How to characterize patent quality with multiple indicators? Evidence based on economic performance of Chinese companies

We establish a framework for using pre-grant and post-grant patent quality indicators to construct composite measures by linking data from the China National Intellectual Property Administration with incoPat. Multicollinearity problems usually arise when various patent quality indicators are included as explanatory variables in regression models. Constructing a composite patent quality index (PQI) and its sub-indices helps to not only avoid multicollinearity problems but also consolidate the information conveyed by multiple indicators of patent quality. We show that these composite indices retain essential information regarding multiple patent quality indicators and offer easy comparisons among patents. Moreover, we find that patent stocks weighted by PQI and its sub-indices are positively and significantly associated with Chinese listed firms' revenue performance or market value.