Analyst Optimism and the Magnitude of Earnings Growth

Paper number: 97/08

Paper date: June 1997

Year: 1997

Paper Category: Discussion Paper

Authors

Richard D.F. Harris
University of Exeter

Abstract

This papaer evaluates analysts' consensus long run earnings growth forecasts. It is shown that the correlation between forecast earnings growth and actual earnings growth is extremely low. Consistent with other studies, forecast earnings growth is found to be too optimistic. This is illustrated by the fact that almost all earnings growth forecasts are positive, while actual earnings growth is more evenly distributed between positive and negative values. For the sample of companies for which actual earnings growth is positive, there is a stron positive correlation between forecast earnings growth and actual earnings growth. In sharp contrast for a sample of companies for which actual earnings growth is negative, there is a strong inverse correlation between forecast earnings growth and actual earnings growth. This suggests that analysts are to some extendt able to forecast the magnitude of earnings growth but not its sign. It is further shown that analysts' forecasts of long run earnings growth are incorporated into the market's expectation of future earnings growth. However, there is evidence that the market attaches more weight to forecasts that have the correct sign, even though the sign of the actual earnings growth is not known at that datae that the forecast is made.

JEL Classification Nos: C22, D84
Keywords: Analysts' Forecasts; Earnings Growth; Rational Expectations; Panel Data.

Corresponding Author: Richard D.F. Harris, Department of Economics, University of Exeter, Amory Building, Rennes Drive, Exeter, EX4 4RJ, Great Britain, tel: (44) 1392 263215, fax: (44) 1392 263242, email: R.D.F.Harris@exeter.ac.uk