On the Informativeness of Unexpected Exclusions from Street Earnings

We create a measure of unexpected exclusions from street earnings by partitioning total exclusions from street EPS (i.e., reported GAAP EPS less street EPS) into those that are forecasted by analysts (forecasted exclusions) and those that are not forecasted but are excluded from street EPS after GAAP EPS is reported (unexpected exclusions). Relative to forecasted exclusions, unexpected exclusions are larger in magnitude and more strongly associated with both recurring and non‐recurring, as well as income‐decreasing GAAP earnings items. We observe that both earnings announcement returns and analysts’ forecast revisions are positively associated with unexpected exclusions, controlling for unexpected street earnings. At the same time, analysts do not efficiently incorporate unexpected exclusions into their forecast revisions, as we also observe that analysts’ nextquarter forecast errors are positively associated with current‐quarter unexpected exclusions. Taken together, our evidence suggests the need to reconsider the informativeness of exclusions from street earnings.