Skip to main content

How Does Recency Bias Cause You to Overweight Recent Financial News?

A stock has rallied 50% in the past year. All the financial news about it is positive: analyst upgrades, earnings beats, industry tailwinds. You're convinced the rally will continue. You buy the stock. But six months later, it falls 30%, and you're shocked. Or a stock has fallen 40% over the past year. All the news is negative: downward guidance, competitive pressure, sector weakness. You avoid it entirely. But six months later, it rebounds 40%, and you missed it.

In both cases, you're falling prey to recency bias: the tendency to give more weight to recent information and to assume that recent trends will continue into the future. Because the recent period (the past year) shows a strong trend (up or down), you assume the trend will persist. You extrapolate the recent past into the near future. But this is often wrong. Trends reverse, momentum fades, and mean reversion is common in financial markets.

Quick definition: Recency bias is the cognitive tendency to give more weight to recent information and to assume that recent trends or conditions will continue, while underweighting historical averages and the possibility of mean reversion.

Recency bias is one of the most costly interpretation mistakes in financial investing. It causes you to buy stocks near peaks (when they've already rallied and everyone is positive), and sell stocks near bottoms (when they've already fallen and everyone is negative). It causes you to overweight whatever has recently worked (tech in 2020, value in 2022) and underweight whatever recently didn't (value in 2020, tech in 2022). It's the reason many investors systematically buy high and sell low, not through dramatic errors but through the quiet distortion of recent trends.

Key takeaways​

  • Recent trends feel predictive but often aren't. A stock has rallied 50% this year; this doesn't predict next year's return. In fact, past performance has little correlation with future performance.
  • Financial markets exhibit mean reversion. Assets that have outperformed tend to underperform in the future, and vice versa. Ignoring this is recency bias.
  • Recency bias + financial news = dangerous feedback loop. Recent outperformance gets positive news coverage; recent underperformance gets negative coverage. Both reinforce recency bias.
  • Recency bias affects asset allocation. It causes investors to overweight whatever has recently worked and underweight whatever hasn't, leading to poor diversification.
  • Long-term data is the antidote. Checking 10-year or 20-year performance records reduces the pull of recent trends.

How Recency Bias Works in Financial News​

Recency bias has been extensively documented in academic literature on financial decision-making. The Federal Reserve's consumer finance research and SEC investor education materials both highlight how recent performance and recent news disproportionately influence investor behavior, often to their detriment over long time horizons.

The Trend Extrapolation Error​

A sector has outperformed for the past two years. Tech stocks have beaten energy and financials. Growth has beaten value. The news is full of stories about tech's dominance, AI's potential, the structural decline of energy. Investors extrapolate: "Tech will continue to outperform." They rotate capital into tech, away from other sectors.

But the very fact that tech has already outperformed significantly is evidence that it might underperform in the future. If tech has gone from 15% of the market to 35%, it's become overweight relative to fundamentals. The probability of mean reversion—tech underperforming as it normalizes—has increased. Yet recency bias makes you assume the recent trend will continue, not reverse.

This is a dangerous extrapolation. It causes investors to buy sectors right when they're becoming expensive, and to sell sectors right when they're becoming cheap. The investors feel like they're "following the trend" or "looking where the growth is." In reality, they're being manipulated by recency bias into a contrarian position (overweighting what's already worked).

The News-Narrative Reinforcement​

Financial journalists write stories about trends. When a sector has outperformed, journalists write about why it's the future (new technology, changing consumer preferences, regulatory tailwinds). These narratives are plausible, and they're written about a sector that has actually outperformed. An investor reads the narrative and extrapolates: "This sector will continue to outperform because of these trends."

But the narrative is often based on past performance (we can see that the sector has outperformed) and plausible future expectations (we can imagine why it might continue). The narrative feels predictive because it's post-hoc explanatory. In reality, past performance and plausible narratives don't predict future performance very well.

Here's the dangerous feedback loop: (1) A sector outperforms. (2) Journalists write about why it's winning. (3) Investors read the narratives and extrapolate. (4) Investors rotate capital into the sector, driving further outperformance. (5) Further outperformance generates more confident narratives. (6) Investors become more convinced and rotate more capital. (7) The sector becomes increasingly expensive and overweight. (8) Eventually, the trend reverses, causing mean reversion that's even more severe because the sector has become even more overweight.

The Performance-Chasing Error​

An investor reviews performance data. Fund A has returned 20% per year for the past five years. Fund B has returned 10% per year. The investor concludes: "Fund A is better; I should invest in Fund A." This is recency bias. The investor is extrapolating the past five years into the future, assuming Fund A will continue outperforming.

But research shows that past performance is a poor predictor of future performance. In fact, there's a slight tendency for past winners to underperform past losers in the future (mean reversion). Investors who chase past performance—buying high-returning funds and selling low-returning funds—systematically buy high and sell low.

This is so common that financial advisors have a rule: "Don't invest based on past performance." Yet investors routinely ignore this rule, falling prey to recency bias. They see that a growth fund has returned 25% and a value fund has returned 5%, and they overweight growth. Two years later, value has returned 20% and growth has returned -5%, and they regret the overweighting. But if they've rotated out of growth and into value (another recency bias move, now overweighting what's recent), they'll regret that, too, when the cycle reverses again.

The Economic-Cycle Recency Bias​

The economy is in expansion; unemployment is falling. Investors extrapolate: "The expansion will continue indefinitely." They overweight cyclical stocks (companies that do well in expansions). But expansions end. Recessions come. The investors who extrapolated the expansion end up overweight cyclical stocks right before the recession, taking large losses.

Conversely, the economy enters recession; unemployment is rising. Investors extrapolate: "The recession will deepen; we're heading for a depression." They flee to defensive stocks and bonds. But most recessions are brief (average length: 11 months). The investors who extrapolated the recession miss the recovery and get whipped by the market rally that typically begins before the recession is even officially over.

The Volatility Recency Bias​

Markets were volatile last month; you assume they'll be volatile next month. Markets were calm last month; you assume they'll be calm next month. But volatility is actually somewhat unpredictable and doesn't have a strong autocorrelation (recent volatility is not strongly predictive of future volatility). Yet investors often adjust their risk exposure based on recent volatility, creating a pattern of buying calm markets right before they become volatile, and selling volatile markets right before they calm down.

A Framework for Detecting Recency Bias​

Real-world examples​

Historical price data from Federal Reserve FRED database and Census Bureau housing data document these cases:

Case 1: The 2000 Tech Crash (Recency Bias Building Bubble, Then Crashing)​

In the late 1990s, tech stocks had outperformed dramatically. Investors who had been in tech for years had massive gains. Financial news was full of stories about the "new economy," the internet, and the "death of distance." Investors extrapolated: "Tech will dominate forever. The old economy (manufacturing, finance, energy) is obsolete."

Recency bias made investors overweight tech dramatically. Money poured into tech stocks, IPOs, and dot-coms. By early 2000, the NASDAQ was trading at a P/E ratio of 200+, divorced from any reasonable earnings estimate. The extrapolation of recent outperformance had driven valuations to unsustainable levels.

Then the bubble burst. From 2000 to 2002, the NASDAQ fell 75%. Investors who had extrapolated the recent tech outperformance into the future (recency bias) suffered massive losses. The irony: investors who had been cautious about tech in the 1990s, who didn't extrapolate the trend, had better returns over the 20-year period than investors who chased tech into the bubble.

Case 2: The 2008 Housing Crash (Recency Bias in Assets, Not Stocks)​

Housing prices had risen for years in the 2000s. Investors extrapolated the trend: "Housing always goes up. Housing is a safe investment." Financial media was full of stories about real estate wealth, flipping houses, and the safety of home equity. Investors, trusting in the extrapolation of recent trends, invested heavily in housing, refinanced at low rates, and took on leverage.

The crash came. Housing prices fell 30%+. Investors who had extrapolated the recent trend of rising prices got badly hurt. The tragedy is that a check of long-term housing-price history would have shown that real housing prices (adjusted for inflation) had been flat or declining for much of the 20th century. The rapid rise of the 2000s was an anomaly, not a new regime. Recency bias made the anomaly seem like the new norm.

Case 3: The Growth vs. Value Cycle (Repeated Recency Bias Errors)​

From 2017 to 2020, growth stocks outperformed value stocks. By 2021, growth had outperformed value for several years. Investors extrapolated: "Growth will continue to outperform. Value is dead." Investors rotated heavily into growth, out of value.

Then in 2022, value outperformed growth. By 2023, growth was rallying again. The investors who had extrapolated growth's outperformance out of value in 2021, and who then extrapolated value's outperformance in 2022, had rotated twice in ways that lost money both times. The recency bias cycles were destroying returns.

A disciplined approach would have been to maintain a consistent allocation and not chase recent trends. But recency bias made this hard for investors to do.

Case 4: Energy Stocks and Climate Bias (2010–2020)​

In the 2010s, energy stocks underperformed dramatically. Oil prices fell from $100 to $50. The financial narrative emphasized the "death of fossil fuels," the inevitability of renewable energy, and the existential risk to energy companies. Investors extrapolated: "Energy stocks will continue to underperform forever."

Investors sold energy stocks and bought renewable-energy stocks (which didn't exist as a category yet; they bought tech stocks and clean-energy ETFs instead). But by 2022, energy stocks had become cheap on both valuation and earnings metrics. When energy prices rebounded in 2022, energy stocks rallied 50%+ while many growth stocks fell. Investors who had extrapolated the 2010s underperformance into the future (recency bias) missed the rebound.

The tragedy is that energy stocks were genuinely offering good value by 2020, but the recency bias of their long underperformance made them appear "bad" to investors, who avoided them just as they were becoming attractive.

Common mistakes​

  1. Buying a stock or fund because it's rallied recently. You notice a stock has returned 50% in the past year. You buy it, assuming the rally will continue. But recency bias is making you extrapolate the recent performance. Past performance has little correlation with future performance. Always ask: "Is this asset now fairly valued, or has recent outperformance made it expensive?"

  2. Selling a stock or fund because it's underperformed recently. A fund has returned 2% while the market returned 10%. You sell it, assuming underperformance will continue. But you might be selling right before the fund rebounds. Past underperformance is not predictive of future underperformance. In fact, mean reversion suggests underperformers might outperform.

  3. Assuming a bull market will continue forever. The market has rallied for three years. You're convinced "the bull market is intact." You increase stock allocations. But bull markets don't last forever. The probability of a pullback increases the longer the bull market lasts, not decreases. Don't extrapolate recent bull-market strength into indefinite continuation.

  4. Assuming a bear market will continue indefinitely. The market has fallen 20%. You're convinced "we're heading into a depression." You sell everything. But bear markets are typically brief. Most of the time, you miss the recovery by assuming the recent downtrend will continue. Bear markets are buying opportunities, not confirmation of a lasting decline.

  5. Extrapolating sector strength or weakness. Tech has outperformed for three years; you assume it will forever. Energy has underperformed for a decade; you assume it will forever. But sectors cycle. Recent outperformance is often a sign that a sector is becoming overvalued and due for a reversion. Avoiding overvalued sectors (due to recency bias) and buying undervalued sectors is a better long-term strategy.

FAQ​

In some cases, yes. If a company's earnings have been accelerating due to structural reasons (new market, new product, market-share gains), the trend might continue. If an economic trend (like aging population or technological adoption) has been driving long-term change, it might continue. But the bar is high. Most short-term trends (past year or two) are not predictive. You need a structural reason for the trend to believe it will continue.

How much weight should I give to recent performance when making investment decisions?​

Use a weighted average: weight the recent period (past year or two) much less heavily than your historical average. A stock has returned 50% in the past year but 5% per year over the past 10 years. Your expected future return might be closer to 8% (weighted average, skewed toward the long-term average) than to 50% or 5%. This avoids both recency bias (overweighting the past year) and ignoring the recent period (which might contain valuable information).

Is recency bias different from availability bias?​

Yes. Recency bias is about overweighting recent data because it's recent. Availability bias is about overweighting data that's easy to come to mind (vivid, memorable, widely covered). Both lead to overweighting recent information, but the mechanisms are different. A recent crash (which is both recent and vivid) triggers both biases. An old crash that's widely remembered (not recent, but highly available) triggers availability bias but not recency bias.

Can I use recency bias strategically by betting against it?​

Somewhat. Mean reversion strategies exist, and they have positive expected value in some contexts. A stock falls 50%; you expect mean reversion and buy, expecting it to partially recover. But timing mean reversion is hard. A stock that falls 50% might fall another 50% (think of fallen unicorns or failed startups). Betting against recency bias isn't a no-brainer; it requires confidence that mean reversion will occur, which requires deeper analysis than just checking that something is "down" recently.

Do professional investors avoid recency bias?​

No. Professional investors, especially those under performance pressure, are often vulnerable to recency bias. A fund that has underperformed for a year faces redemptions, so it has pressure to overweight what's recently worked. This can cause professionals to chase performance just like retail investors do. However, good investment firms have processes and disciplines that resist recency bias, making them less vulnerable than the average investor.

Summary​

Recency bias is the tendency to overweight recent information and assume recent trends will continue into the future. In financial markets, this bias is costly because trends frequently reverse due to mean reversion, valuation mean reversion, and the cyclical nature of markets and economies. Recency bias causes investors to buy high (when assets have recently outperformed and become expensive) and sell low (when assets have recently underperformed and become cheap). Financial news amplifies this bias by narrating recent trends as if they're predictive of the future. You can protect yourself by: (1) checking long-term performance data and comparing recent performance to historical averages; (2) asking whether recent outperformance has made an asset expensive (reducing future return potential); (3) maintaining consistent allocations rather than chasing recent trends; and (4) recognizing that mean reversion is a powerful force in financial markets, and that what has recently worked is often due for underperformance. The goal is not to completely ignore recent information, but to weight it appropriately alongside historical context and fundamental valuation. This balanced approach reduces the damage caused by recency bias over the long term.

Next​

→ Overweighting Loud Voices in Market News