Text-Based Network Industries and Endogenous Product Differentiation

We study how firms differ from their competitors using new time-varying measures of product differentiation based on text-based analysis of product descriptions from 50,673 firm 10-K statements filed yearly with the Securities and Exchange Commission. This year-by-year set of product differentiation measures allows us to generate a new set of industries and corresponding new measures of industry competition where firms can have their own distinct set of competitors. Our new sets of industry competitors better explain specific discussion of high competition by management, rivals identified by managers as peer firms and firm characteristics such as profitability and leverage than do existing classifications. We also find evidence that firm R&D and advertising are associated with subsequent differentiation from competitors, consistent with theories of endogenous product differentiation