CEO Retirement, Corporate Governance and Conditional Accounting Conservatism

Abstract

Based on 16,604 observations between 1994 and 2006, this study revisits the ‘horizon problem’ by examining how CEO retirement affects conditional accounting conservatism. We hypothesize and find that firms become less conservative in their financial reporting before the retirement of their CEOs, and that strong corporate governance mitigates the effect of CEO retirement. The literature concerning the horizon problem has suggested that CEOs manipulate earnings to boost short-term performance before they leave their companies (Dechow, P. M., & Sloan, R. G. (1991). Executive incentives and the horizon problem: An empirical investigation. Journal of Accounting and Economics, 14(1), 51–89; Smith, C. W., & Watts, R. L. (1982). Incentive and tax effects of executive compensation plans. Australian Journal of Management, 7(2), 139–157), but the evidence is mixed. By examining conditional conservatism, we avoid some of the methodological difficulties that confront researchers when examining either real or accrual earnings management. Ours is the first study to provide evidence on how the horizon problem shapes conditional accounting conservatism.

Acknowledgements

We thank Professor Laurence van Lent and the two anonymous reviewers for their constructive comments. We also would like to thank the conference participants at the 38th European Accounting Association Annual Conference and those at the 2016 International Academic Business Conference San Francisco, Professor Woody Wu and Professor Zhijun Lin, for their very helpful suggestions.

Notes

1 While we cluster based on year and industry fixed effects, our results are qualitatively unchanged if we include firm fixed effects rather than industry fixed effects.

2 We have also examined whether the external and internal governance measures are substitutive or complementary in alleviating the horizon problem. Our sample firms are classified into three groups, with Group 1 having both fewer antitakeover provisions and strong internal monitoring mechanisms, Group 2 either fewer antitakeover provisions or strong internal monitoring measures (but not both), and Group 3 weak in both external and internal governance mechanisms. We find that the coefficients on the interaction terms of our interest are either insignificant or significantly positive for the first group, but consistently and significantly negative for the third group. For the second group, the coefficients are mostly insignificant. Overall, these results suggest a complementary rather than a substitutive effect of corporate governance mechanisms.

3 We do not include performance measures because performance is more related to forced turnover rather than retirement. However, our propensity score matching conclusions are the same if we include in the likelihood of retirement the following variables such as last year’s stock return and ROA, average ROA over CEO career, and post-retirement income measures including wealth and outside directorships.

4 For brevity, we do not present the results of the additional conservatism measures in this paper. A detailed description on those measures and the related results are available at: https://sites.google.com/site/crcgac/home/additional-tables.