Why Factor Performance Decays: Understanding Alpha Decay and Crowding in Long-Term Investing

The momentum factor delivered approximately 10% annual returns in the 1990s. Today, that figure is closer to 2%. Value, quality, and size have followed similar trajectories. This isn't market noise or a statistical fluke - it's a predictable consequence of success. As investors discover and copy profitable factor signals, the returns erode. Understanding this dynamic is critical for anyone building a long-term investment strategy.

What Are Factors and Why Did They Work?

Factor investing is an investment approach that involves targeting quantifiable factors that can explain differences in security returns. Size and value are two of the five most widely accepted factors, with their focus on small-cap companies and relatively cheap securities, respectively. Academic research documented these premiums extensively. Since 1963, the size factor has captured 208.86% cumulative returns, which translates to 1.87% annually.

The value factor has shown outperformance, achieving an annualized return of 2.92% from 1963 to the present.

These weren't massive returns, but they were measurable, persistent, and - until recently - ignored by most investors. The early research advantage is gone. A growing body of evidence documents the decay of factor premia following academic publication. McLean and Pontiff found that approximately 50% of anomaly alpha disappears post-publication, consistent with investors learning from research and arbitraging away returns.

How Factor Crowding Works

Factor crowding occurs within quantitative equity portfolio management. As factor investing gains popularity, the symptoms of crowding have become increasingly pronounced. The mechanics are simple: as more capital chases the same signal, competition for finite profit erodes everyone's returns.

When N agents discover and trade the same profitable signal, they compete for a fixed "alpha capacity" K. In Nash equilibrium, each agent earns alpha equal to K divided by N. The math is brutal. If 100 investors discover a factor that generated 5% annual alpha, and they all trade it identically, that 5% splits into 0.05% per investor (before costs).

Significant signs of crowding exist in well known equity signals, such as Fama-French factors and especially Momentum. But not all factors decay at the same rate.

The Difference Between Mechanical and Judgment Factors

Not all factors crowd equally. The model fits "mechanical" factors - those with unambiguous, easily replicated signals like momentum ("buy recent winners") - but fails for "judgment" factors like value, where the signal ("what is cheap?") admits multiple interpretations.

Momentum is mechanical: past performance is objective and easy to implement. Once published, institutional money flooded in, and the premium compressed. Value is stickier. Two analysts can disagree sharply on whether a stock is "cheap" or a "value trap," creating persistent disagreement and higher barriers to full crowding.

Factors with high barriers to entry are likely to maintain their performance over time and are less prone to crowding-induced declines in efficacy. Conversely, factors with low barriers to entry are prime candidates for crowding arbitrage.

Quantifying the Decay

Research using game-theoretic modeling reveals the precise shape of decay. For momentum, hyperbolic decay achieves R² of 0.65, outperforming linear decay (0.51) and exponential decay (0.61). This suggests decay accelerates early - the steepest drops happen first - then flattens.

More concerning: Crowding accelerated post-2015. Training on 1995-2015 and predicting 2016-2024, the model over-estimates remaining alpha (0.30 predicted vs. 0.15 actual). The rise of factor ETFs has democratized factor investing. Over-prediction of alpha correlates with factor ETF volume growth, suggesting that democratization of factor investing amplified crowding.

When Crowding Cuts Both Ways

The type of factor determines whether crowding helps or hurts. Factor premiums are classified as divergent (such as momentum), inherently destabilizing due to positive feedback loops and lack of fundamental anchors; or convergent (such as value), having self-correcting negative feedback loops and fundamental anchors.

For divergent factors like momentum, crowding is catastrophic. Everyone piling in to "buy winners" pushes prices up artificially, eventually leading to a reversal when the crowd exits. For convergent factors like value, crowding can create the opposite effect: if everyone's pushing prices down by selling "cheap" value stocks, those stocks become even cheaper, actually reinforcing the fundamental case.

Implications for Your Portfolio

Understanding factor decay changes how you build a portfolio. Pure factor chasing - rotating into the most crowded signals - is now a losing game. The premiums that delivered consistent returns for 40 years have compressed and become fragile.

More reliable for long-term investors is diversification across uncrowded sources of return. As long as asset classes do not exhibit perfect positive correlation, owning exposure to both will inherently yield diversification benefits. Positive correlation does not negate the benefits of diversification; meaningful benefits exist even when correlation is consistently above zero.

This doesn't mean ignore factors entirely. Over the long term, factors have delivered a clear and consistent premium to the broad global equity market, even as short-term variability increases. Rather, it means:

• Avoid pure factor concentration. Don't allocate 50% to momentum or value alone.

• Prefer judgment-intensive factors. Value, quality, and profitability have higher barriers to entry than momentum.

• Combine multiple factors. Because some factors are not correlated, combining two or three can improve performance.

• Rebalance sparingly. Frequent rebalancing to chase factor tilts incurs trading costs that offset any premium. Rebalancing quarterly or monthly produced no improvement in long-term risk or returns; it simply drove up the turnover rate and the number of rebalancing events (and potential transaction costs!).

The markets have learned. Factor premiums still exist, but they're smaller, more fragile, and less forgiving of poor execution. Long-term investors need to adapt accordingly.

Prop firm rules change frequently - always confirm the current rules with your firm. Trading futures involves substantial risk of loss.

References

• Multi-Factor Investing: Beyond the Traditional Factors | Pacer ETFs
• Size vs value: what to consider when factor investing | EBI
• How correlation impacts diversification in investing | Saxo
• Effects of Factor Crowding | CXO Advisory
• Portfolio Rebalancing: When, Why, And How Much It Matters | Calculatorian

Key definitions

Factor investing - An investment approach targeting quantifiable characteristics (such as size, value, or momentum) that explain differences in security returns across a portfolio.

Factor premium - The excess return earned by securities exhibiting a particular factor characteristic, measured relative to the broad market over a defined period.

Factor crowding - The erosion of factor returns as increasing capital competes for the same profitable signal, dividing available alpha among more participants.

Mechanical factor - A factor with an unambiguous, easily replicated signal (such as momentum based on past price performance) that is quick to arbitrage once identified.

Judgment factor - A factor requiring subjective interpretation of the underlying signal (such as value, where "cheapness" admits multiple definitions), creating persistent disagreement and higher barriers to crowding.

Divergent factor - A factor that lacks fundamental anchors and is subject to positive feedback loops, such that crowding amplifies price movements and eventual reversals (example: momentum).

Convergent factor - A factor anchored to fundamentals and subject to negative feedback loops, such that crowding reinforces rather than undermines the underlying signal (example: value).

Alpha decay - The decline in excess returns following academic publication or widespread investor adoption of a trading signal.

Educational research on historical data only - not investment advice, not a signal, and never a performance promise. Past results do not predict future performance. Drafting uses AI assistance; every citation is link-verified before publication and every paper is re-audited weekly against the library's editorial standard. Last reviewed by the PropLedger research pipeline: 2026-08-26. Educational research on historical data; not financial advice.