Do Behavioral Biases Influence the Length of Sell-Side Analysts' Observable Careers?

Analysts' observable careers are short. Around half of the analysts we study have observable lives of approximately three years. Using the IBES database as a proxy for the length of analysts' careers, we find that distraction and herding shorten observable lives. There are important roles for economically rational influences. The length of observable careers is reduced by inaccuracy in forecasting. Issuing a first forecast early but, on average, delaying forecasts for the portfolio of firms they cover also increase analysts' life spans. We find no robust association between life span and optimism or overconfidence.