Entry into Innovation Areas

Abstract

This paper investigates the multi-area entry decisions of listed U.S. electronics firms. I apply Latent Dirichlet Allocation to over 630,000 patents to define twenty innovation areas and merge them with market-implied valuations. I estimate a static entry game with moment inequalities. Firms enter areas where expected patent valuations are higher. Holding those valuations fixed, additional rivals reduce expected profit. Firms gain from occupying more areas at once. A large yearly per-area cost and rival congestion still limit expansion. From observed portfolios, a 25% reduction in that cost raises to 22-27% the share of firms that occupy an extra area.