Recency bias in investing: why recent returns are a poor guide to future performance

Markets move. After they move in one direction for long enough, investors begin to believe the direction will continue—not because the evidence supports it, but because recent experience dominates their expectation. Benartzi & Thaler (2001) documented this pattern in 401(k) contribution data: participants allocated more to equity funds after strong recent equity returns, extrapolating recent performance into future expectations. Recency bias is one of the most reliably costly errors in investment behaviour.

What recency bias is

Recency bias is a cognitive shortcut in which the mind assigns disproportionate weight to recent observations relative to the full historical distribution. The brain is not designed to weight all evidence equally—recent events are more vivid, more available, and easier to recall. The result is that expectations are anchored to what just happened rather than to what the long-run evidence suggests is likely.

Benartzi & Thaler (2001), Naive Diversification Strategies in Defined Contribution Saving Plans, American Economic Review, in their study of 401(k) participant behaviour, found that allocation decisions were systematically influenced by recent fund performance. When equities had performed well in the preceding period, participants increased their equity allocations. The decision was driven by the recent observation, not by any change in the long-run expected return of equities—which was unaffected by the recent price movement.

How it manifests in investing

Recency bias produces several recurring and costly investment patterns.

Chasing recent fund performance. Investors disproportionately allocate to funds that performed best in the most recent one-to-three-year period. Those funds attracted capital because their recent returns were high—but recent high returns in actively managed funds are a weak predictor of future returns and a strong predictor of mean reversion. The investor who buys last year's top performer is often buying at the peak of that fund's relative performance cycle.

Abandoning strategies after drawdowns. A strategy that has recently experienced a drawdown looks, through the lens of recency bias, like a broken strategy. The investor who abandons it at the trough—switching to a different strategy or to cash—locks in the loss and misses the recovery. The recent loss dominates the assessment of the strategy's long-run merit.

Over-allocating to assets at cycle peaks. When an asset class has delivered strong recent returns, recency bias makes those returns feel normal and sustainable. Investors allocate more precisely when expected future returns are lowest, because valuations have already risen in response to the recent performance they found so compelling.

The cost

The pattern of buying into recent strength and selling into recent weakness is self-defeating at scale. Benartzi and Thaler's research on 401(k) allocation, conducted across a dataset of over 1.5 million participants, documented return-chasing behaviour that systematically disadvantaged participants relative to a stable allocation. Goetzmann & Peles (1997), Cognitive Dissonance and Mutual Fund Investors, Journal of Financial Research, found that fund investors systematically overestimated their own past returns—a finding consistent with recency bias distorting not just forward expectations but backward memory. The combined effect is an investor who remembers recent gains as larger than they were, extrapolates them forward, and allocates accordingly—a reliable path to buying high.

What helps

Systematic strategies evaluate signals consistently across defined time windows rather than weighting recent observations selectively. A momentum signal calculated over a 12-month window, for example, incorporates the full year of price history—not only the most recent weeks. The strategy's rules prevent the investor from abandoning it on the basis of short-term underperformance, because the sell decision is governed by the signal, not by the investor's current assessment of how the strategy feels. The discipline of following a defined process is one of the most effective counterweights to recency bias.

Recency bias in pfolio

pfolio's systematic strategies evaluate momentum signals over defined windows of 6–12 months, incorporating the full period of price history rather than weighting recent weeks more heavily. The rules specify when a signal changes; the investor's current perception of how the strategy feels is not an input. When a systematic strategy is experiencing a drawdown, the exit condition is the momentum signal flipping negative—not the investor's assessment that the recent poor run is likely to continue. This removes the primary mechanism through which recency bias causes investors to abandon strategies at their worst moment. For context on how pfolio strategies have performed across prior drawdown periods, see pfolio Insights.

Related articles

• Overconfidence bias
• Myopic loss aversion
• Action bias