Slow Diffusion of State-Level Information and Return Predictability

We use a new disclosure-based approach to show that value-relevant information about publicly-traded firms is geographically distributed within the United States and the market is slow in aggregating this information. Firm fundamentals such as earnings and cash flows can be predicted using the fundamentals of other firms in economically relevant U.S. states, but sell-side equity analysts and institutional investors do not fully incorporate this information in their earnings forecasts and trades, respectively. Consequently, a Long−Short trading strategy that exploits the slow diffusion of geographic information earns an annual, abnormal risk-adjusted return of about 9%.