Print
E-MailThe following post comes to us from Ilia Dichev, Professor of Accounting at Emory University; John Graham, Professor of Finance at Duke University; Campbell Harvey, Professor of Finance at Duke University; and Shivaram Rajgopal, Professor of Accounting at Emory University.
While hundreds of research papers discuss earnings quality, there is no agreed-upon definition. We take a unique perspective on the topic by focusing our efforts on the producers of earnings quality: Chief Financial Officers. In our paper, The Misrepresentation of Earnings, which was recently made publicly available on SSRN, we explore the definition, characteristics, and determinants of earnings quality, including the prevalence and identification of earnings misrepresentation. To do so, we conduct a large-scale survey of 375 CFOs on earnings quality. We supplement the survey with 12 in-depth interviews with CFOs from prominent firms.
Why ask CFOs about earnings quality? We argue that CFOs are in a unique position to provide insights on earnings quality. The reason is simple: the CFOs’ choices determine, to a large extent, the quality of earnings. Almost all of the research on earnings has been empirically based where quality is assumed to be a function of observables, such as accruals and real decisions to massage earnings. In contrast, our approach focuses on the person that makes decisions with respect to accruals and real actions.
We find that CFOs believe that quality earnings are sustainable and predictable, with few one-time items, and solid backing by cash flows. Earnings quality is determined in about equal measure by uncontrollable factors like industry and economic conditions, and controllable factors like internal controls and corporate governance.
We also find that in any given period a remarkable 20% of public firms use discretion within GAAP to intentionally misrepresent their earnings. The magnitude of such misrepresentation is surprisingly large—of firms that do it, an average of 10% of the reported earnings is misrepresented. In addition, the misrepresentation goes both ways with a full one-third of perpetrators low-balling their reported earnings.
The most popular reasons for earnings misrepresentation are desire to influence stock price, related internal and external pressures to hit targets, and executive compensation and career concerns. Misrepresentation is also difficult to detect for an outsider. Accounting rules have become complicated, and the motivation for real actions like cutting valuable R&D initiatives often cannot be disentangled between business decisions and earnings management. Nevertheless, the CFOs provide a useful list of “red flags” that they would look for if they were searching out misrepresentation.
More than a decade after the Enron scandal, increased regulatory scrutiny following the Sarbanes Oxley Act and the demise of Arthur Andersen, the practice of earnings distortion continues unabated. As long as incentives to manage earnings to influence stock prices remain, either via managerial equity ownership or via pressure from the managerial labor market or from stakeholders such as analysts to “hit the numbers,” we believe that earnings misrepresentation will live on.
The full paper is available for download here. Tables and references are available here.
