Evaluating Patent Data as Equity Selection Factors
Summary
The document surveys a platform’s patent database and describes how patent characteristics are turned into equity-selection factors. It groups patent information by patent type, application or grant status, and measures such as counts, claims, citations, family size, and document attributes, then combines those measures with time dimensions. The study proposes ranking and backtesting the resulting factor set.
It emphasizes that patent counts are strongly related to industry and company size, which can make a single-factor portfolio unfairly favor certain sectors or large firms. Basic patent measures may not capture genuine innovation; overseas patent-family filings are presented as a potentially more informative signal because they are harder and costlier to obtain. The text gives a research framework and qualitative observations, but no detailed performance statistics or test setup. It is an older implementation intended for study, and the planned work on neutralization, factor combinations, and machine-learning applications is not presented as completed analysis.
Key ideas
- Patent features can be grouped by patent type, process status, measure, and time period to form equity factors.
- The study proposes ranking and backtesting a broad set of patent-derived factors.
- Patent counts are associated with industry and company size, creating potential portfolio exposures.
- Overseas patent-family filings are suggested as a more selective proxy for valuable innovation.
- The document gives limited test details and describes the implementation as an older version for study.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.