This paper provides evidence of security analyst (SA) superiority relative to univariate time-series (TS) models in predicting firms' quarterly earnings numbers and shows that SA forecast superiority in our sample is attributable to: (1) better utilization of information existing on the date that TS model forecasts can be initiated, a contemporaneous advantage; and (2) use of information acquired between the date of initiation of TS model forecasts and the date when SA forecasts are published, a timing advantage.

This study examines the association between abnormal returns and five alternative proxies for the market's assessment of unexpected quarterly earnings. We examine the role that measurement error potentially has in multiple regression tests of abnormal returns (occuring around the time of earnings announcements) on an unexpected earnings proxy and other non-earnings variables. The results indicate a potential measurement error interpretation of such multiple regression tests. We examine three procedures which reduce, to an unknown degree, the measurement error problem. Our procedures appear to be more (less) effective at reducing measurement error for small (large) firms and recent (non-recent) forecasts.