Combining Short-term Trading Signals

Should investors dismiss short-term signals as unexploitable due to high trading frictions? In their May 2022 paper entitled “Beyond Fama-French Factors: Alpha from Short-Term Signals”, David Blitz, Matthias Hanauer, Iman Honarvar, Rob Huisman and Pim van Vliet investigate whether investors can extract a material net alpha by applying efficient trading rules to a composite of several short-term signals from a liquid global universe of stocks. The composite signal each month normalizes and averages the following individual signals:

1. Industry-relative 1-month reversal
2. 1-month industry momentum
3. Analyst earnings revisions over the last 30 days
4. Same-calendar month average return over the past 10 years
5. 1-month daily idiosyncratic volatility

They hypothesize that integrating signals with low correlations offers diversification benefits, thereby boosting gross returns and suppressing volatility. They each month rank stocks on individual and composite signals into fifths (quintiles) and measure alphas of equal-weighted quintile portfolios using a 6-factor (market, size, book-to-market, profitability, investment and momentum) model of stock returns. They then consider a more efficient trading strategy is each month long (short) stocks currently in the top (bottom) X% plus the stocks as selected in previous months and still among the top (bottom) Y% of stocks, with X=20/Y=20 representing conventional full quintile rotation and X=10/Y=50 representing an efficient trading rule. They assume 0.25% average 1-way trading frictions. Using daily prices and end-of-month signal data for all stocks in the MSCI World Index and regional 6-factor model returns during December 1985 through December 2021, they find that:

(more…)