Follow the Leader: Using the Stock Market to Uncover Information Flows between Firms
Anna Scherbina, Bernd Schlusche
Review of Finance, Volume 24, Issue 1, February 2020, Pages 189–225, https://doi.org/10.1093/rof/rfy038
News developments experienced by one firm, such as labor force scandals, intellectual property disputes, or geopolitical challenges encountered in other countries, can also impact other firms in similar circumstances. If investors are slow to react to relevant news of other firms, a firm at the forefront of a valuation-relevant news development may lead the returns of other firms until the issue is resolved. Consistently, this paper shows that an individual stock can have a collection of “bellwether” stocks that are able to forecast that stock’s return.
Each stock’s leaders are identified as the set of stocks whose returns have Granger-caused the stock’s returns over a trailing one-year window. After identifying the set of leaders for each stock, the leader signal is calculated by averaging the leaders’ returns in the current week. The paper shows that thus calculated leader signals predicts the followers’ returns in the following week. During the period 1996-2016, the strategy of buying stocks in the highest leader-signal quintile and selling short stocks in the lowest leader-signal quintile within each industry (thereby, forming industry-neutral portfolios) generated an annualized return of 16.12% for equal-weighted and 8.32% for value-weighted portfolios.
While previous literature has shown that stocks with high levels of investor attention can lead the returns of more neglected stocks by being first to react to common market- and industry-wide news, this paper demonstrates that leader stocks can have lower levels of investor attention and belong to a different industry than follower stocks. Moreover, by taking an agnostic approach on the direction of the information flows between firms, the methodology used in the paper can help uncover new channels regarding the cross-predictability of stock returns.
The results suggest that the return predictability is explained by an underreaction to relevant news of other firms. The predictability of leaders’ returns is stronger following weeks with a high overall news intensity, when investors are more distracted by competing news. Additionally, the predictive ability of smaller-stock leaders is stronger than that of larger-stock leaders because investors are more likely to overlook the relevant news developments of smaller stocks.
Given that the return predictability appears to have a behavioral explanation, it shares similarities with other return anomalies: It has declined over time, and it is stronger for smaller stocks with lower levels of attention from sophisticated investors. Moreover, the paper shows that short sellers appear to trade on the anomaly by increasing short positions in response to negative leader signals.