotional resistance to realizing losses that should trigger selling.
The practical consequences for portfolio performance are severe. Research by Shefrin and Statman (1985) documented the 'disposition effect'—the empirical tendency for investors to sell winning positions too quickly (capturing gains) and hold losing positions too long (avoiding realized losses). This behavior is the direct manifestation of sunk cost fallacy and loss aversion in portfolio management. It results in portfolios that become increasingly concentrated in losers while winners are systematically trimmed, effectively creating a negative momentum bias that destroys alpha over time.
Professional investment organizations combat the sunk cost fallacy through several structural mechanisms. Pre-defined stop-loss rules (e.g., sell any position that falls more than 20% from the entry price regardless of thesis) remove the emotional decision-making from loss realization. Investment review frameworks that require periodic re-underwriting of positions—asking 'would we buy this at the current price given current information?'—force analysts to evaluate positions based on forward expected return rather than prior cost. Position sizing discipline, through which each position is sized based on current conviction rather than historical cost, prevents the psychological over-weighting of positions that happen to be large due to prior unrealized appreciation.
The sunk cost fallacy extends beyond individual stock selection to broader corporate and fund management decisions. A hedge fund manager who has spent 18 months developing a complex credit strategy that has not generated returns may continue allocating resources and capital to the strategy due to the psychological weight of the prior investment, even as better opportunities present themselves. Corporate boards famously fall into this trap with large capital projects: the Concorde aircraft was continued by British and French governments long after it was clear the project was economically unviable, largely because of prior development expenditure—giving rise to the 'Concorde Fallacy' as a synonym for sunk cost reasoning.
Example
A hedge fund analyst initiates a long position in Retailer Corp at $80 per share, with a price target of $110 and stop-loss policy at $64 (−20%). The stock falls to $64 as same-store sales data disappoints, triggering the stop-loss level. However, the analyst has spent two months building the thesis and feels the weakness is temporary—a classic sunk cost influence. Instead of honoring the stop-loss, the analyst argues to maintain the position. The stock continues to fall to $45 over the next quarter as structural headwinds materialize. The loss grows from a $16/share stop-loss exit to a $35/share realized loss—a cost more than double what the stop-loss would have produced. The $80 purchase price was a sunk cost irrelevant to the decision at $64; only the forward expected return at $64 was relevant.
Related terms
Alpha Availability Heuristic Basis Confirmation Bias Disposition Effect Hedge Fund January Effect Loss Aversion Mean Reversion Bias Recency Bias Stock