The Accrual Anomaly: Accrual Originations, Accrual Reversals, and Resolution of Uncertainty*

• 成果类型: Article
• 署名作者: Fedyk, Tatiana; Singer, Zvi; Sougiannis, Theodore
• 署名单位: University of San Francisco; Universite de Montreal; HEC Montreal; University of Illinois System; University of Illinois Chicago; University of Illinois Chicago Hospital
• 刊物名称: CONTEMPORARY ACCOUNTING RESEARCH
• ISSN/ISSBN: 0823-9150
• DOI: 10.1111/1911-3846.12538
• 发表日期: 2020
• 页码: 885-916
• 关键词: earnings management; balance-sheet; Cash flows; overconfidence; forecasts; investors; OPTION; INFORMATION; QUALITY; prices

摘要

We combine a fundamental property of accruals and a behavioral phenomenon to provide an explanation for the accrual anomaly. The fundamental property is accruals that originate must subsequently reverse. The behavioral phenomenon is individuals tend to underestimate the variance of noisy signals in various domains of decision making. We argue that originating accruals represents a noisier signal than reversing accruals because the uncertainty of whether originating accruals will eventually convert into cash is high, while there is no uncertainty regarding reversing accruals. If investors underestimate the variance of originating accruals but understand reversing accruals, then originating accruals will be mispriced while reversing accruals will not. To test this prediction, we first develop and empirically validate a novel method for ex ante detecting accrual originations and their reversals. Then we document that investors face increased uncertainty when accruals originate and decreased uncertainty when accruals reverse, and we provide evidence that only originating accruals are mispriced. We further demonstrate that our findings can be useful for improving the accrual-based trading strategy by ex ante detecting and removing accrual reversals from extreme accrual portfolios. Overall, we provide a behavioral explanation for the accrual anomaly that is consistent with the mispricing of originating accruals only.