CEO Social Network Connections and Management Forecasts

Abstract

This study investigates whether CEOs who have more network connections are able to provide more accurate earnings forecasts. Prior literature documents that social network connections among individuals can serve as private information channels. CEOs who have more network connections can gather more information and resources needed to make accurate forecasts. However, CEOs may overreact to the private information they obtain, leading to more biased earnings forecasts. In addition, the private information travels via CEOs' network connections are not necessarily correct or useful. Overreaction and incorrect information could lead to less accurate earnings guidance. Using BoardEx Network Profiles with 4,508 firm-year observations from 2000 to 2011, we document that CEOs with more social network connections make less accurate earnings forecasts. This negative effect of social network is robust to various specifications, including controlling for factors affecting earnings forecasts, addressing concerns about crosssectional correlation using the Fama-MacBeth approach, and adopting the Newey-West procedure to account for heteroscedasticity and serial correlation. Our results suggest that CEOs' social networks negatively affect the accuracy of their earnings guidance.