Do Analysts Chase Prices When Prices Have No Information?

Abstract

In a recent survey of analysts, 96% claim that returns are not very useful as earnings forecast model inputs. I find, though, that analysts actually do incorporate returns into their earnings forecasts, even if those returns have no underlying earnings information. This leads to forecast error, which is worse among inexperienced analysts and those with reduced attention. Finally, I show that the market is unable to anticipate this error, leading to mispricing that is not resolved until the earnings announcement date. The literature has so far been unable to explore these issues causally. However, I am able to do so by developing a novel identification strategy that isolates the non-earnings component of the Federal Open Market Committee (FOMC) announcement and separates out concurrent information using intraday and instrumented returns.

Keywords: analyst forecasts, busy analysts, experienced analysts, fatigued analysts, market inefficiency

JEL Classification: G3, G17, M4, M41