When Everyone Trades the Same Factor Playbook

For decades, academic researchers have catalogued hundreds of patterns in the stock market — statistical regularities linking firm characteristics to future returns. These persistent return patterns, unexplained by standard risk models, are known as anomalies. They now form the intellectual backbone of a multi-trillion-dollar industry called factor investing, implemented through mutual funds, hedge funds, and ETFs worldwide.

The premise is straightforward: buy stocks with characteristics associated with high returns (the right, or “long leg”) and short stocks with characteristics associated with low returns (the left, or “short leg”).

Anders Posselt and Mads Kjær, authors of the March 2026 study, “Anomaly-Driven Demand,” examined what happens to market prices when millions of investors simultaneously follow the same mechanical rules to rebalance the same portfolios? If you allocate to factor strategies, this paper has interesting findings as to where your returns are actually coming from.

Research Design

Anomaly strategies are constructed by sorting stocks on a trait, or characteristic — say, book-to-market for value investing. Because those characteristics change over time, portfolio constituents are not static. Stocks drift in and out of the long and short legs every month as their characteristics update.

Each time a stock enters the long leg of a value strategy, value investors must buy it. When it exits, they must sell. This is mechanical, predictable rebalancing — analogous to what happens when a stock is added to or removed from an index like the S&P 500.

Now extend that logic across to what researchers have dubbed a “zoo’ of anomalies” — hundreds of different published anomalies simultaneously. A stock that enters the long leg of many anomalies at once is being targeted by a large, diverse population of factor investors all rebalancing in the same direction, at approximately the same time. The authors call this cumulative buying (or selling) pressure Anomaly-Driven Demand (ADD).

With ADD defined, the authors then tested whether it actually predicts returns.

You can read the rest of my Alpha Architect article here.