Article ID: | iaor19961516 |
Country: | United States |
Volume: | 41 |
Issue: | 8 |
Start Page Number: | 1392 |
End Page Number: | 1396 |
Publication Date: | Aug 1995 |
Journal: | Management Science |
Authors: | Elgers Pieter T., Lo May H., Murray Dennis |
Keywords: | investment, time series & forecasting methods |
This study assesses the effectiveness of using systematic components of cross-sectional forecast errors from prior years in order to adjust current analysts’ earnings forecasts. The empirical results document that a significant component of the cross-sectional MSE in analysts’ forecasts is systematic, and that parameter estimates from earlier periods enable the elimination of a substantial portion of the systematic errors in current forecasts. Further improvements in forecast accuracy are attained by the incorporation of prior-year excess security returns in order to reduce unsystematic error.