Earnings Management from the Bottom Up

The following post comes to us from Felix Oberholzder-Gee and Julie Wulf, both of the Strategy Unit at Harvard Business School.

In our paper, Earnings Management from the Bottom Up: An Analysis of Managerial Incentives below the CEO, which was recently made publicly available on SSRN, we analyze all components of compensation packages for CEOs and for managers at levels below that of the CEO. Pay-for-performance contracts are a critical instrument to align the interests of principals and agents (Jensen and Meckling, 1976). While it can be optimal to make the agent the residual claimant of the firm’s profit, under numerous conditions principals are better off employing weaker incentives. These include situations with poor measures of performance and multitasking environments, when agents reduce their motivation in response to financial incentives, and when principal and agent have differing priors.

Another cost of high-powered incentives is that they provide managers with incentives to manipulate the firm’s reported earnings. For example, equity incentives can entice managers to boost reported earnings just before they exercise options or sell stock. There are now a number of academic studies – and many anecdotes – that document this link between the structure of chief executive officer (CEO) compensation and various measures of earnings manipulation (e.g., Beneish and Vargus, 2002; Bergstresser and Philippon, 2006; Peng and Roell, 2008). These papers generally focus on one component of compensation for the top position—equity incentives for the CEO. In this paper, we extend this literature by analyzing all components of compensation packages for CEOs and for managers at lower levels. To our knowledge, this study is the first that analyzes the relationship between CEO, division manager, and chief financial officer (CFO) compensation and earnings management in a large sample of firms.

We are interested in positions below the CEO because it is unclear if all or even most financial misreporting is decided at the top. There are many examples of managers at lower levels in the corporate hierarchy “cooking the books,” in some cases without the knowledge of senior management. For instance, at the H.J. Heinz Company, division managers received bonuses only when earnings increased from the prior year. These incentive plans led managers, among a longer list of improper accounting practices, to manipulate the timing of shipments, falsify dates on sales invoices, and recognize advertising expenses in the wrong period. Senior managers were unaware of these practices for seven years. They were ultimately discovered during investigations associated with a lawsuit that Heinz filed against a competitor. In addition to division managers, the importance of the CFO’s role in financial reporting and the numerous recent corporate fraud cases suggest that CFOs can significantly affect accounting quality. Consistent with this, Jiang, Petroni and Wang (2010) find that the magnitudes of accruals are more sensitive to CFO equity incentives than to those of the CEO. Although these examples and findings certainly suggest that managers below the CEO level could play a significant role in earnings management, to date much of the literature is focused on CEOs, mainly due to data limitations. This paper fills the gap.

Since the quality of financial reporting is difficult to assess, we use various measures of earnings manipulation in our study, including discretionary accounting accruals, end-of-year excess sales and class action litigation. While most of the existing literature employs cross-sectional methods, we analyze a panel over a long period (1986-1999) during which the structure of compensation contracts varied considerably. The panel structure of our data allows us to control for time-invariant unobserved heterogeneity, both at the level of the firm and the managerial position. While most of the existing literature uses discretionary accounting accruals, we also use a measure of excess fourth-quarter sales that is measured rather precisely because identification stems from variation in the beginning of firms’ fiscal years, which is plausibly exogenous to earnings manipulation.

We find that the effects of incentive pay on earnings management vary considerably by both type of incentive pay and position. For instance, companies report significantly higher discretionary accruals, excess sales and have a higher incidence of future lawsuits when CFOs are paid larger bonuses. Importantly, the magnitudes of these effects are much larger for CFOs in comparison to both CEOs and division managers. We find a small positive effect on accruals from current grants of stock options to division managers, but in contrast to prior research, we find no effect from equity incentives for CEOs or CFOs on accruals once we include bonus pay in our regressions. Turning to excess 4th quarter sales, we find little effect of division manager pay on excess sales in the fourth quarter which is at odds with the explanation that division managers reallocate effort to make annual budgets. Instead our findings suggest that CFOs aggregate financials to make bonus targets by shifting revenue to the 4th quarter of the fiscal year. We also find larger excess 4th quarter sales when CFOs have more stock options—a result that is at odds with the expectation that managers should lower reported earnings in advance of options grants to reduce exercise prices. Overall, our results suggest that the primary focus of compensation committees on equity incentives for CEOs overlooks a critical component in curbing earnings manipulation. If one wanted to weaken incentive pay to get more truthful reporting, diluting bonuses—particularly the CFO’s–would be the place to start.

The full paper is available for download here.

Both comments and trackbacks are currently closed.