Abnormal returns from predicting earnings thresholds

This study examines the performance of a trading strategy based on the prediction of firms concurrently reporting a positive earnings change and meeting analysts' earnings forecasts. The evidence indicates that a model predicting both earnings thresholds concurrently can yield excess returns that are incremental to predicting only one earnings threshold. Further, I find that the prediction of forecast errors is relatively more important than predicting earnings changes as the incremental benefit from predicting earnings changes concurrently with forecast errors is small relative to a model that predicts only forecast errors. The results hold while controlling for various risk factors and known anomalies.