Recency Bias: Why Last Year's Best Strategy Loses Next Year
The Invisible Filter That Distorts Your Judgment
Ask yourself: which feels more "real" — a backtest showing 30 years of data, or your personal experience of the last three years? For almost every human being, the answer is the same. The recent experience wins. It feels more vivid, more relevant, more trustworthy. The 30-year backtest is abstract. The last three years are something you lived.
This is recency bias — the tendency to overweight recent experience when making decisions about the future. It is one of the most pervasive and costly cognitive biases in investing, and it operates so smoothly that most investors don't even notice it shaping their decisions.
How Recency Bias Works
Human memory is not a neutral recording device. It weights recent events more heavily than distant ones. This is generally adaptive — in most areas of life, recent experience is a better predictor of the near future than distant experience. The restaurant that served you a bad meal last week is probably still serving bad meals. The route that was under construction yesterday is probably still under construction today.
But financial markets don't work this way. In fact, markets exhibit a powerful and well-documented tendency toward mean reversion at the strategy level — meaning that strategies and asset classes that outperformed in the recent past tend to underperform in the near future, and vice versa. Recency bias takes the one domain where recent experience is systematically misleading and treats it as the most reliable guide.
The Pattern: Chase, Suffer, Abandon, Repeat
Recency bias creates a devastating cycle that plays out over and over across market history:
- A strategy or asset class outperforms for 2-3 years. It attracts attention, media coverage, and capital.
- Investors pile in, attracted by the recent track record. "This strategy returned 25% per year for the last three years — clearly it's the best approach."
- The outperformance ends (or reverses) due to mean reversion, changed conditions, or crowding.
- Investors suffer underperformance for 1-2 years, then abandon the strategy — usually at the point of maximum underperformance.
- They move to whatever has been outperforming recently, and the cycle repeats.
The net result: the investor systematically buys high and sells low at the strategy level. They hold each strategy during its worst period and miss its best period.
Real-World Examples
Value vs. Growth: 2017-2023
From 2017 through 2020, growth stocks crushed value stocks. The Russell 1000 Growth index returned over 130% cumulatively, while the Russell 1000 Value index returned roughly 40%. Recency bias told investors that growth was the only strategy worth holding. Value investing was declared "dead" by numerous publications.
Investors who abandoned value in 2020 missed a sharp reversal. In 2022, the Russell 1000 Growth fell approximately 29%, while the Russell 1000 Value fell only 8%. The "dead" strategy outperformed by over 20 percentage points in a single year. But by then, most recency-biased investors had already sold their value positions and concentrated in growth.
Tactical vs. Buy-and-Hold: 2010-2019
The 2010-2019 decade was one of the strongest and most uninterrupted bull markets in history. Buy-and-hold in the S&P 500 returned approximately 13.5% annualized. Tactical allocation strategies, with their built-in defensive mechanisms, captured less of the upside — perhaps 9-11% annualized, depending on the strategy.
Recency bias told investors that tactical allocation was inferior. "Why pay the cost of drawdown protection in a market that only goes up?" Many abandoned their tactical approaches for passive buy-and-hold during 2018-2019.
Then 2020 arrived. The S&P 500 fell 34% in 23 trading days. Then 2022 brought another 25% decline over several months. The investors who had abandoned tactical protection based on the last decade's smooth sailing suffered the very drawdowns they had originally sought to avoid.
Commodities: The 2022 Trap
In 2022, commodities were one of the few asset classes delivering positive returns. Energy stocks soared. Gold held up well. Agricultural commodities spiked. Recency bias drove a wave of investors into commodity-focused strategies. "Commodities are the new asset class to own. Stocks and bonds are done."
In 2023, commodities broadly declined while stocks rallied strongly. The investors who shifted into commodities based on 2022's performance experienced the classic recency-bias outcome: they bought the outperformance at its peak and held through the reversion.
Mean Reversion: Why Recent Winners Tend to Become Future Losers
The phenomenon recency-biased investors fail to account for is mean reversion — the tendency for extreme performance (both good and bad) to normalize over time. This operates at multiple levels:
| Level | Mechanism | Typical Cycle |
|---|---|---|
| Asset classes | Outperformance attracts capital, raising valuations to unsustainable levels | 3-7 years |
| Sectors | Hot sectors become crowded, margins compress, expectations reset | 2-5 years |
| Strategies | Crowded strategies lose edge; unfashionable strategies regain it | 3-10 years |
| Fund managers | Top-quartile managers regress toward the mean as conditions change | 1-5 years |
This is why chasing last year's best-performing strategy is so reliably destructive. You're not buying a proven winner — you're buying an asset or strategy at the moment when reversion to the mean is most likely.
Why Blending Strategies Beats Picking the "Best" One
If individual strategies cycle between outperformance and underperformance, the logical implication is powerful: you shouldn't try to pick the single best strategy. You should blend multiple strategies with different characteristics.
A blend of strategies that respond differently to different market conditions — some that thrive in trending markets, some that excel during mean-reverting conditions, some that emphasize defense — creates a portfolio with more consistent returns. When one strategy is in its "off" cycle, others are in their "on" cycle.
Blending doesn't eliminate underperformance — no approach does. But it dramatically reduces the magnitude and duration of drawdowns, which in turn reduces the emotional pressure that drives recency-biased strategy-switching. When your blended portfolio has a modest down year instead of a devastating one, the temptation to abandon it is manageable.
The Antidote: Systematic Lookback Periods
Recency bias operates by letting the most recent 1-3 years of experience dominate your decision-making. Systematic strategies counter this by using predefined lookback periods that are calibrated, not to what feels right, but to what has been empirically shown to predict future returns.
Momentum strategies, for example, typically use 6-12 month lookback periods — long enough to capture genuine trends, short enough to be responsive. These lookback windows are set in advance based on decades of research, not on whatever happened last quarter. The strategy doesn't "know" or "care" that growth stocks outperformed value stocks over the last three years. It evaluates current momentum over its predefined window and allocates accordingly.
This is a fundamental advantage of rules-based systems: they are immune to the recency bias that plagues human decision-makers. The system's "memory" is exactly as long as it was designed to be — no longer, no shorter. It doesn't matter how vivid or emotionally charged recent market events were. The lookback period doesn't stretch or shrink based on how you feel about recent returns.
Committing to the Process, Not the Outcome
The deepest challenge of recency bias is that it makes process-oriented investing feel wrong. When your strategy underperforms for 1-2 years, every fiber of your being says the strategy is broken and you need something different. Recency bias makes the underperformance feel permanent — as if the strategy has fundamentally lost its edge.
But the historical record is clear: every worthwhile strategy has multi-year periods of underperformance. The strategies that deliver the best long-term results are often the ones that are hardest to hold during their "off" periods. Recency bias is the specific mechanism by which investors abandon good strategies at the worst possible time.
The solution isn't more discipline. It's building an investment process that doesn't require you to fight recency bias with willpower. Commit to a blend of strategies, execute the monthly signals, and accept that some strategies will lag in any given year. The blend ensures you're never fully exposed to any single strategy's off-cycle. And the systematic signals ensure that your decisions are based on data, not on the distorted lens of recent experience.