Recency Bias: Assuming Trends Continue

The human mind is a pattern-recognition engine tuned to recent experience. After a decade of rising tech valuations and a brutal oil crash in 2020, investors came to assume these trends would persist forever. Recency bias—the tendency to overweight recent events and assume they represent the true state of affairs—is one of the most dangerous behavioral traps in investing, and arguably the most valuable to exploit as a value investor.

Quick definition: Recency bias is the cognitive tendency to believe that recent patterns and trends will continue indefinitely, giving disproportionate weight to recent data while ignoring longer-term statistical evidence.

Key Takeaways

• Recency bias explains why investors buy what's been winning and avoid what's been losing, precisely the opposite of value discipline
• Markets in strong uptrends (like FAANG stocks 2015–2020) create an illusion of permanence that hides fundamental deterioration
• Value investors profit by recognizing that exceptionally good or bad periods are typically mean-reverting, not structural
• The bias is reinforced by narrative—media coverage, analyst reports, and peer commentary amplify the sense that "this time is different"
• Exploiting recency bias requires the psychological fortitude to ignore short-term underperformance and act when trends appear most entrenched
• Quantifying reversion—building models of historical cycles—helps override the emotional pull of the recent trend

Why Recency Bias Is So Powerful

Recency bias operates on multiple levels. First, it's automatic. When you see five years of positive data, your brain defaults to assuming the sixth year will be positive. Second, it's reinforced by incentive structures. Fund managers, sell-side analysts, and media outlets all profit from chasing recent winners. A fund that did poorly for three years faces redemptions; one that's riding the recent wave attracts inflows. Nobody gets fired for buying what's been working.

Third, recency bias is self-validating in the short run. If tech keeps rising, tech investors look smart for three more years, even if they're sitting on a bubble. The trend itself becomes the evidence. "See, I was right—it's still going up."

The Mechanics of Recency Bias

When a stock or sector has outperformed for several years, four things happen:

1. Cognitive Drift Investors unconsciously begin to believe the outperforming asset class has changed. "It's not just a tech stock anymore; it's a growth story." "Oil isn't cyclical anymore—peak demand is priced in." The original thesis gets rewritten to justify continued ownership.

2. Narrative Capture The financial press, desperate for clicks and engagement, promotes stories that align with recent performance. "Why Tech Will Dominate the 2020s" gets written 50 times. "The Case for Deep Value" gets written twice a quarter and dismissed as contrarian noise.

3. Risk Perception Distortion After five years of low volatility and rising prices, investors perceive risk as low. The chart looks smooth, so the asset feels safe. Volatility has fallen, so risk premia compress. This is precisely when tail risks are most severe.

4. Institutional Herding If recency bias were just individual retail investors, value investors could easily arbitrage it. But when institutional money—pension funds, endowments, hedge funds—all pile into recent winners based on recent track records, they create structural price distortions that persist for years.

Real-World Examples

FAANG 2010–2020: After a decade of outperformance, Big Tech had compressed valuation multiples so dramatically that you could buy the S&P 500 ex-Big Tech at a cheaper forward P/E than you could buy the five largest companies. Most investors saw this and thought, "Good thing I'm in the winners." Value investors saw it and began accumulating beaten-down industrials, financials, and energy. The 2021–2022 rotation proved the latter group correct—but only after another 12 months of underperformance.

Oil 2008–2019: After crude crashed from $147 to $26, investors became convinced that oil demand would be permanently depressed, that the age of cheap fossil fuels was over. Energy stocks traded at 5–7x earnings while tech traded at 25–30x. Between 2020 and 2022, oil rebounded (because demand did bounce back), and energy stocks delivered the strongest returns in any sector. Recency bias about peak demand created a decade-long mispricing.

Interest-Rate Sensitive Stocks 2010–2021: After 11 years of ZIRP (zero interest rate policy), investors became convinced that rates would never rise meaningfully. Banks, insurance companies, and REITs—all interest-rate sensitive—suffered chronic undervaluation because everyone assumed their economics were permanently capped. In 2022, rates rose sharply, and these sectors rallied 30–50% as the recency bias unwound.

Chinese Tech 2020–2021: After a five-year rally, Chinese tech (Alibaba, JD.com, Baidu) was treated as unstoppable. When regulatory headwinds emerged in 2021, investors with recency bias remained shocked. Value investors who had been waiting for a 40–50% correction got their opportunity and bought at valuations unseen in a decade.

Common Mistakes

Mistake 1: Extrapolating a Bull Market Indefinitely After 10 years of rising stock prices (2009–2019), many investors assumed the bull was structural. In early 2020, when COVID hit and stocks crashed, panic selling revealed how deep the recency bias ran. Investors who had lived through the 2008 crash intellectually knew crashes happened, but emotionally felt they were in a "new era." By the time they forced themselves to sell, the reversion was well underway.

Mistake 2: Dismissing Mean Reversion as an Old Idea A corollary of recency bias is the dismissal of historical patterns as obsolete. "Mean reversion doesn't apply to tech anymore." "Cyclicality is dead in energy because of climate policy." Every cycle, investors are convinced their cycle is different. Value investors exploit this by waiting for the inevitable reversion and buying when mean-reverting assets are cheapest.

Mistake 3: Conflating Recent Valuations with Fair Value Because you've never seen a stock trade below a certain multiple, you assume that's the "floor." A stock that traded at 15x for five years begins to feel like it's cheap at 20x. But if the true long-term average is 12x, you're still paying above fair value. Recent history is not a guide to fair value; earnings power and capital requirements are.

Mistake 4: Over-Allocating to Recent Winners Investors often boost allocations to assets after they've performed well, compounding the recency bias. This is the opposite of rebalancing. A disciplined rebalancer would trim winners and add losers—the essence of contrarian behavior.

FAQ

Q: Isn't following trends rational if markets are efficient? No. If markets were efficient, trends wouldn't persist long enough for recency bias to create profits. Markets are efficient on average over long periods, but highly inefficient over medium-term periods (2–10 years), which is where recency bias operates. The existence of exploitable recency-bias-driven mispricings implies markets aren't efficient in the strong form.

Q: How do I know if I'm a victim of recency bias? Ask yourself: "If I saw the chart of this stock/sector 10 years in the past, what was my view?" If your answer is "I'd think it was overvalued," and the fundamental metrics haven't improved proportionally, you're likely being influenced by recency bias. Also ask: "Is my thesis dependent on this trend continuing?" If yes, you're vulnerable.

Q: What's the best way to quantify mean reversion? Use z-scores of valuations (how many standard deviations a metric is from its 10–20 year mean) and historical reversion rates. For example, if P/E ratios revert to the mean within 3–5 years on average, and a stock is trading at 2.5 standard deviations above its 15-year mean, you can model the time and magnitude of likely reversion. This isn't perfect, but it beats relying on narrative.

Q: Aren't there structural changes that justify new valuations? Yes, sometimes. The shift from hardware to software, the rise of cloud computing, and the emergence of network effects in tech were all real structural changes. But they were priced in far before they created the extremes seen in 2020. By the time a structural change is five years into its manifestation and embedded in consensus, it's usually already priced. The real opportunity is buying companies before structural tailwinds become obvious.

Q: How long does a reversal typically take? Mean reversion can take 2–10 years depending on the magnitude of the mispricing, the catalyst required, and the degree of structural change. The lesson: exploit recency bias by buying undervalued assets for the long term, not by trying to time a reversal precisely.

Related Concepts

• Mean Reversion: The statistical tendency for prices to revert toward long-term averages. Recency bias is the cognitive trap that prevents investors from acting on this.
• Narrative Fallacy: The human compulsion to construct stories around data. Recency bias + narrative fallacy creates powerful conviction in trends that won't persist.
• Contrarian Investing: The discipline of betting against recent trends and consensus sentiment. This is the opposite of recency bias.
• Market Cycles: Historical patterns show that nearly all markets—stocks, commodities, currencies, real estate—experience cycles. Recency bias makes these cycles invisible until they reverse.
• Trend Reversal: The moment a trend breaks is almost always when recency bias is deepest. This is why the best buying opportunities coincide with maximum pessimism.

Summary

Recency bias is a powerful force that creates systematic mispricings. By convincing investors that recent trends are permanent, it pushes valuations to extremes. Value investors win by recognizing these extremes, building patience to endure underperformance while the trend persists, and deploying capital when mean reversion begins.

The battle against recency bias is not primarily intellectual—most sophisticated investors can articulate mean reversion cogently. The battle is psychological: staying disciplined when the crowd is convinced "this time is different," when your portfolio is down 20% while everyone else is up 40%, and when five years of data seem to contradict your contrarian thesis.

But the mathematics of mean reversion are inexorable. Extreme valuations eventually normalize. The only question is whether you have the temperament to wait for it.

Next

Read about The Action Bias to understand how the urge to trade constantly works against the patient, long-term discipline required to exploit recency bias.