Hindsight bias
What is Hindsight bias?
Hindsight bias is the tendency, after an outcome is known, to view it as more predictable or inevitable than it appeared beforehand.
Decision journals should freeze the information set, alternatives, probabilities, and constraints at the time. Later reviews can add outcomes without editing the original entry. This supports learning from process, distinguishing bad luck from a weak decision and good luck from skill, while avoiding a new story fitted to known results.
Organizations should reward honest uncertainty and accurate process records; punishing every adverse outcome encourages vague forecasts and destroys the evidence needed for genuine learning.
How narratives change
Memory reconstructs prior beliefs using current knowledge. Warning signs that fit the result become salient, competing possibilities fade, and ambiguous forecasts are remembered as precise. Market commentary can turn a surprise into an obvious consequence within hours. The outcome may have been plausible without having been probable or investable in advance.
Portfolio consequences
Hindsight can create false confidence, unfairly reward or punish managers, encourage overfitted strategies, and prevent learning. A good process with an adverse random outcome can be abandoned, while a reckless profitable decision is repeated. Investors may believe they called turning points that their dated positions and forecasts did not actually anticipate.
Example
After a company fails, observers say its decline was inevitable because one risk was disclosed. Before failure, several outcomes depended on financing, customers, and policy. A proper review compares the assigned probabilities, position size, evidence, and contingency plan with what was knowable then, rather than treating eventual failure as proof of certainty.
Decision controls
Record forecasts, probability ranges, assumptions, alternatives, invalidation conditions, and positions before outcomes. Preserve contemporaneous data and meeting notes. Review both process and result using base rates. Score calibration over many predictions. Premortems identify failure paths in advance, while postmortems should avoid rewriting the original thesis.
Practical interpretation
Ask what probability was assigned, what other outcomes were credible, and what trade was feasible at the time. Separate predictable from profitable because timing, valuation, carry, and risk limits matter. Use dated evidence to challenge confident recollection. The aim is accountable learning, not excusing every loss as unforeseeable uncertainty.
Sources and further reading
- Behavioral Patterns of U.S. Investors, U.S. Securities and Exchange Commission