Using Patent Citation Patterns to Infer Innovation Market Competition

Abstract

We propose an empirical strategy for estimating competition in innovation markets based on a model of creative-destruction. Our method relates firms’ market return on equity to information about patent citation patterns. Two innovations we introduce are using daily abnormal stock returns rather than annual measures of Tobin’s q and creating citation patterns related to the area of science a firm patents in as represented by the detailed patent classification system. We find that firm’s market value increases when its patent portfolio is cited and when there are citations to an area of science in which their patent portfolios are prominent. Holding this effect constant, we find that citations from patents in the same area of science tend to reduce market value. We interpret these findings as consistent with more citations indicating more valuable intellectual property but citations from competing technologies decreasing its value.