Skip to main content
← Glossary · Behavioral Economics

Disposition Effect

The tendency to sell winning investments too early and hold losing investments too long, driven by loss aversion and mental accounting.

What Is the Disposition Effect?

The disposition effect is the tendency to realize gains too early and losses too late. Investors sell stocks that have risen (locking in the emotional win) and hold stocks that have fallen (avoiding the emotional pain of admitting loss). The same pattern shows up in product decisions, feature investments, and A/B test management.

Also Known As

  • Marketing teams: "campaign hold-on problem"
  • Sales teams: "holding losers"
  • Growth teams: "feature-kill avoidance"
  • Product teams: "sunk-cost retention"
  • Behavioral science: Shefrin and Statman's (1985) disposition effect

How It Works

An A/B test shows a clear negative result by week two — the variation is losing. Instead of calling it, the team extends the test, segments the data, and spends two more weeks looking for a subgroup where the variation wins. They eventually find one marginally significant slice and ship. Six months later, overall metrics are down. The team couldn't bring themselves to "sell the loser" and realize the loss.

Best Practices

  • Do pre-commit to stopping criteria before launching any test or initiative.
  • Do build decision rules that remove emotion from kill/continue choices.
  • Do celebrate killed experiments and sunsetted features as program wins, not losses.
  • Don't let teams redefine success metrics mid-test to rescue a losing variation.
  • Don't extend tests indefinitely "just to be sure."

Common Mistakes

  • Keeping underperforming features alive because "we've already invested so much."
  • Letting winning tests ship without proper validation, while losing tests get extended indefinitely.
  • Treating kill decisions as punishment rather than hygiene.

Industry Context

  • SaaS/B2B: Feature portfolios that accumulate zombie features; experimentation programs that never call losers.
  • Ecommerce/DTC: Product lines that stick around past profitability; holding losing inventory.
  • Lead gen/services: Channels that lost ROAS months ago but still get budget.

The Behavioral Science Connection

Hersh Shefrin and Meir Statman documented the disposition effect in 1985, initially in stock market behavior. It combines loss aversion (losses feel twice as painful as gains feel good) with mental accounting (each investment is its own mental account, and closing a losing account feels like personal failure). It's a direct consequence of Prospect Theory's asymmetric value function.

Key Takeaway

Pre-commit to your kill criteria before launch — emotional resistance to realizing losses is universal, but decision rules bypass it.