Cross-Firm Information Flows and the Predictability of Stock Returns
Anna Scherbina, UC Davis
Bernd Schluschey, Federal Reserve Board

The authors identify leader stocks based on their ability to Granger-cause returns of other stocks and show that such leaders can reliably predict their followers’ returns out of sample. Leaders’ predictive ability is robust to firm- and industry-level controls and works at the level of individual stocks rather than industries. Many leaders cannot be easily detected using ex-ante firm characteristics: They often are small, belong to a different industry than their followers, or exhibit only a short-lived leadership. The authors find support for the conjecture that leaders tend to be at the center of important news developments that also affect their followers by showing that, all else equal, firms with greater news coverage have a larger number of followers. We furthermore find that, consistent with the view that equilibrium mispricing is related to arbitrage costs, more heavily traded stocks react to their leaders’ signals with a shorter delay. Finally, we present evidence that sophisticated investors trade on leaders’ signals.

Understanding how information flows and ultimately impacts asset prices can help investors better understand price dynamics. The results of this study help identify the impact of media coverage on price formation. They also can help investors better decide when it is most important to trade quickly.