Dividend Policy in the Era of Big Data

The releases of real-time satellite data of U.S. retail firms’ parking lot traffic reduce information asymmetry between managers and outside investors. Using the staggered releases of satellite data as a quasi-natural experiment, we test the competing dividend theories based on information asymmetry and agency costs. We find that retail firms substantially increase dividend payouts after their satellite-based traffic data are released, and the increase in dividends is concentrated in firms with poor investment opportunities. Further analyses show that the effect of satellite data release is stronger when firms have more entrenched managers, less severe financial constraints, or higher ownerships by sophisticated investors. Additionally, we find that firms finance the dividend increase by reducing low-quality investment while their high-quality investment (R&Ds) remains intact. These results support the “outcome model” that dividend payout is a complement of corporate governance, and show that big data can have substantial effects on firms’ corporate policies.