Determinants and value relevance of UK CEO pay slice

Abstract

This paper studies the CEO pay slice (CPS) of UK listed firms during the period 2003 to 2009. We investigate the determinants of CPS. We study the links between CPS and measures of firm performance. We find that firms with higher levels of corporate governance ratings and those with more independent boards tend to have higher CPS. In addition, we find that CEOs are more likely to receive lower compensation when they chair the board and when they work in firms with large board size. We also find that higher CPS is positively associated with firm performance after controlling for the firm-specific characteristics and corporate governance variables. We get compatible results when we examine the association between equity-based CPS and firm performance. Our results remain robust to alternative accounting measures of firm performance. Our results suggest that high UK CPS levels do indeed reflect top managerial talent rather than managerial power.

Keywords:

• CEO pay slice
• firm performance
• corporate governance ratings
• corporate governance mechanisms

JEL Classifications:

• G30
• M15

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. In this study we use CEO pay slice instead of the CEO pay inequality measure developed by Forbes et al. (2014) for two reasons. First, Forbes et al. (2014) calculate a measure of CEO pay inequality as the compensation of the CEO divided by remuneration to all the executives on the board. This measure is inversely associated with the size of the board. Prior research (i.e. Guest 2009) shows that the board size is negatively associated with corporate performance using UK data. As a result, examining the effect of CEO pay inequality on firm performance is vulnerable to the correlated omitted variable problem, because CEO pay inequality and firm performance are related to board size. Second, our focus in this paper is to investigate the association between CEO’s relative pay and firm’s future performance, and one benefit of using CPS is that we may compare and reconcile our findings to those based on US data (i.e. Bebchuk et al. 2011). If we do not use CPS, it will be difficult to triangulate our results with previous studies.

2. The responsibilities of the remuneration committee include: (1) the determination of company-wide policy on remuneration; (2) the determination of individual remuneration packages for each executive director and other senior executives if appropriate; (3) reporting directly to shareholders on behalf of the board of directors on all matters relating to executive remuneration (Hughes 1996).

3. Among the six provisions, four limit shareholder rights (staggered boards, limits to shareholder amendments of bylaws, supermajority requirement for merger and charter amendment) and the other two make potential hostile takeover more difficult (poison pills and golden parachute). Therefore, a higher index score implies that the firm is entrenched, or has weak governance.

4. Carter et al. (2010) use two sets of proxies to capture the talent of executives. The first set of variables captures characteristics of the executives’ managerial position and professional profile at the prior firm, while the second set of variables measures firm performance and financial reporting quality of the executives’ former employer over their managerial tenure.

5. We regress firm performance in year t+1 on CPS in year t in the empirical analysis.

6. Consistent with previous literature, we exclude financial institutions and utility firms.

7. UK publicly listed firms are required by the Directors' Remuneration Report Regulations (2002) to disclose information on executive as well as non-executive’s compensation (including cash compensation, share options and long-term incentive schemes) in a separate ‘Director’s remuneration report’ as part of the annual report.

8. We drop firms for which CEO is not among the five most highly paid executives.

9. We follow the corporate governance literature by including board size, board independence and institutional ownership in our models (see for example McConnell and Servaes, 1990; Larmou and Vafeas, 2010).

10. The Shea partial R² for the instrument variables in all our models is higher than 0.50. In addition, the first stage regression in our models indicates a good significant level and power. Hence, the instruments in our case are not weak.

11. We exclude ROE as one of the control variables in these analyses.