Earnings management is the use of accounting methods by managers to manipulate financial statements in order to inaccurately convey the financial performance of their company. Earnings management is costly to stakeholders and society because it decreases the informativeness of earnings thereby distorting the economic efficiency of the stock market. Moreover, earnings management can also potentially result in large, negative stock returns and deteriorating performance upon detection. Accounting researchers have been interested in understanding the incentives behind earnings management for quite some time.
One prominent theory of the motivations for earnings management involves stock option compensation. Because firm managers’ personal portfolios are typically less diversified than shareholders’ portfolios, managers typically prefer less risk than shareholders. Firms alleviate this difference in risk preferences by linking managerial wealth to risk-taking through stock options. Traditionally, regulators and researchers argue that because earnings management increases stock return volatility, risk-taking incentives should also encourage earnings management (Greenspan 2002; Coffee 2005; Armstrong et al. 2013; Levitt 1998; Knowledge at Wharton 2003; Jensen 2002). However, in our recent work, we re-examine this commonly held belief based on three potential theoretical and methodological shortcomings of prior research in this area.
First, it is hard to isolate whether stock options cause earnings management, because boards tend to grant more options—and executives are more likely to manage earnings—when firms enjoy large growth opportunities. Boards increase risk-taking incentives when growth opportunities expand because these incentives enable executives to share in the gains of growth. Similarly, stock prices are more sensitive to earnings when the economy expands and creates growth opportunities. Thus, growth opportunities might result in stock options being correlated with earnings management without being the cause of it.
Second, prior theory predicts that boards rely on stock options when it is difficult to know what information executives have and what the optimal choice for shareholders is. However, managers are also inherently more likely to manage earnings in these settings of low transparency, as the probability of detection is lower. If boards offer greater risk-taking incentives when facing more difficult monitoring, then the monitoring could create a positive correlation between risk-taking incentives and earnings management. However, such an association would not be due to risk-taking incentives inducing earnings management.
Lastly, more risk tolerant executives value stock options more highly than less risk tolerant executives. As such, more risk tolerant executives are likely to match with firms offering greater risk-taking incentive compensation. However, these risk tolerant executives are also more likely to manage earnings even absent stock options. Thus, risk-taking incentives may not cause earnings management but simply represent the matching of risk tolerant executives with firms. Such endogeneity complicates how one interprets a positive association between risk-taking incentives and earnings management.
Methodology and results
We begin with replicating a prominent, published study that shows a positive association between risk-taking incentives and earnings management (Armstrong et al. 2013). When we do not control for our three aforementioned concerns, we find that the risk-taking incentives provided by stock options encourage executives to manipulate accruals, restate financial statements, and engage in fraud. However, when we control for growth opportunities and monitoring, this positive association vanishes. Thus, we find evidence consistent with our first two concerns, namely that there is not an economically or statistically important association between risk-taking incentives and earnings management once we properly control for growth opportunities and monitoring.
To address our third concern about executives matching to firms, we employ two complementary statistical methods to sort out causation from correlation. First, we exploit an exogenous change in the financial accounting standard for stock options called FAS 123R (it has since been renamed to ASC 718). FAS 123R required firms to expense stock options and therefore led to decreases in stock option compensation. Importantly, FAS 123R did not uniquely affect risk averse executives. Second, we employ an identification methodology called instrumentation. Instrumentation attempts to root out changes in risk-taking incentives that are not related to changes in managerial risk aversion. Across both tests, we again fail to find an association between the amount of risk-taking incentives and magnitude of earnings management. Thus, prior work showing risk-taking incentives causing earnings management may be an artifact of risk tolerant executives matching to firms that offer risk-taking incentives, consistent with our third concern. While we cannot conclude that there is no association between all forms of risk-taking incentives and all forms of earnings management, our results suggest the need to re-evaluate concerns over stock-option compensation.
Our study is significant to multiple parties. First, board members, regulators, and shareholders should find our results important. Regulators, the popular press, and academics have expressed concerns that equity-based compensation impairs financial reporting quality. In contrast, our results suggest risk-taking incentives do not have a pervasively positive relation with earnings management. Second, academics should find our research informative as we answer calls from researchers in various disciplines. Research on equity-based compensation has too often neglected to consider “the causes and consequences of accounting rules” (Murphy 2013). We look forward to the ongoing debate on the role that executive compensation plays in earnings management.
Michael Mayberry is the Jack Kramer Term Assistant Professor in the Fisher School of Accounting at the University of Florida. He currently serves on the editorial advisory boards of multiple academic journals, including The Accounting Review, The Journal of the American Taxation Association, and The International Journal of Accounting.
Hyun Jong Park is an Assistant Professor in the Fox School of Business and Management at Temple University.
Tian Xu is an Assistant Professor at Texas A&M University- Corpus Christi.
This post is adapted from their paper, “Risk-Taking Incentives and Earnings Management: New Evidence,” available on SSRN.