Peer effects of star-analysts' departure: New evidence from China

Abstract

This paper examines how the departure of star analysts influences the performance of their non-star peers. Using manually collected data from China, we observe that non-star analysts enhance their forecast accuracy after a star analyst leaves. This effect is particularly pronounced when non-star analysts have previously worked within hierarchical teams or when the departing star analyst held significant internal resources within the brokerage. Additionally, the performance improvements are more substantial in environments with greater promotion opportunities and in brokerages characterized by a collective organizational culture. To establish causality, we leverage the suspension and subsequent reform of the star analyst voting system as an exogenous shock. A difference-in-difference (DID) analysis demonstrates that brokerages more affected by this shock exhibit larger improvements in analyst forecast accuracy. These findings offer new insights into the celebrity effect, highlighting how changes in team structure and internal competition influence analyst performance.