How Much Do Analysts Influence Each Other's Forecasts?
This paper develops and applies a new approach for disentangling the influence of analysts on each other's earnings forecasts from the effects of correlated information shocks. We estimate that, on average, each cent a new forecast by an analyst is above (below) another analyst's most recent forecast causes the other analyst to revise her forecast upwards (downwards) by between 0.21 and 0.36 cents. More reputable analysts are more influential, while those that tend to be optimistic are less influential and are influenced more by the forecasts of other analysts. We do not find support for career concerns-driven herding or anti-herding. Finally, we find that more influential analysts are more likely to subsequently be ranked as All-Stars and to move from a less to a more prestigious brokerage house, and less likely to leave the analyst profession, suggesting that influence is a desirable characteristic.