Print
E-MailDavid De Angelis is Assistant Professor of Finance at at the Jesse H. Jones Graduate School of Business at Rice University and Yaniv Grinstein is Adjunct Professor of Finance at the Samuel Curtis Johnson Graduate School of Management at Cornell University. This post is based on their recent article, forthcoming in the Journal of Financial and Quantitative Analysis. 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).
In this article, we investigate a market-for-talent rationale for the use of relative performance evaluation (RPE) in CEO compensation. Our premise is based on Gibbons and Murphy (1990), who show RPE in CEO compensation can be used as an efficient way to compensate CEOs for their talent. In their setting, firms learn CEO talent from CEO performance relative to peer CEOs and compensate their CEOs according to their relative performance in order to retain their talent. We study the talent-retention hypothesis in two ways. First, we examine the contractual terms of RPE in CEO compensation and analyze the extent to which they are consistent with talent-retention motives. Second, using an improved empirical specification to detect RPE on a long panel of compensation data of CEOs of US firms, we test whether the transferability of CEO talent relates to a stronger use of RPE in CEO compensation.
The evidence regarding the contractual terms indicates RPE could play a role also for talent-retention purposes. We find the peers chosen to benchmark CEO performance tend to be the same peer CEOs who are also used to benchmark CEO compensation. The latter are considered the relevant talent pool of candidates for a CEO position (Bizjak, Lemmon, and Naveen (2008), Faulkender and Yang (2010, 2013), Albuquerque, De Franco, and Verdi (2013)). The high overlap between performance peers and compensation peers suggests firms in general view talent peers—CEOs whom the firm competes for talent—as close substitutes for CEO performance peers. Clearly, talent peers and performance peers should highly overlap also if performance peers are used to eliminate a common shock, because talent peers are usually CEOs who work in firms similar to the firm in question. Interestingly, we find performance peers are not necessarily involved in the same activities as the firm. About a third of all performance peers in our sample do not even belong to the same 2-digit SIC code as the firm.
In the cross-sectional analysis, we examine whether RPE is more prevalent when CEO talent is more transferrable. In such cases, CEOs would require a commitment to be paid for their outside opportunities in order to stay with the firm. To test this prediction, we study cross-sectional variations in the presence of RPE in CEO compensation across a large panel of firms between 1992 and 2017.
We find that as CEO talent becomes less transferrable, firms rely less on RPE in CEO compensation. We use five different proxies to capture the transferability of CEO talent: (i) whether CEO talent is general enough for CEOs to easily transfer their talent across firms (Custódio, Ferreira, and Matos (2013)); (ii) whether the CEO is also the founder of the firm, because the founder is likely to have strong firm-specific talent that cannot be easily transferred to other firms; (iii) whether the CEO is around retirement age, because CEOs around retirement age are likely to retire due to social norms (Jenter and Lewellen (2015)) and thus are less likely to move to other firms; (iv) whether firms in the industry tend to promote insiders as their new CEOs (Cremers and Grinstein (2014)); and (v) whether the state in which the firm is headquartered enforces non-compete clauses that restrict managers from leaving their firms for competitors (Garmaise (2011)). Overall, our evidence shows that when CEO talent is less transferrable, firms tend to rely less on RPE, consistent with the talent-retention hypothesis.
Our primary contribution is to shed new light on an under-explored role of RPE in CEO compensation. In that sense, our article answers the recent call made in Edmans, Gabaix, and Jenter (2017) for more work that analyzes the importance of other explanations beyond moral hazard for observed CEO compensation contracts. Our evidence suggests labor-market motives play a role in the design of CEO compensation contracts. The talent-retention hypothesis has cross-sectional implications for when RPE is useful and when it is not: These implications can help future research connect RPE either to good governance or to optimal contracting. It thus extends the literature that studies the possibility that performance-based compensation is used to retain talent.
The complete article is available here.
