Skip to main content

Recency & Availability Bias

Updated Jun 24, 2026 at 2:35pm

Research Draft High 1,169 words

Recency bias and availability bias are two closely related cognitive shortcuts that distort how investors judge probability and likely future outcomes. The availability heuristic — formalized by Amos Tversky and Daniel Kahneman in 1973 — is the tendency to estimate how frequent or likely something is by how easily examples come to mind. Recency bias is a specific, temporal case of it: the most recent events are the easiest to recall, so they get overweighted relative to the longer historical record. The core tension is that ease of recall is a poor proxy for true frequency or probability. Vivid crashes, recent rallies, dramatic stories, and round-the-clock news are all highly "available," but availability is driven by salience, recency, and emotional impact — not by base rates. The result is a systematic mismatch between how investors expect markets to behave and how they actually do.

How it's formed (the mechanism)

Both biases stem from how human memory retrieves information. In their original work, Tversky and Kahneman argued that people substitute the question "how probable is this?" with the easier question "how easily can I think of an instance?" Instances that are recent, vivid, emotionally charged, or frequently reported are retrieved faster and judged as more common. This is a fast, effortless System-1 process that conserves cognitive load but deviates from Bayesian reasoning, because retrievability correlates only imperfectly with actual frequency.

In markets, three reinforcing channels make the effect strong:

  • Recency weighting in memory. Recent returns dominate the recall set, so investors extrapolate the last 6–12 months forward. Greenwood and Shleifer (2014), studying six independent survey series of investor return expectations from 1963–2011, found those expectations are highly correlated with past returns and the level of the market — yet negatively correlated with model-based future expected returns. In other words, surveyed investors are most optimistic exactly when subsequent returns tend to be lowest.
  • Cued recall and emotion. Jiang, Liu, Peng and Yan (Investor Memory and Biased Beliefs: Evidence from the Field, QJE 2025) provide field and survey evidence (a large sample of Chinese retail investors) that high recent returns cue more positive memories, which in turn generate more optimistic forecasts — an explicit memory-based microfoundation for return extrapolation.
  • Salience / attention. Barber and Odean's work on attention-grabbing stocks shows individual investors disproportionately buy stocks that are recently in the news, have unusual volume, or extreme prior-day returns — using availability to prune an unmanageably large choice set.

How it's used in practice

In practice these biases are mostly something to guard against rather than exploit, and the application is the same across investing styles:

  • Performance chasing. The clearest footprint is money flowing into whatever recently performed best. Frazzini and Lamont's "dumb money" study (2008) found that retail mutual-fund flows reliably move toward recent winners and that, on average, these reallocations reduce investors' wealth over the long run — high-sentiment (heavily-bought) stocks subsequently underperform low-sentiment ones — because the chased funds and growth stocks tend to mean-revert.
  • Return extrapolation in expectations. Investors who lived through a long bull market price in continued gains; those scarred by a crash stay defensively positioned through the recovery. Vanguard and Schwab note retail expectations turn bullish after good years and bearish after bad ones — the opposite of what valuation implies.
  • Risk misjudgment. After a crash (2000, 2008, 2020), perceived crash probability spikes and stays elevated long after it should; in calm markets, tail risk feels unavailable and gets underpriced. The same mechanism feeds momentum-style trend-following on the way up and capitulation at the bottom.
  • Practical countermeasures that the literature and practitioners converge on: anchoring decisions to long-run base rates and full-cycle data, rules-based rebalancing (which mechanically sells recent winners), written investment plans, and lengthening the look-back window used to form expectations.

Adoption, debate & evidence

Availability is one of the most established findings in cognitive psychology — Kahneman won the 2002 Nobel in economics partly for this body of work, and the heuristic is taught as foundational in behavioral finance (CFA curriculum, Investopedia, academic texts). The existence of return extrapolation is unusually well-evidenced for a behavioral claim: it appears across multiple independent survey datasets (Greenwood–Shleifer), in fund flows (Frazzini–Lamont), and now in memory experiments (QJE 2025).

Where honest caveats apply:

  • Folklore vs. measured. The direction is robust; the magnitude is regime-dependent and sample-specific. Any single dollar or percentage figure for the "dumb money" wealth cost is a particular study's estimate over a specific period and sample — not a universal constant. Treat such numbers as illustrative, not a law.
  • Boundary with rational explanations. Some return-chasing is defensible (flows to genuine skill; time-varying risk premia). The behavioral and rational camps still debate how much of the flow-performance relationship is bias versus rational learning.
  • Debiasing is hard. Research on whether recency can be trained away is mixed; awareness alone does little, while structural rules (forced rebalancing, pre-commitment) work better than willpower.

Strengths & limitations

As heuristics, availability and recency are genuinely useful: they let an investor act under uncertainty without processing every data point, and recent information is sometimes the most relevant (a regime really did change). The failure mode is treating a small, salient, recent sample as representative of the true distribution.

The #1 misuse is extrapolating recent performance into the future — buying after a run-up, selling after a drawdown — which is mechanically the inverse of "buy low, sell high." A close second is letting a single vivid event (a memorable blow-up, a friend's 10-bagger) override base-rate reasoning about how often such outcomes actually occur. These biases are also the engine behind bubble and panic dynamics when they aggregate across the market.

Sources

Dispute flags: the existence of return extrapolation / recency overweighting is well-supported; specific cost magnitudes for flow-chasing are single-study, period-specific estimates and should not be quoted as fixed constants. The behavioral-vs-rational interpretation of flow-chasing remains contested.