In determining how much to pay their CEOs, U.S. companies typically use “benchmarking,” setting compensation at or above the median of what peer firms pay their CEOs. This practice is designed to enable firms to retain talented CEOs and to remain competitive in the labor market (Oyer and Schaefer 2005 [1], Bizjak, Lemmon, and Naveen 2008 [2]). However, some aspects of the practice are controversial. Critics contend that CEOs can influence boards and compensation consultants to select peer companies with higher-paid CEOs, thereby justifying greater increases in their pay.
What We Investigate
In a new paper, we study whether firms tend to obfuscate some of their benchmarking practices when disclosing their benchmarking peers. We examine whether firms tie their CEO’s pay to that of highlycompensated CEOs at a few, prominent “star” firms that are not declared to be compensation peers under SEC requirements and that lack significant economic ties to the benchmarking firms. We refer to this benchmarking behavior as reaching for the -stars. A factor that might facilitate undisclosed star-linked pay is that CEO compensation is usually evaluated relative to that of CEOs at peer firms. Hence, compensation increases might be less likely to trigger proxy adviser and shareholder disapproval – as we find – as long as similar star-driven pay increases are also observed at several other firms.
What Is a Star Firm?
We construct a compensation peer network using the peers declared by firms themselves and identify a small group of star firms with the greatest influence in the network. We use “eigenvector centrality” (EC), a measure used widely in network theory, as our primary measure of compensation influence (Bonacich 1972 [3]). We refer to the 50 firms with the highest EC as “star” firms. We test whether these star firms influence CEO pay at firms that do not declare the stars as direct compensation-benchmarking peers.
What We Find
We estimate the relation of the focal firm’s CEO pay to the median pay of star firms using regression analysis, controlling for the median pay of the focal firm’s declared peers, the focal firm’s size, its industry, the year, and other determinants of CEO pay. We find that focal firms’ CEO pay is significantly correlated with the stars’ CEO pay, after controlling for the focal firm’s declared peers’ CEO pay and various firm attributes. The correlation is stronger in low-profile focal firms.
The elasticity of focal firms’ CEO pay with respect to the stars’ median CEO pay (star-driven pay effect) is around 0.68. This suggests that a one standard deviation increase in the median CEO pay of stars corresponds to an approximate increase of $400,000 in CEO pay for the average focal firm. In comparison, a one standard deviation increase in declared compensation-peers’ pay is associated with an increase of around $1,440,000. While the declared compensation peer group exerts a larger influence on focal firm CEO pay, we find that stars that are undeclared as compensation peers by focal firms have a significant influence on a focal firm’s CEO pay.
Additionally, we find that star-driven pay effects are concentrated among focal firms that are peripheral in the compensation peer network (i.e., have below median EC). These low EC firms are less visible and are not often declared as benchmarking peers in the compensation peer network, which is worrisome from a corporate governance standpoint since this increases CEO pay levels for these focal firms for no apparent economic reasons.
Governance and Reputed Consultants
Consistent with this being an agency problem, we find evidence that star-driven pay effect for the low EC firms is significantly stronger among firms with higher agency costs such as those with lower stakes held by institutional investors, less independent and more co-opted boards, and longer CEO tenures and CEOs who also serve as president or chair of the board. We show that star-driven pay effects are asymmetric and that focal firms respond positively only to increases in star CEO pay but not to decreases. Furthermore, reaching-for-the-stars benchmarking is related to lower firm profitability and less firm value.
In setting pay, designing incentive contracts and benchmarking compensation, compensation consultants often play an important role in assisting focal firms and selecting appropriate compensation peers (Bettis et al. (2018) [4]) We consider the role of reputable compensation consultants ( those with top market shares in the industry) in reaching for the stars. We find that, with these consultants, the effect of reaching for the stars among low EC focal firms is largely mitigated.
Ruling Out Alternative Explanations
We study whether the CEO labor market and the urge to retain talent explain reaching for the stars by low EC focal firms. We do not find any empirical support for this explanation. We investigate whether reaching for the stars reflects the propagation of pay across firms – an outcome of system-wide compensation peer benchmarking – or the propagation] of pay from prominent firms that define the market for CEO pay in the economy and find no support for this explanation either. Finally, we consider whether the focal firm and star firms have any connections that would facilitate the sharing of information – common directorships, compensation consultants, product markets, industry, or geography –and find no support for this alternative explanation.
Was Reaching for the Stars Triggered by SEC Disclosure Requirements?
Firms did not have to disclose compensation peers in the United States prior to the 2006 SEC compensation disclosure rules, which rules required firms to disclose information on their compensation benchmarking peers. We therefore examine whether reaching for the stars emerged only after the disclosure rules were implemented, and our results suggest that it did. We conjecture that the regulations had the counter-intuitive effect of making pay in the compensation network more visible to firms and led to increases in CEO pay levels for low EC firms.
Our evidence indicates that the 2006 SEC disclosure rules enabled firms to recognize star firms in the compensation peer network, possibly prompting the emergence of star-driven pay effects. Our findings on star-driven pay are novel to the literature and highlight the potentially unanticipated consequences of requiring greater disclosure of compensation arrangements.
Overall, our findings highlight that star-driven pay is primarily an outcome of agency problems and, interestingly, resulted with more compensation disclosures enacted through regulations.
Herd Immunity?
We find that, unlike other types of excessive pay, star-driven pay does not appear to trigger a negative response from shareholders and proxy advisers. This supports the notion of herd immunity: star-linked pay might receive less scrutiny – since CEO pay is judged relative to that at similar firms, many of which might be influenced by CEO pay at star firms.
Convergence Trends in CEO Pay
Finally, we examine whether star-driven pay can partly account for some observed, but little understood, convergence trends in CEO compensation. Recent studies report a significant convergence in CEO compensation across firms and find CEO pay trends to be markedly different between larger, higher-pay firms and smaller, lower-pay firms (Edmans, Gabaix, and Jenter 2017 [5], Cabezon 2021 [6], and Jochem, Ormazabal, and Rajamani 2021[7]). We show that star-driven pay contributes significantly to these compensation trends. While only suggestive, these results indicate that a better understanding of the mechanisms underlying star-driven pay could be important to understanding broad trends in CEO pay.
ENDNOTES
[1] Oyer, P., and S. Schaefer. 2005. Why do some firms give stock options to all employees?: An empirical examination of alternative theories. Journal of financial Economics 76:99–133.
[2] Bizjak, J. M., M. L. Lemmon, and L. Naveen. 2008. Does the use of peer groups contribute to higher pay and less efficient compensation? Journal of Financial Economics 90:152–168.
[3] Bonacich, P. 1972. Factoring and weighting approaches to status scores and clique identification. Journal of Mathematical Sociology 2:113–120.
[4] Bettis, J. C., J. Bizjak, J. L. Coles, and S. Kalpathy. 2018. Performance-vesting provisions in executive compensation. Journal of Accounting and Economics 66:194–221.
[5] Edmans, A., X. Gabaix, and D. Jenter. 2017. Executive compensation: a survey of theory and evidence. The Handbook of the Economics of Corporate Governance 1:383–539. Chapter 7.
[6] Cabezon, F. 2021. Executive compensation: The trend toward one size fits all. Working Paper.
[7] Jochem, T., G. Ormazabal, and A. Rajamani. 2021. Why have CEO pay levels become less diverse? Working Paper.
This post comes to us from professors Vikram K. Nanda at the University of Texas at Dallas, Swaminathan L. Kalpathy at Texas Christian University’s M.J. Neeley School of Business, and Yabo Zhao at the Chinese University of Hong Kong (Shenzhen). It is based on their recent paper, “Reaching for the ‘Stars’ in CEO Compensation,” available here.