Abstract
This paper investigates the impact of managerial bad news hoarding on the quality of financial analyst forecasts, using a panel dataset of US public firms spanning from 1995 to 2022. Using stock price crash risk to measure managerial bad news hoarding, we document a positive association with the quality of analyst forecasts in terms of greater accuracy and lower forecast dispersion. This association is mainly driven by the intensified herding behavior exhibited by the financial analysts when forecasting firms with higher stock price crash risk. Such analysts’ herding is accompanied by reduced forecast dispersion and also facilitates the aggregation of individual analysts’ advantageous information, leading to improved forecast accuracy. Despite the benefits of herding on improving analysts’ forecast quality, we further show that analysts tend to exhibit a more biased and optimistic view towards firms exposed to higher crash risk. This heightened forecasting bias is due to the anchoring effect, a cognitive bias where analysts would rely excessively on a reference point rather than adjusting their forecasts sufficiently to the new information presented. Our results remain robust across a rich set of econometric specifications, accounting for potential endogeneity and alternative measures of independent and dependent variables.