Uncovering Expected Returns: Information in Analyst Coverage Proxies

This study demonstrates that standard analyst coverage proxies contain information about firm-level expected returns. I decompose analyst coverage proxies into abnormal and expected components using a simple characteristic-based model and show that firms with abnormally high coverage outperform firms with abnormally low coverage by approximately 80 basis points per month. Abnormal analyst coverage also predicts firms’ earnings news, suggesting analysts allocate coverage by anticipating changes in firms’ future profitability. Using mutual fund flows as an exogenous shock to firm-level mispricing, I also show analysts increase coverage for stocks that appear underpriced. Together, these findings indicate analysts allocate resources across firms in proportion to their expected returns, which makes analyst coverage reflective of expected returns and thus a noisy measure of information asymmetry and dissemination. JEL Classifications: G10, G11, G12, G14, M40, M41 ∗I thank Lily Fang, Nick Guest, Travis Johnson, Jinhwan Kim, Chris Noe and seminar participants at Boston University and the 2015 Fidelis Capital Management Summit for helpful feedback and comments. Contact: [email protected], E62-677 100 Main Street, Cambridge MA 02142. Uncovering Expected Returns: Information in Standard Analyst Coverage Proxies 1