Information Environment and Return Predictability of Conglomerates

This paper examines how the change in information environment affects investors’ ability to process complex information. To verify this, a return of pseudo conglomerate is created using the methodology of Cohen and Lou (2012), who recognize conglomerate’s return as a portfolio return and construct the value-weighted return based on the sales of its industry segments. This study verifies that the return of Pseudo conglomerate in the past well explains the return of current conglomerate. Unlike standalone firms, conglomerates are composed of several segments, so it takes a long time to determine how information affects a firm’s value as a result of high information processing complexity. In this case, we expect that the predictability of the past Pseudo conglomerate’s returns is expected to be high. This paper also investigates whether the quality of the information environment, which affects the complexity of information processing, affects the predictive power of conglomerate’s returns. As a result, the return of a standalone firm can predict the return of a conglomerate and the predictive power of Pseudo conglomerates’ returns decreases as the information environment improves. This study sets the passage of the Sarbanes-Oxley Act of 2002 (SOX), which improves the quality of information and finds that the predictive power is significantly reduced to less than half in the post-SOX. On the contrary, explanatory power increases during periods when investors’ distraction is high due to news outside the capital market. Even after considering the conventional risk factors as control variables, the information effects are significant, confirming the Pseudo conglomerate return’s robustness.