Analyst following and count-data econometrics

This paper reexamines the determinants of the number of analysts following a firm using econometric models based on count distributions. We replicate Bhushanʹs (1989) analyst-following study to demonstrate the effects of using count-data econometrics, in lieu of OLS, in studying phenomena where the dependent variable ranges among nonnegative integers. In contrast with the original paper, our findings indicate the number of institutional investors is inversely related with analyst following. We also provide econometric evidence to support the preferred use of the negative binomial model in estimating cross-sectional, analyst-following regressions.