Social and Behavioral Sciences Commons^™

Robust Measures Of Earnings Surprises, Chin-Han Chiang, Wei Dai, Jianqing Fan, Harrison Hong, Jun Tu

Research Collection Lee Kong Chian School Of Business

Event studies of market efficiency measure an earnings surprise with the consensuserror (CE), defined as earnings minus the average of professional forecasts. Even if asubset of forecasts can be biased, the ideal but difficult to estimate parameter-dependentalternative to CE is a nonlinear filter of individual errors that adjusts for bias. We showthat CE is a poor parameter-free approximation for this ideal measure. The fractionof misses on the same side (FOM), by discarding the magnitude of misses, offers a farbetterapproximation. FOM performs particularly well against CE in predicting thereturns of US stocks, where bias is potentially large, than that of international …