Patent Approval Factors for Stock Selection and CSI 500 Enhancement
Summary
This report studies whether company patent data can help predict stock performance. It compares factors based on patent counts and types, including invention patents, utility models, and design patents. The report treats granted invention patents as the most informative type because they have passed examination and received legal authorization. Patent coverage is reported as higher in manufacturing and technology-related industries, with several named sectors exceeding 70% coverage.
In backtests, one-year counts of granted invention patents and the number of associated companies show predictive and stock-selection potential. Applied to CSI 500 constituents, the patent-count factor produces reported annualized excess return of 4.20%, with relative volatility of 5.08% and information ratio of 0.83; coverage among constituents is 62%. A research-efficiency factor combining R&D expense with granted patent counts is less stable and suffers a notable drawdown from early 2017. Results are specific to the report’s data, universe, and test period; the summary does not establish that patent signals will persist or generalize.
Key ideas
- Granted invention patents are presented as more informative for stock selection than other patent types.
- One-year granted invention patent counts and related-company counts show predictive potential in the reported tests.
- Patent data coverage varies by industry and is relatively high in several manufacturing and technology sectors.
- The patent-count factor reportedly enhances the CSI 500, while the R&D-efficiency factor is less stable.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.