34 Pages Posted: 8 Dec 1998
Date Written: December 1997
We evaluate the ability of the mean analyst forecast to effectively summarize analysts' information. We show analytically that even if analysts possess the ability and intention to forecast earnings truthfully, the mean forecast underweights analysts' private information. Thus, the mean does not adequately aggregate the full set of information individual analysts use in making their forecasts. Since the mean underweights private information, the problem worsens as the number of analysts forecasting earnings increases. We show that a positive multiple of forecast revision can be used to reduce the impact of improper information aggregation. We show empirically that forecast errors are positively related to forecast revision, and this relation is increasing in the number of forecasts made. Our results have implications for researchers who use the mean analyst forecast to proxy for the market's expectations of earnings.
JEL Classification: G10, G20, G29, M41
Suggested Citation: Suggested Citation
Kim, Oliver and Lim, Steve C. and Shaw, Kenneth W., The Use of Forecast Revision in Reducing Built-in Biases in Mean Analyst Forecasts (December 1997). Available at SSRN: https://ssrn.com/abstract=140997 or http://dx.doi.org/10.2139/ssrn.140997
By John Graham