When it comes to momentum investing, everyone is always looking for a better way to implement a momentum-based stock selection strategy (the same goes for a value strategy). We highlight a few methods in our book, Quantitative Momentum, as well as on our blog. We recently came across a paper from 2006 that has an improvement on a baseline momentum investing strategy, titled “Firm-specific attributes and the cross-section of momentum.”
The abstract is below:
This paper identifies observable firm-specific attributes that drive momentum. We find that a firm’s revenues, costs, and growth options combine to determine the dynamics of its return autocorrelation. We use these insights to implement momentum strategies (buying winners and selling losers) with both numerically simulated returns and CRSP/Compustat data. In both sets of data, momentum strategies that use firms with high revenue growth volatility, low costs, and valuable growth options outperform traditional momentum strategies by approximately 5% per year.
The Theory of Firm-Specific Information and Momentum Investing
The paper looks to improve a cross-sectional momentum strategy by incorporating firm-specific information. The goal of the paper is to see if there are certain conditions that exist which will cause a higher autocorrelation of returns at the firm-level. Why does this matter? Here is a direct quote from the paper:
The profitability of momentum strategies is a cross-sectional result: winners realize higher average returns than losers. Suppose one can identify firms with time-varying return
autocorrelation and can restrict a momentum strategy to those firms whose autocorrelation is conditionally higher than average. Ceteris paribus, this restricted strategy results in enhanced profits because winners (losers) with relatively high autocorrelated returns have more persistent expected returns than winners (losers) from an unrestricted strategy.
