Do CEO Paycuts Really Work?

Gerald J. Lobo is Arthur Andersen Chair in Accounting at University of Houston C. T. Bauer College of Business; Hariom Manchiraju is Assistant Professor of Accounting at Indian School of Business; and Sri S. Sridharan is John and Norma Darling Distinguished Professor in Financial Accounting at Northwestern University Kellogg School of Management. This post is based on their recent article, forthcoming in the Journal of Accounting and Public Policy. Related research from the Program on Corporate Governance includes Paying for Long-Term Performance by Lucian Bebchuk and Jesse Fried (discussed on the Forum here).

Boards of directors (boards) often cut CEO pay following poor performance. These paycuts can go beyond the general pay-for-performance relation. Agency theory suggests that such paycuts can act as a disciplining mechanism against the CEO and, therefore, can lead to better performance in subsequent periods. Consistent with this line of reasoning, there is some empirical evidence that firm performance improves following a CEO paycut.

In our article, Accounting and Economic Consequences of CEO Paycuts, forthcoming in the Journal of Accounting and Public Policy, we examine the possibility that cutting the pay of an incumbent CEO might also induce an adverse response. Specifically, we examine whether, in response to paycuts, CEOs actually increase their efforts to improve the underlying economic performance of the firm, or simply resort to managing measured performance through activities such as accruals manipulation and real activities management. These latter activities may be designed to boost reported earnings in the short-run at the expense of long-term shareholder value. Since CEO pay is often linked to reported earnings performance, CEOs have incentives to engage in earnings management after a paycut because such activities can lead to faster improvement in reported performance and, hence, to speedier restoration of their pay to prior levels. Thus, the efficacy of a CEO paycut as a disciplining mechanism is unclear.

To answer this question, we identify 1,496 instances of CEO paycuts for a sample of non-financial firms in Execucomp over the period 1994-2013. We classify a decrease in CEO pay as a “paycut” if the CEO’s incentive compensation (i.e., bonus plus stock-based pay) that is not directly tied to performance is reduced by at least 25% from the previous year. By doing so, we are able to identify planned or deliberate paycuts arising from a revised contract as opposed to paycuts that follow as a consequence of poor performance based on an existing or prior managerial contract. We then use the propensity-score-matching (PSM) procedure to select a control sample that includes firms that did not initiate a CEO paycut, but are similar to the firms that did initiate a CEO paycut, using factors that prior research has shown to be associated with a CEO paycut.

We then proceed to examine changes in earnings management before and after a CEO paycut for the firms that had a CEO paycut and benchmark these changes against this control sample. We consider both accruals-based and real activities-based earnings management in our analysis. Our analysis reveals a significant increase in the magnitude of abnormal accruals and abnormal discretionary expenditure after a CEO paycut compared to the period prior to the pay cut. The firms with a CEO paycut would have reported much lower profits in the year following the paycut had they not engaged in these earnings management activities. In contrast, we do not observe a similar increase in earnings management for the matched control firms.

We then examine cross-sectional variation in the proclivity of CEOs to manage earnings in the year following the paycut. Prior research suggests that one role of corporate governance in financial reporting is to ensure compliance with financial accounting requirements and maintain the credibility of financial statements. We focus on the role of two particular features of corporate governance—CEO power vis-à-vis the board, and institutional ownership in the firm. We find that the proclivity to manage earnings after a pay cut is higher for firms whose CEOs are more entrenched (as proxied by a higher E-index) because these CEOs are less likely to be constrained when their power relative to the board’s is greater. On the other hand, we find that the ability of CEOs to opportunistically manipulate earnings in the year after the paycut decreases in the presence of dedicated institutional ownership because more effective monitoring implied by dedicated institutional ownership inhibits CEOs from engaging in such opportunistic behavior.

Finally, we examine the impact of paycuts on long-term profitability and risk. We find that one-year-ahead return on assets, cash flow from operations to total assets (CFO~t+1~), and stock returns increase after a CEO paycut (relative to the control sample), but only for those firms that have low levels of earnings management after the paycut. In contrast, the future performance is lower following a CEO paycut for firms that exhibit high levels of earnings management. These results also rule out the possibility that the earnings management proxies capture operational changes expected to happen after a CEO paycut rather than opportunistic CEO behavior to mask poor reported accounting performance in the short-run. We also find that one-year-ahead idiosyncratic return volatility increases (decreases) significantly after a CEO paycut (relative to the control sample) for firms that have high (low) levels of earnings management.

Our results raise two questions that require further explanation. First, why would a CEO engage in earnings management only after the board cuts his pay and not do so before the paycut? The CEO may have been able to avoid the paycut and other negative ramifications of poor performance by engaging in earnings management prior to the paycut. Second, why would a board that is able to cut the CEO’s pay tolerate earnings management behavior, which imposes significant agency costs on the firm? These two questions are interrelated and are best explained by examining the firms’ and managers’ behavior in the pre- and post-paycut periods in greater detail.

We posit that CEOs can time their earnings management so that they attract less scrutiny. The pre-managed earnings needs to be bumped up significantly to show an improvement in ROA for year T (the year of the paycut) over the ROA for year T-1. This exceedingly high level of earnings management is likely to attract considerably more scrutiny when compared to earnings management in year T+1 (the year after the paycut) when the benchmark to beat is relatively low. Thus, if the CEO is powerful and can influence the pay-setting process, he can devise a less risky strategy such as accepting a paycut in the period of poor performance to placate stakeholders, and subsequently have the board restore the pay to earlier levels when the firm’s reported accounting performance improves, albeit via real activities and accruals earnings management. This tacit understanding between the board and the CEO can also possibly explain why a board first cuts the CEO’s pay and then either tolerates earnings management or fails to formally recognize it. In this manner, the CEO is also able to avoid the negative publicity, scrutiny, and political constraints associated with high compensation during times of poor performance.

In summary, we examine the effectiveness of paycuts to induce CEO effort. Our analysis shows that the reported improvement in financial performance after a CEO paycut is superficial in several instances. The pressure to achieve a quick turnaround and to restore compensation and reputation in the labor market to pre-paycut levels may lead CEOs to engage in both real earnings management and accruals management. Hence, our findings suggest caution before considering CEO paycut as a strategy to induce greater CEO effort, as it can have unintended consequences in the absence of strong monitoring mechanisms.

The full article is available here.

Both comments and trackbacks are currently closed.