Factor Momentum II

Does chasing factor performance work?

August 2026. Reading Time: 10 Minutes. Author: Nicolas Rabener.

SUMMARY

  • Chasing factor performance is a viable strategy
  • Provides diversification benefits given a low correlation with equities
  • Some skepticism is warranted given a lack of successful funds

INTRODUCTION

In August 2018, we explored whether chasing factors like value or quality in the stock market was a profitable investing strategy. After all, most investors only allocate to factors once returns have started to look attractive. Our analysis focused on 25 equity factors, where we constructed a long-short portfolio by selecting the top and bottom 5 factors based on their performance over the last 12 months. We explored this across markets, and while returns were positive from 2002 to 2018, performance was inconsistent. However, there is no reason investors should limit themselves to equity factors: trend-following funds, also known as CTAs, pursue the same strategy using futures across asset classes.

In their 2023 paper “The Co-Pricing Factor Zoo“, Alexander Dickerson, Christian Julliard, and Philippe Mueller analyzed 18 quadrillion models for the joint pricing of bonds and stocks but identified only around 50 factors as robust sources of risk. The authors made this factor dataset publicly available, which we will use to further explore factor momentum.

FACTOR MOMENTUM

The dataset from Dickerson et al. covers 54 factors, categorized into equity factors, fixed income factors, and non-tradable factors. This includes well-known factors such as stock-based momentum and bond carry, as well as more esoteric ones like behavioral and inflation volatility. Returns are available from 1987 to 2022.

The excess return from the U.S. stock market is included as a factor, but we don’t want to trade equities, since there is already ample research on this. We therefore exclude all factors with a correlation greater than 0.75 with the stock market, reducing the universe to 47 factors.

We create an index by selecting the top 5 performing factors based on the trailing 12-month performance for the long portfolio, while simultaneously shorting the five worst-performing factors. The portfolio is rebalanced monthly, and we as