Recency Bias in Investing: How Chasing Last Year’s Winners Quietly Costs You 1–2% a Year
In Morningstar’s 2024 Mind the Gap study, the average mutual fund returned 7.3% annualized over the past decade — but the average investor in those same funds earned only 6.0%. That 1.3-percentage-point gap is, in plain English, what recency bias in investing costs the typical person every year. The funds didn’t fail their investors. The investors faded in after the run-up and sold after the drawdown, then repeated the cycle with the next hot category.
Most people don’t notice this happening. The pattern hides inside reasonable-sounding logic: “this strategy has been working,” “this sector is finally turning,” “this manager is on a streak.” This article walks through what recency bias in investing actually looks like in a real portfolio, the 20-year data set that debunks the “hot streak” story, and the three discipline rules that quietly defeat it.
The Common Belief: Last Year’s Winners Will Keep Winning
If you ask a casual investor how they pick a fund, the answer is almost always some version of: I look at the recent returns. Open any brokerage app and the default sort is “1-year return” or “5-year return,” descending. Open Morningstar or Yahoo Finance and the star ratings sit right next to the trailing-period performance. Open a financial news site and the headline funds are almost always the ones near the top of a recent leaderboard.
The implicit assumption is straightforward: a fund that just beat its peers has some edge — better managers, smarter allocation, a sector exposure that’s working — and that edge will persist. So the “safe” thing to do is buy what’s been working and avoid what hasn’t.
This belief is the cleanest possible expression of performance chasing, and it’s the mental shortcut that drives the largest fund flows in any given year. The 2021 cathode-ray ARK Innovation rush, the 2022 energy-sector chase, the 2024 mega-cap tech pile-in — every cycle, retail capital rotates toward whatever just won. The shortcut feels like prudence. Looking only at fresh evidence feels like updating your priors. But the data on whether recent performance actually predicts the next period’s performance is unusually clear, and unusually brutal.
What Recency Bias in Investing Actually Looks Like in Your Brain
Recency bias is the tendency to weight recent events more heavily than older ones when forecasting the future. In investing, it shows up in three measurable ways:
- Compressed memory of base rates. Markets have long been roughly 7–10% annualized real returns over multi-decade horizons. But after a year of 25% returns, that number starts feeling like the new baseline. After a year of –15%, the same brain treats 7% like a fantasy.
- Pattern-matching on small samples. Three months of outperformance starts to feel like “this strategy works.” Two years of a sector leading the market feels like a structural shift. The brain treats short streaks as signal even when statistical inference says they’re noise.
- Asymmetric attention to vivid evidence. A friend who tripled their money in a single name is louder in your memory than the silent hundred who lost on similar bets. A crash you lived through anchors your risk perception for years; one you read about in a textbook barely registers.
None of this is irrational in everyday life. If a restaurant has been good for the last three visits, going back is a fine bet. The problem is that financial markets aren’t restaurants. They’re competitive, mean-reverting, and dense with mispriced narratives. The mental machinery that helps you in low-stakes daily life is exactly the machinery that hurts you here.
I started watching for this bias in my own portfolio a few years back, mostly out of curiosity about whether I was as disciplined as I thought. As a software engineer, I lean toward systems and rules over gut, and the honest answer was: not as disciplined as I assumed. I had a documented investment plan and still found myself reading more about the funds that had recently done well, quietly nudging contributions toward them, and feeling slightly worse every time I looked at the laggards. That last part — the discomfort of holding the loser — is the engine driving the gap that Morningstar measures.
The Data That Debunks the “Hot Streak” Story
If recency-driven picks worked, we would see strong persistence in top-quartile fund performance. The data set that disproves this most cleanly is S&P Dow Jones Indices’ Persistence Scorecard, which has tracked actively managed US equity funds for two decades.
The latest scorecard, released in early 2025, finds that of the actively managed domestic equity funds in the top quartile of performance for the five years ending December 2019, less than 2% remained in the top quartile for the subsequent five-year period. Put more bluntly: being a top-quartile fund almost guarantees falling out of the top quartile. The phrase “past performance does not guarantee future results” isn’t boilerplate — it’s a literal description of what the data shows.
The S&P SPIVA report, also annual, lengthens the time frame. Over the 15 years ending in 2024, roughly 88% of all actively managed US large-cap equity funds underperformed the S&P 500 benchmark. Recency bias in investing isn’t just unprofitable inside fund picks — it’s unprofitable at the asset-class and strategy level too. The thing chasing wins is almost always a more aggressive position, levered into the past trend, exactly when that trend is most expensive.
Vanguard’s 2023 research piece on rebalancing offers the same conclusion from the opposite direction: portfolios that mechanically trimmed winners and added to losers across asset classes produced more consistent risk-adjusted returns than portfolios that let the recent winners run. The disciplined rebalancer was, almost by definition, betting against recency.
The Real Cost of Recency Bias in Investing: A 20-Year Comparison
What does a 1.3% annual gap actually compound into over a working life? The math is worth seeing on a single line.
The table below assumes two investors each contribute $500 a month for 25 years. The first sticks to a broad index fund earning the fund’s reported return. The second behaves like the average investor in the Morningstar data — same fund, but with timing decisions that shave 1.3 percentage points off the realized return each year.
| Investor | Annual Return | Total Contributions | Portfolio at Year 25 |
|---|---|---|---|
| Disciplined indexer | 7.3% | $150,000 | $437,000 |
| Performance chaser (recency-driven) | 6.0% | $150,000 | $359,000 |
| Gap (cost of recency bias) | — | — | $78,000 |
$78,000 of lifetime portfolio value, lost to a behavior that nobody consciously chose — that’s the cost most investors pay for letting recent returns drive contributions and reallocations. The pattern looks small year by year. It only looks expensive when you compound it.
This is closely related to the gambler’s fallacy in investing, where the brain reads streaks as signals about what comes next. Recency bias and the gambler’s fallacy run on the same underlying flaw — the assumption that small samples carry information they don’t.
Want to see what disciplined investing turns into over 25 years — without the recency gap?
What to Do Instead: Three Discipline Rules That Defeat the Bias
The defense against recency bias is not willpower — willpower fails at the worst possible moments, because every market cycle eventually produces a year where the disciplined plan looks stupid. The defense is structure: rules that take the in-the-moment decision out of your hands.
1. Mechanical Rebalancing on a Calendar
Pick a date — the first weekend of January is a common one — and on that date, rebalance back to your target allocation. If you set 80% stocks and 20% bonds, and stocks have run hot to 86/14, you sell stocks and buy bonds until you’re back at 80/20. If stocks have crashed to 74/26, you sell bonds and buy stocks. The action is the same regardless of how it feels.
Vanguard’s research finds the absolute level of returns barely changes between annual, quarterly, or threshold-based rebalancing — what matters is that some rule exists. The rule itself is the antidote, because every rebalance is, by construction, a bet against the recent winners.
2. Automated Contributions That Ignore the News
Dollar-cost averaging is the structural cousin of rebalancing. When the same dollar amount enters the market every two weeks regardless of headlines, recency loses its grip on the contribution decision. You buy more shares when prices are low (when recency bias would have you nervous) and fewer when prices are high (when recency bias would have you piling in).
This is the same principle behind the three-fund portfolio approach — remove the optionality, and you remove most of the behavioral cost. A portfolio you don’t touch can’t be steered into a recency-driven mistake.
3. A Written Investment Policy You Read Twice a Year
One page. What your target allocation is. What triggers a change (life events, not market events). What you will not do (chase last year’s top fund, time the market, sell during a drawdown). Read it in January and July. The recency bias loses much of its power once you’ve already committed, in writing, to what you would do under exactly the conditions you’re now facing.
The written plan also gives you a script for the conversation in your own head. When the temptation to chase shows up — and it always does — you don’t have to argue with it. You just point at the page.
This is the same pattern that defeats status quo bias in financial decisions in the other direction: structure beats willpower, and the rule beats the impulse. The two biases are mirror images — one tells you to chase movement, the other tells you to do nothing — and the same defense (a written rule) shuts both down.
Why This Specific Bias Is So Hard to Notice
Most cognitive biases announce themselves in obvious ways — you feel the urge, you ignore it, you move on. Recency bias is different because every decision it produces feels analytical. You’re looking at real data. The data is just from too short a window.
This is why the bias survives even after people learn about it. The brain doesn’t feel like it’s overweighting the recent past. It feels like it’s being responsive to fresh information — which is, in most contexts, a good intellectual habit. The fix has to live outside the moment of decision: in calendar-driven rules, in automation, in policies written down when you were calm.
The same principle applies to the way present bias affects retirement contributions. Both biases survive because the in-the-moment decision feels reasonable to the person making it. Both get neutralized by the same trick: pre-commit, automate, ignore.
Frequently Asked Questions
Is recency bias in investing the same thing as performance chasing?
Performance chasing is the behavior. Recency bias is the cognitive engine that produces the behavior. You can describe most performance-chasing fund flows — investors piling into last year’s top-quartile funds, fleeing the bottom-quartile ones — as the visible footprint of recency bias operating at scale. The Morningstar “Mind the Gap” data is essentially measuring the cost of recency bias across the entire mutual fund universe.
Doesn’t looking at recent returns at least help filter out bad funds?
It’s less useful than it feels. Recent returns are dominated by the fund’s sector exposure and the market’s recent regime — not by manager skill. A fund can be in the top quartile because it’s heavy on what just worked, and in the bottom quartile next year for the identical reason. The actual filters that have predictive power are expense ratio (lower is better, full stop), turnover (lower is generally better for tax efficiency), and tracking-error consistency for index funds. Those metrics are stable across time. Recent return is not.
What if a fund or sector has been outperforming for 10+ years — isn’t that a real signal?
It can be, but the honest answer is that a decade of outperformance also means a decade of valuation expansion — the fund or sector is now more expensive relative to its fundamentals than it was 10 years ago. The forward return is mathematically lower for the same earnings stream. This is the reason mean reversion shows up so reliably in long-horizon equity data: prices can’t outrun fundamentals forever. The 10-year track record looks like signal, but a meaningful chunk of it is just the multiple going up. Recency bias gets you to ignore that part.
Photo by Anne Nygård on
Unsplash
