Which Measure of Investors’ Earnings Expectations is More Value-Relevant? Evidence from Within-Quarter Individual Analysts’ Forecasts

• In Event: M&A, Trading Volume, Analysts

Abstract

Previous research documents financial analysts’ forecasts are a superior measure of investors’ earnings expectations relative to time-series models, and there are predictable patterns in analysts’ forecasts and forecast errors. We assume investors are aware of these patterns. If so, an analyst’s forecast is an incomplete representation of investors’ expectations, and researchers can improve upon these forecasts by incorporating additional conditioning information. Stated differently, using additional conditioning information enables generation of forecasts that better capture unobservable investors’ expectations. The more complete the conditioning information, the more effective will be this measure of investor EPS expectations.
We investigate four alternatives to the typically-used forecast change specification of changes in investor expectations that employ broader conditioning information sets. Using data from 2005 – 2015, we compare the association of abnormal stock returns and our four measures of changes in expectations and determine which best represents investors’ expectations. We also employ hedge-portfolio returns tests to gauge the relative ability of our alternative models to accurately capture investors’ expectations of earnings.
Both sets of empirical tests show a ‘bandwagon’ expectations model, which includes two previous forecast measures, provides the most significant improvement over the conventional approach. Our findings demonstrate there are significant benefits to considering information beyond the previous analyst forecast when attempting to represent investors’ expectations of earnings. Given these findings, future research may wish to continue to refine this approach to enhance our understanding of the relation between accounting information and share prices as well as the time series properties of analysts’ forecasts and forecast revisions. In conclusion, this paper provides useful information to accounting academics interested in studying these phenomena, as well as equity investors who are end-users of financial analysts’ forecasts.

Authors

• Ray J. Pfeiffer, Texas Christian University

• Karen Tucker Teitel, College of the Holy Cross

• Mahmoud Wahab, University of Hartford

• Susan Wahab, University of Hartford