Analysts' Herding Propensity: Theory and Evidence from Earnings Forecasts

We model and estimate analysts' herding propensity with I/B/E/S annual earnings forecast data. Compared to prior studies, our paper has three unique features. First, we directly estimate analysts' true posterior beliefs of a firm's earnings rather than using their own prior forecasts as a proxy. Second, we provide evidence on analysts' herding behavior at the analyst and more aggregate levels rather than the forecast level. Third, we test the usefulness of our herding propensity estimates out-of-sample rather than insample because out-of-sample results have higher practical value. We document pervasive herding behavior. At the aggregate levels, we find that herding propensity varies positively with forecast horizon and analyst coverage, but negatively with analysts' general experience and brokerage size. At the analyst level, we find that about 75% (15%) of the analysts in our sample tend to herd (anti-herd). Our herding propensity cannot be explained by the walkdown-walkup patterns of analysts' earnings forecasts. Finally, we show that our herding propensity estimates are useful in explaining the cross-sectional variation in analysts' out-of-sample herding behavior and forecast accuracy.