Forecasting A-Share Returns with Gross Profitability Trends

Summary

This research note proposes a stock-selection factor based on the trend in quarterly gross profitability. It estimates the trend as the time coefficient in a linear regression, using either sequential-quarter or year-over-year observations. The study evaluates both versions with single-factor sorts, conditional two-factor sorts, and cross-sectional regressions. It reports stronger results for the year-over-year trend factor, including positive average IC and RankIC, a positive-month hit rate, and a monthly long-short return. The factor’s sorted returns reportedly remain monotonic after controlling for common factors.

The note also examines portfolio industry composition and compares the trend factor with gross-profit growth. It reports modest IC correlation with gross-profit growth but stronger correlation with the original gross-profitability factor, suggesting the trend measure is not simply a growth-rate substitute. The findings concern the A-share market and the document warns of model misspecification and changing factor effectiveness. The supplied text is an abstract rather than the underlying report, so details such as sample construction, transaction costs, and robustness checks are unavailable.

Key ideas

• The factor measures the time trend in quarterly gross profitability using a linear regression slope.
• The study compares sequential-quarter and year-over-year versions with sorts and cross-sectional regressions.
• The year-over-year trend version is reported to perform better than the sequential-quarter version.
• The factor’s reported relationship with gross-profit growth is weaker than its relationship with gross profitability itself.
• The evidence is specific to A-shares, and factor effectiveness may change over time.

Tags

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.