Conflict of Interest or Information Sharing? Evidence from Affiliated Analyst Performance in Korea*

Abstract

This paper examines the accuracy and optimism of forecasts made by affiliated analysts compared to those made by independent analysts. Using Korean chaebol firms, we test the information-sharing hypothesis and the conflict of interest hypothesis. Our results show that the forecasts made by the affiliated analysts are less accurate and more optimistic than those made by the independent analysts. This finding is inconsistent with the information-sharing hypothesis but supports the conflict of interest hypothesis. The results also indicate that the forecast accuracy of affiliated analysts is not related to the internal sales transaction ratio or the idiosyncratic information of the group firms. We also find that the forecast accuracy of affiliated analysts increased after Regulation Fair Disclosure came into effect in 2002 in Korea, which finding is in disagreement with the information-sharing hypothesis. Our results reveal that the forecast inaccuracy and optimistic bias of the affiliated analysts are not directly related to the individual ownership relationship of the affiliated group firms with the forecasting firm. Our results provide further evidence of analysts' conflict of interest that arises from earnings-related intragroup propping behavior.