The effect of Innovation Similarity on Asset Prices: Evidence from Patents’ Big Data

Abstract

Through textual analyses of 7.7 million patents, we develop a novel intercompany innovation similarity measure which enables us to find that technologically connected firms cross-predict one another’s returns. Investors impound information about firms’ technological connectedness, although not immediately and fully. Buying (shorting) shares of technological peers earning high (low) returns during the previous month yields a 1.29% monthly return. Firms’ return predictability increases with patent complexity or limited technological disclosures but decreases with better information transparency. These results suggest investor inattention explains technology momentum. Unlike momentum stemming from simpler class-based technological links, our Big Data text-based return predictability remains active.

Keywords: Big Data; Textual analysis; Patents; Innovation similarity; Investor attention; Return predictability; Technological diversification

JEL Classification: G11; G12; G14; O31; C55