Noise in Expectations: Evidence from Analyst Forecasts

Abstract

Analyst forecasts outperform econometric forecasts in the short run but underperform in the long run. We decompose these differences in forecasting accuracy into analyst information advantage, forecast bias, and forecast noise. We find that noise and bias increase strongly with forecast horizon while analyst information advantage decays rapidly. Noise increase with horizon generates a mechanical reversal in the sign of the Coibion and Gorodnichenko (2015) regression coefficient at longer horizons, independently of over-/underreaction. A parsimonious model with bounded rationality and a noisy cognitive default à la Patton and Timmermann (2010) matches the term structures of noise and bias jointly.

Keywords: subjective expectations, noise, term structure of expectations, forecast bias, forecast noise, bounded rationality, machine learning

JEL Classification: C53, D79, D83, D84, D90, E37, E70, G17, G32, G40, G50, G59