Print
E-MailThis post comes to us from Roman Inderst, Professor of Finance and Economics at the Goethe University Frankfurt, and Holger Mueller, Associate Professor of Finance at New York University.
In the paper, CEO Replacement under Private Information, forthcoming in the Review of Financial Studies, we derive joint implications for the optimal CEO compensation and replacement policy based on a model of “information-based entrenchment.” In our model, the CEO’s desire to become entrenched is endogenous and does not derive from exogenously specified private benefits of control. Rather, it derives from the optimal compensation scheme inducing the CEO to work hard, which must promise him an ex-post quasi rent (in the form of generous on-the-job pay) in case he continues. This biases the CEO towards continuation which, together with his private information at the interim stage, drives a wedge between efficient CEO replacement and actual CEO turnover.
High-powered incentive pay, and possibly also severance pay, can mitigate CEO entrenchment. Incentive pay ensures that the CEO’s expected on-the-job pay is high precisely when the firm value under his continued leadership is high, thus aligning the CEO’s continuation preferences with those of the firm at the replacement stage. The role of severance pay in our model is more nuanced, as an increase in severance pay must be accompanied by a simultaneous increase in the CEO’s on-the-job pay. Otherwise, there would be too high a reward for failure and the CEO would shirk. As a result, each dollar of severance pay constitutes a dollar of rent for the CEO. Importantly, however, whether severance pay can mitigate entrenchment depends on the structure of the CEO’s on-the-job pay. As our model shows, the firm gets the biggest “bang for the buck” (i.e., the biggest reduction in entrenchment) if an increase in severance pay is accompanied by a simultaneous increase in incentive pay, not base pay.
Our model abstracts from many real-world features that are likely to affect CEO turnover and compensation in practice. For instance, the board’s role may not be confined to monitoring (and replacing) the CEO but it may additionally include giving the CEO valuable advice, as in Adams and Ferreira (2007). Also, our model assumes that the board acts in the firm’s best interest when designing the optimal CEO compensation and replacement policy. Alternatively, it has been argued that boards are captured by the CEO, maximizing the CEO’s utility instead of firm profits (e.g., Bebchuk and Fried, 2004). Realistically, the truth will probably lie somewhere in between these two polar views. Extending our model along these lines might provide a fruitful avenue for future research.
The full paper is available here.
