The Effect of Disclosure on the Pay-Performance Relation

The following post comes to us from Gus DeFranco and Ole-Kristian Hope, both of the Rotman School of Management at the University of Toronto, and Stephannie Larocque of the Department of Accounting at the Mendoza College of Business, University of Notre Dame.

An important task for boards is to oversee executive compensation. The effectiveness of boards in carrying out this monitoring responsibility, however, is widely debated. In the paper, The Effect of Disclosure on the Pay-Performance Relation, forthcoming in the Journal of Accounting and Public Policy, we argue that disclosure improves board effectiveness. First and as a general point, disclosure can improve transparency, which facilitates the monitoring of management and hence causes managers to act more in the interests of shareholders. Such monitoring is potentially valuable since managers will not always act in the best interest of shareholders.

Second, increased disclosure can allow external stakeholders, such as institutional investors, analysts, and the media, to develop their own independent and informed views on firms’ decisions. These external stakeholders can pressure the board to act in the interests of shareholders. Although board members have fiduciary responsibilities to shareholders, members have incentives to support the CEO. Serving on a board can provide a salary or other compensation as well as prestige and valuable business and social connections. Further, factors such as collegiality, team spirit, a natural desire to avoid conflict within the board, friendship, and loyalty can also align board members with CEOs. Agency theories argue that pressures from external investors are necessary to encourage managers to pursue value-maximizing policies.

Third, increased disclosure can potentially improve the quality of boards’ information sets. While boards will receive private information, the board depends largely on management for its information. This information could be biased or distorted. The CEO has little incentive to provide information that could cause board members to revise downward their assessment of the CEO’s talent or abilities. A publicly disclosed signal will undergo more scrutiny by external stakeholders, and more care will be taken in the process of disseminating it.

The theoretical intuition for our prediction follows that of agency theory. Better information can lead to a better understanding of the relation between the manager’s actions and performance, hence reducing noise in the estimated relation. Improved transparency can also allow stakeholders to more easily filter out the noise caused by factors unrelated to management’s actions on performance, consistent with the “noise reduction” role of information. Furthermore, disclosure can improve the precision of the performance signal. For example, the literature shows that voluntary disclosure by firms results in more informationally efficient stock prices. Thus, we predict that the pay-performance relation will be stronger for firms that provide management guidance, our empirical proxy for higher disclosure.

The analysis in this study supports our prediction. In a regression of CEO compensation (salary plus bonus) on accounting and stock price performance, we document that the coefficients on both performance measures are higher for firms issuing guidance. These tests are in changes, which controls for potentially correlated factors that do not change over time. Our results suggest that this difference is economically meaningful.

Management guidance is a voluntary disclosure and hence it is possible that the decision to disclose is related to the compensation decision. We address the potential endogeneity of managements’ decision to issue guidance in three ways. First, we use a Heckman self-selection model, in which the first stage predicts whether guidance was issued and the second stage estimates the pay-performance relation. Second, we use a matched-sample approach. Third, we study a subsample of firms in which the decision to disclose is likely unrelated to compensation (i.e., firms initiating disclosure immediately following the passage of Reg. FD). Our results are robust to these tests. We also investigate alternative explanations for our results. Our results are also robust to controls for the information environment, the asymmetric sensitivity of compensation to performance, variations in investment opportunities, and controls for the previous year’s and next year’s disclosure decisions. Replicating our main tests using conference calls as an alternative measure of disclosure provides corroborating evidence.

Our study contributes to the literature that examines how agency problems can be mitigated through increased transparency. Several empirical studies relate better disclosure to better firm performance. These papers implicitly assume that disclosure leads to better monitoring, which in turn leads to better performance. By focusing on the pay-performance relation, we establish a more direct link between disclosure and monitoring.

In addition, our study informs the current debate about the role of management guidance. Critics have called for an end to management guidance. They purport that such disclosures create incentives for firms to manage earnings upwards, distort earnings, or act myopically. Whether managers do so is an open question and the empirical evidence is mixed. We are agnostic about the costs of management guidance, and we do not suggest an equilibrium amount of management guidance. We simply point out that improved monitoring of CEOs represents a potential benefit that should be considered in analyses of management guidance.

The full paper is available for download here.

Both comments and trackbacks are currently closed.