Financial Similarity Momentum for Chinese Equity Selection

Summary

This research proposes a stock selection factor that estimates similarity between companies using ten financial measures spanning profitability, solvency, operating efficiency, and growth. After cross-sectional normalization, it uses cosine similarity to weight the monthly returns of financially similar firms, producing a financial momentum signal. The authors also neutralize the signal for industry, size, and monthly reversal effects.

The report presents historical tests on Chinese equities, including information coefficients, ranked portfolios, factor correlations, and Fama–MacBeth regressions. It reports that the factor had low correlation with many conventional factors and showed stronger performance around financial reporting periods. Combining it with a benchmark financial factor or an AdaBoost-based factor also improved selected test metrics. These results are historical and depend on the authors’ sample, construction choices, and assumptions; the report says the cause of the reporting-season effect needs further investigation and warns that market or policy changes may weaken the signal.

Key ideas

• The factor measures cosine similarity across ten normalized company financial indicators.
• It weights peer firms’ returns by financial similarity to estimate a stock-level momentum signal.
• The study reports low correlation with many established factors and positive historical selection results.
• The signal performed more strongly around financial reporting periods in the reported tests.
• Historical backtests do not establish that the effect will persist in changing markets.

Tags

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