Abstract
Value and momentum are among the most-studied factor anomalies in equity markets. Both have been documented across decades, regions, and asset classes, and both feature prominently in the systematic equity toolkit. A common practical choice when implementing them is to merge the two scores into a single composite ranking used to select stocks for a portfolio.
This working paper argues that combining value and momentum at the signal level is structurally counterproductive. The two anomalies target fundamentally different parts of the cross-section. Value tilts toward stocks the market has marked down. Momentum tilts toward stocks the market is currently pushing up. Blending them into one ranking forces them to compete for the same name on the same date, and the resulting portfolio inherits a diluted signal rather than the sum of the two.
The empirical work compares two implementations side by side. The first merges value and momentum into a single composite score at the signal level. The second runs each factor as a separate portfolio sleeve and then weights the two sleeves together. The volatility-matched comparison consistently favors the second design across the equity universes and rebalancing horizons tested. The paper documents the mechanism behind that difference, the conditions under which it strengthens or weakens, and the practical implications for portfolio construction.
The takeaway for systematic equity investors is to treat value and momentum as separate building blocks rather than as ingredients of a single recipe.