Evaluating Chinese Patent Data as Equity Selection Factors
Summary
This study examines a database of 486 patent-derived factors for Chinese listed companies and tests their usefulness in stock selection. It organizes patent information by type, procedural status, metric, and lookback period. Measures include patent counts, claims, citations, international family filings, and technology lifecycle. The report explains that basic patent counts tend to be more widely covered, while detailed measures such as foreign family patents are sparse. Patent measures also vary strongly by industry and company size, making unadjusted single-factor comparisons potentially misleading.
The authors rank the factors with backtests and discuss how patent quantity, citations, international filings, growth, and research efficiency may reflect different aspects of innovation. Results vary widely: the reported Sharpe ratios range from 0.043 to 0.536, and some patent family measures appear more informative than simple counts. The evidence is preliminary and largely cross-sectional or historical; observed returns may reflect industry and size exposures. The report recommends neutralization, combining patent measures with other company data, and further model testing before treating the factors as robust signals.
Key ideas
- Patent factors measure different aspects of corporate innovation, including counts, claims, citations, and international patent families.
- Basic patent counts have broader coverage, while detailed patent measures can be sparse.
- Patent data are strongly associated with industry and company size, which can confound single-factor tests.
- The reported factor backtests show mixed performance and require further validation and neutralization.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.