Calibrating Innovation Proxies with an Accurate Tobin’s Q and Appraised Innovation Value

Researchers have used patent counts, citation weighted patent counts, as well as research and development spending to measure innovation with, at times, conflicting results. We benchmark the validity of these innovation proxies using a novel data set of appraised tangible and intangible assets. This new approach enables the correct replacement cost specification for Tobin’s Q and provides an appraisal of innovation assets. Additionally, we include trademarks as a new proxy to capture un-patentable and undisclosed innovation. We find that two innovation proxies, the market response to new patents and trademark counts, are both significant and consistent predictors of a corrected Tobin’s Q and the value of innovation assets. However, our results indicate that three other proxies — patent count, citation-weighted patent count, and research and development expenditure — are not robust measures for capturing innovation