Robust Anomalies? A Close Look at Accrual‐Based Trading Strategy Returns

Abstract

The last 40 years have seen an extensive literature documenting so‐called anomalies in major capital markets. Evidence of ‘abnormal’ returns associated with trading strategies based on readily observable phenomena such as accounting‐based data involves experimental design choices that can be expected to influence the results. We show how evidence of an accrual anomaly in Australia is sensitive to research design specifications such as the choice of proxy for total accruals; the definition of abnormal returns (i.e. the return generating model); the impact of data trimming as a response to exceptionally large returns; and the choice between value and equal weighting of returns. We show that research design choices do matter and help reconcile conflicting prior evidence of any accrual anomaly in Australia. More broadly, our results suggest the need for caution in drawing inferences from trading strategy tests which claim to identify anomalies.