CEO Compensation and Relative Company Performance Evaluation

Abstract

This study examines how firm performance influences CEO pay in U.K. public listed companies. Specifically, it explores evidence of relative performance evaluation, following the recommendation in the U.K. Combined Code on Corporate Governance to link executive pay to relative firm performance. Using a panel of CEOs drawn from 204 of the largest, nonfinancial U.K. companies, between 2003 and 2007, we provide new and convincing evidence that shows basic pay and annual bonus is determined relative to annual FTSE 350 market performance and long-term incentive payouts are determined relative to 3-year FTSE 350 industry sector performance. Our results provide robust evidence that is consistent with the principal-agent framework of executive pay and firm performance. We demonstrate that it is crucial for research to distinguish between the different elements of pay and the different performance conditions that attach to those elements in order to establish a comprehensive understanding of the pay-for-performance relationship. The study provides confirmation that remuneration committees consider own firm performance relative to peer group performance in setting up CEO compensation contracts. We conclude that changes introduced to improve corporate governance practice in the field of executive pay are working to the benefit of shareholders.

Keywords

corporate governance executive pay firm performance relative performance evaluation

Introduction

Executive pay holds a central role in corporate governance policy and its function has become even more influential over the past two decades. In the United Kingdom, the Directors’ Remuneration Report (DRR) Regulations¹ and the U.K. Corporate Governance Code² and in the United States, the Securities and Exchange Commission³ place significant importance on the relationship between firm performance and executive pay and recognize the value owners put on relative performance evaluation (RPE). The use of RPE in determining executive compensation also has a strong theoretical foundation.⁴ An incentive contract based exclusively on absolute firm performance will penalize or reward the executive for factors outside of their control (e.g., a global economic recession or boom). A more efficient incentive contract must exclude the effects of market-wide random events, which are beyond the control of the executive. In RPE theory, the executive’s performance is determined by a comparison to the performance of other executives facing similar market risk. As a result, executive compensation should be determined relative to the peer group performance of firms exposed to the same market risk. The academic literature on RPE is inconclusive, but overall there is very little support for it in executive compensation research. The vast majority of empirical tests of RPE have been conducted in the United States, although there are a few studies in the United Kingdom and elsewhere. Overall, there is mixed evidence of RPE in U.S. studies and very little support in the U.K. literature.

A contention of our study is that one of the reasons for the often-inconclusive findings about pay and performance and the limited evidence of RPE is the way that academic researchers measure pay. A measure of executive pay that is widely used in U.K. studies is cash compensation and is usually the sum of basic salary and annual bonus. Cash compensation is a measure of realized pay because it only includes guaranteed pay or pay that has met performance criteria, but crucially, its major drawback is that it does not include payouts from stock options or long-term incentive plans (LTIPs) that are designed to vest according to absolute and/or relative firm performance.

All recent U.K. research on executive pay and firm performance that do allow for stock options and LTIPs measure target long-term pay and not realized pay. That is the maximum incentive opportunity (sometimes adjusted for the probability of vesting) rather than the amount of the award that actually vests after performance conditions have been satisfied. The actual payout can vary between zero to full vesting of the award. To some extent this might matter less in the United States where performance conditions on stock option awards are less prevalent. But in the United Kingdom, performance conditions are extensively used in stock option plans and LTIPs ever since Greenbury⁵ first suggested the use of performance vesting criteria attached to the awards of long-term incentives. Prior academic studies disregard these performance criteria and as a consequence instead focus on long-term pay opportunity. That is, prior research focuses on the amount the executive can potentially receive if the performance criteria are met in full. In contrast, this study treats realized incentive payments as distinct from maximum incentive opportunities with the intention to gain a more insightful understanding of the relationship between firm performance and chief executive pay.

We question whether one of the reasons for not finding evidence of RPE in executive compensation is the measures of pay and firm performance used by academics, which may not reflect how remuneration committees tie executive pay to relative firm performance. In this article, we make a distinction between the different measures of executive compensation and firm performance in ways that previous studies have not done. We begin by defining compensation in terms of realized pay—that is, the remunerative rewards that executives do receive based on actual performance. Next, we recognize that total realized pay consists of three key elements and we incorporate these in our study. First, basic pay paid in cash. Second, bonus pay, which is based on a single year’s performance, paid in cash, but sometimes with a mix of cash and stock (with one or both elements perhaps deferred). Third, long-term pay, based on longer term absolute and/or relative firm performance measured against a peer group and paid in stock or stock options. Finally, we address the different performance requirements for the different elements that make up total chief executive compensation and, in particular, the short-term performance requirements of short-term incentives versus the long-term performance requirements of long-term incentives.

The findings suggest that remuneration committees consider own firm performance relative to market peer group performance when making adjustments to chief executive basic pay. Furthermore, annual bonus plans and long-term awards are also designed with consideration given to peer group performance so that bonus payments are linked to benchmark market performance and long-term incentive payouts are linked to benchmark industry sector performance. Overall, these findings, as summarized in Table 1, contribute to the understanding of the relationship between firm performance and CEO pay and provide robust evidence that is consistent with the principal-agent framework of executive pay and firm performance.

Table 1.

Key Findings: How Firm and Peer Group Performance Affect CEO Pay.

Pay element	Performance evaluation
CEO basic pay	Lagged short-term stock return and lagged market peer group performance
CEO bonus pay	Short-term stock return and market peer group performance
CEO long-term pay	Lagged long-term stock return and lagged industry peer group performance

U.K. Focus and International Relevance

The United Kingdom provides an ideal setting to investigate RPE following the recommendation in the Greenbury Report and the Combined Code on Corporate Governance⁶ to consider linking executive pay to relative firm performance. It is apparent from U.K. remuneration reports that long-term incentives are designed so that awards vest, to some extent, according to a firm’s performance relative to the average performance of a predetermined peer group.

The widespread use of performance vesting criteria on all long-term incentive awards distinguishes the United Kingdom from other economies. In the United Kingdom, grants of long-term incentives (stock and stock options) are awarded contingent on future firm performance. This is not the same for U.S. long-term incentive awards where it has been common practice to grant large awards of stock options, which vest only according to time and are not contingent on future firm performance. However, there is some recent evidence to suggest that relative performance criteria attached to the vesting of long-term awards is becoming more prevalent in the United States and other countries and therefore the findings of our study will also have international relevance.

Theoretical Framework

An agency relationship describes a position where the owners of a firm entrust responsibility to executives who are expected to behave in a way commensurate with the interests of the shareholders. Agency “costs” arise when the executive’s and the shareholders’ interests are not fully aligned. An incentive contract is thought to be the most effective means of addressing the agency problem. Executive compensation should be related to the success of the firm via an incentive contract, which is designed to ensure that executives are acting in the interests of shareholders. An incentive contract tied purely to absolute firm performance is thought not to be optimal if the performance of the firm is correlated with market performance since it will reward or penalize the executive for overall market performance. A more efficient compensation contract ought to exclude the effects of market wide events, which are outside the control of the executive.

RPE theory suggests that how well or poorly an executive performs is determined by a comparison to the performance of other executives in a peer group, and as a consequence, executive compensation will also be relative. A peer group is a group of firms that face similar market risks. This might, for example, include firms in the same industry sector. Average peer group performance provides information on the common uncertainty faced by firms in the same market. According to the theory, the assessment of an executive’s performance in comparison to a competitor’s performance filters the “uncontrollable risk” from the executive’s own incentive compensation and protects them from the common uncertainty associated with the market benchmark.

In this study, we test whether RPE is used to determine realized CEO compensation using measures of short- and long-term firm performance and both industry sector and FTSE 350 peer groups. We do this by conducting separate regression analyses on the different elements of CEO pay (basic pay, bonus pay, long-term pay and total pay).

If RPE is present in executive compensation, then an executive will gain superior compensation for relatively better performance compared to a peer group. Conversely, executives will receive a reduced amount of compensation if they underperform their competitors.

Research Design

We employ fixed effects estimation using ordinary least squares regression clustered by year and firm CEO, to estimate the relationship between firm and peer group performance on chief executive pay. The fixed effects specification removes the unobserved and time invariant effects on pay by focusing on the variables that change from one year to the next and it is only the differences of variables that remain in the model. The transformation is computed by mean differencing the data. The mean difference is the difference between the observation and the mean of the observations. Time-invariant effects, such as industry sector, that do not change from period to period, are eliminated from the estimating equation because their mean difference is zero. The fixed effects estimates are “unbiased and consistent”⁷even if the unobserved effects, which are assumed to be fixed, are correlated with firm performance. In executive compensation research fixed effects models are often preferred because it is not unreasonable to expect that an executive-specific unobserved effect, such as individual ability, or a firm-specific unobserved effect, such as remuneration policy, may be correlated with firm performance. The fixed effects estimation models are shown below:

B a s i c P a y_{i} = γ_{i} + α + β_{1} * f i r m_{i} + β_{2} * p e e r_{i} + β_{3} * c o n t r o l s_{i} + μ_{i} .

B o n u s P a y_{i} = γ_{i} + α + β_{1} * f i r m_{i} + β_{2} * p e e r_{i} + β_{3} * c o n t r o l s_{i} + μ_{i} .

L o n g - T e r m P a y_{i} = γ_{i} + α + β_{1} * f i r m_{i} + β_{2} * p e e r_{i} + β_{3} * c o n t r o l s_{i} + μ_{i} .

T o t a l P a y_{i} = γ_{i} + α + β_{1} * f i r m_{i} + β_{2} * p e e r_{i} + β_{3} * c o n t r o l s_{i} + μ_{i} .

The subscript _i refer to a CEO-firm pair. γ_i is a time-invariant chief executive/firm specific effect and differs between firms. α is a time trend. µ_i is the idiosyncratic error. Firm is a vector of short- and long-term firm performance variables. Peer is a vector of short- and long-term peer group performance variables. The pay and performance variables are described later in this section.

The strength of the relationship between absolute firm performance and chief executive pay is represented by the coefficient β₁. The CEO pay variables are in logarithmic form, which translates changes in variables into percentage changes and therefore, β₁, the coefficient of firm, is the pay-for-performance elasticity. A key issue is that the regression coefficient represents the approximate percentage change in pay that is associated with a 1% change in shareholder return.

Evidence of RPE is based on the strength and the sign of the coefficient of peer. While controlling for absolute firm performance, β₂ the coefficient of peer, is expected to be significant and negative. The so-called strong form of the RPE hypothesis⁸ expects the coefficient β₂ of the peer explanatory variable to be significant negative and equal in size to the predicted positive coefficient β₁ of the firm explanatory variable. In other words, the sum of the two coefficients is equal to zero (β₁ + β₂ = 0). The predicted result is consistent with the complete filtering out of the market risk component of firm performance. The weak form of the hypothesis simply predicts that β₂ is significantly negative but smaller in size than β₁. This is consistent with only partial filtering out of market performance. In the strong and weak form of the hypothesis, executive pay “should be increasing in firm performance and decreasing in peer group performance.”⁹

Data Collection

We collect original data on U.K. firms covering 5 years from 2003, the year following the introduction of the DRR regulations, to December 31, 2007. The firms were selected from the FTSE 350 list on December 31, 2007 (excluding financial services, real estate and investment trusts). Executives with fewer than 2 consecutive years are excluded. Executives combining the role of Chairman and CEO are also excluded. The final data set comprises an unbalanced panel of 204 companies and 905 executive- years from 2003 to 2007. The sample includes firms from nine different FTSE 350 industry sectors. The chief executive compensation data is hand collected from company remuneration reports in order to facilitate the precision and detail necessary to construct the compensation variables. Other incumbent data such as CEO tenure is also hand collected from annual reports. We use Datastream (a financial database) to collect the firm performance data.

CEO Compensation

Basic or base pay is measured as the annual salary reported in the DRR. Bonus pay is measured as paid annual bonus, plus any guaranteed deferred cash and/or stock compensation with no further performance conditions attached. Long-term pay is measured as the value of performance options vesting in the current year, plus the value of performance stock vesting in the current year, plus the grant value of time-vesting options, plus the grant value of time-vesting stock. Total pay is basic pay, plus any other cash, plus bonus pay, plus long-term pay.

Performance options and time-vesting options are valued using the Black and Scholes¹⁰ option pricing model. Performance stock and time-vesting stock are valued as at the fiscal year-end company stock price. The compensation variables are transformed into natural logarithms in order to impose a constant percentage effect of the independent variable on the dependent variable. The log transformation cannot be applied where a variable has a zero value. Therefore, to facilitate the logarithmic transformation of the data zero bonuses and zero payouts from performance option or performance stock, awards are replaced with the value of £100.

Firm Performance

We use market-based measures of firm performance calculated from the Datastream return index. We prefer market measures of firm performance to accounting measures for several reasons. First, market measures of firm performance and peer group performance have a clear and intuitive link to shareholder interests. Second, in order to evaluate comparative accounting performance it is necessary to match firms according to their fiscal year-end, which reduces the comparator sample size to a small number of firms particularly in some industry sectors. Third, a review of company remuneration reports illustrates that relative total shareholder return is extensively used in long-term compensation arrangements.

We incorporate both short- and long-term firm performance in our models. Short-term stock return is measured as the annual change in the natural logarithm of the Datastream return index. Contemporaneous short-term stock return and lagged stock return are then included since adjustments to compensation are frequently based on the previous year’s firm performance. We measure long-term stock return as the 3-year change in the natural logarithm of the return index. Long-term stock return is lagged 1 year since actual long-term pay is determined at the end of a predetermined performance period, which may or may not be the fiscal year-end.

Peer Group Performance

In tests of RPE, the majority of studies use a measure of average peer group performance; however, there are differences in how the peer group is defined. A number of studies use an industry sector peer group whereas others employ a much broader definition of the market. Many studies use both an industry benchmark and a market index measure. We also take account of both industry sector peer group performance and market peer group performance.

Industry stock return is calculated from a portfolio of firms in the same FTSE 350 sector index. Firm return index data are available from Datastream on a daily basis, which enables industry stock returns to be calculated to match each firm’s own fiscal year-end. The FTSE 350 index stock return (excluding financial services, real estate and investment trusts) is calculated in the same way. Short-term industry peer group performance is measured as the median annual stock return of firms in the same FTSE 350 sector index. Short-term market peer group performance is measured as the median annual stock return of firms in the FTSE 350 index. Long-term industry peer group performance is measured as the median 3-year stock return of firms in the same FTSE 350 sector index. Long-term peer group market performance is measured as the median 3-year stock return of firms in the FTSE 350 index.

Control Variables

Consistent with prior executive pay research, several control variables are included in the CEO pay regression models. Executive pay studies control for firm size because it represents organizational complexity, and therefore, greater skill and effort may be required to manage a larger firm.¹¹ We measure firm size as the natural logarithm of firm sales lagged by 1 year. Our CEO pay models also include the market-to-book ratio as a measure of growth options within the firm.¹² In order to control for firm risk we also include stock volatility.¹³ Volatility is measured as the standard deviation of annualized monthly stock returns over the prior 48 months. CEO tenure is included to capture the CEO’s experience in the role. We measure tenure as the number of years since being appointed CEO. It is then squared so that it can be included in the fixed effects regression models. Finally, a set of year dummy variables are included to filter average changes in CEO compensation due to macro-economic shocks such as price inflation and pay trends. We do not report the control variables in this study.

Results

Descriptive Statistics

Table 2 reports the descriptive statistics for the compensation variables. The mean CEO basic pay data increased from £450,260 in 2003 to £566,700 in 2007 for our sample of firms. The mean CEO total pay nearly doubled from £1,093,969 in 2003 to £2,070,682 in 2007. The median bonus increased progressively as a percentage of median basic pay from 50% in 2003 to 91% in 2007, whereas, the ratio of median long-term pay to median basic pay ranged from 32% in 2005 to 75% in 2007.

Table 2.

CEO Pay Descriptive Statistics (£ 000s).

Variables	2003 Mean (median)	2004 Mean (median)	2005 Mean (median)	2006 Mean (median)	2007 Mean (median)
Basic pay	450.26 (397.50)	473.88 (411.00)	507.57 (450.00)	536.20 (482.13)	566.70 (513.50)
Bonus	338.81 (200.00)	396.86 (250.00)	456.21 (300.00)	530.70 (400.00)	647.56 (469.50)
Long-term pay	419.72 (170.00)	662.02 (178.33)	517.58 (142.28)	704.02 (269.63)	1,072.55 (385.51)
Total pay	1,093.97 (752.52)	1,385.58 (933.00)	1,413.87 (956.80)	1,615.30 (1,104.28)	2,070.68 (1,352.53)

Basic Pay Results

Table 3 shows the results of our fixed effects regression on CEO basic pay. Columns 1 and 2 report the effect of short-term performance on basic pay. Here the stock return coefficients are significant but negative, which is contrary to our predicted result. Industry return (column 1) is not significantly related to basic pay and market return (column 2) is marginally positively related to basic pay. Next, columns 3 and 4 show the effect of lagged short-term performance on basic pay. In column 3 we report a significant negative coefficient for lagged short-term industry return, but the corresponding stock return coefficient is not significant. However, column 4 does show a significant positive relationship between lagged short-term stock return and basic pay (t = 1.69, p < .1). From column 4 we also observe a significant negative association between lagged short-term market return and basic pay (t = −2.69, p < .01). Thus we provide evidence of RPE in CEO basic pay with respect to lagged short-term stock return relative to market return. Columns 5 and 6 report the results of lagged long-term performance on basic pay. The long-term stock return coefficients are significant and positive (column 5: t = 2.98, p < .01; column 6: t = 3.43, p < .01), but we find no significant result for long-term industry or market return on basic pay.

Table 3.

CEO Basic Pay on Firm and Peer Group Performance.

	ln(Basic pay)
	(1)	(2)	(3)	(4)	(5)	(6)
Short-term performance variables
Stock return	−0.04* (−2.39)	−0.05** (−2.83)
Industry return	0.02 (0.96)
Market return		0.05^† (1.70)
Lagged short-term performance variables
Stock return			0.02 (1.52)	0.02^† (1.69)
Industry return			−0.04^† (−1.94)
Market return				−0.06** (−2.69)
Lagged long-term performance variables
Stock return					0.03** (2.98)	0.04** (3.43)
Industry return					0.02 (1.19)
Market return						−0.03 (−1.31)

Note. t-Statistics in parentheses.

p < .05. **p < .01. ***p < .001. ^†p < .1.

Overall, we do not find evidence of RPE in CEO basic pay when using contemporaneous stock returns or lagged long-term returns, but we do find support for RPE when we use lagged short-term stock return and lagged short-term market return. Our results are consistent with remuneration committees adjusting basic pay according to the previous year’s firm performance relative to benchmark market performance. The result is in contrast to most other studies, which measure cash compensation (basic pay plus bonus) or total compensation (cash compensation plus long-term incentives).

Bonus Pay Results

Table 4 reports the results of the fixed effects regressions on CEO bonus pay. Columns 1 and 2 show the effect of short-term performance on bonus pay. When the peer group consists of firms in the same FTSE 350 industry sector (column 1), we find no evidence of RPE; the coefficient is as predicted, negative, but is not significant. But we do find evidence consistent with RPE theory in Column 2 when we use short-term market return. Column 2 shows short-term stock return is positively related to bonus pay (t = 3.49, p < .01) and market return is negatively related to bonus pay (t = −2.14, p < .05).

Table 4.

CEO Bonus Pay on Firm and Peer Group Performance.

	ln(Bonus pay)
	(1)	(2)	(3)	(4)	(5)	(6)
Short-term performance variables
Stock return	1.23** (3.00)	1.36** (3.49)
Industry return	−0.77 (−1.12)
Market return		−1.64* (−2.14)
Lagged short-term performance variables
Stock return			0.49 (1.50)	0.56^† (1.74)
Industry return			−0.35 (−0.76)
Market return				−0.78 (–1.37)
Lagged long-term performance variables
Stock return					0.24 (1.16)	0.26 (1.18)
Industry return					−0.06 (−0.14)
Market return						−0.38 (−0.65)

Note. t-Statistics in parentheses.

p < .05. **p < .01. ***p < .001. ^†p < .1.

Columns 3 and 4 of Table 4 show no evidence of RPE in bonus pay when using the lagged short-term performance variables. Columns 5 and 6 also report no significant result when using lagged long-term performance variables. But as discussed above, we do find evidence of RPE using the contemporaneous short-term performance variables, which is consistent with bonus payouts being determined at the end of the financial year and reported in the annual report and accounts in the year to which they relate. Our results illustrate the importance for research to separate bonus payments from total compensation and to correctly specify the firm performance measure in the analysis of executive bonus payments. Until now research has focused on total cash or total compensation where the findings are mixed.

Long-Term Pay Results

The results for the fixed effects regressions of long-term pay are presented in Table 5. Columns 1 and 2 report no significant relationship between short-term firm performance and long-term pay. We also find no significant relationship when we use lagged short-term performance (columns 3 and 4). In column 5, we do find lagged long-term stock return is significant and positively related to long-term pay (t = 4.06, p < .001) and lagged long-term industry return is significant and negatively related to long-term pay (t = −2.30, p < .05). This is consistent with RPE theory and a principal-agent model of incentives designed to align the interests of executives with long-term shareholder value. In column 6, the coefficient of market return takes the correct sign but is not significant.

Table 5.

CEO Long-Term Pay on Firm and Peer Group Performance.

	ln(Long-term pay)
	(1)	(2)	(3)	(4)	(5)	(6)
Short-term performance variables
Stock return	−0.22 (−0.42)	−0.33 (−0.61)
Industry return	−1.11 (−1.08)
Market return		−1.05 (−0.75)
Lagged short-term performance variables
Stock return			−0.06 (−0.09)	0.03 (0.04)
Industry return			1.12 (1.05)
Market return				1.19 (1.01)
Lagged long-term performance variables
Stock return					2.02*** (4.06)	1.74** (3.27)
Industry return					−1.45* (−2.30)
Market return						−0.95 (−1.01)

Note. t-Statistics in parentheses.

p < .05. **p < .01. ***p < .001. ^†p < .1.

Total Pay Results

Fixed effects regression estimates for total pay are shown in Table 6. Columns 1 and 2 report a significant relationship between stock return and total pay. In columns 3 and 4, the results for lagged short-term stock return are not significant. Lagged long-term stock returns (columns 5 and 6) are significant and positively related to total pay. Evidence of RPE is restricted to the relationships between short-term industry return (column 1: t = −2.18, p < .05) and market return (column 2: t = −3.72, p < .001) and total pay. There is no significant evidence of total pay being determined relative to lagged short-term performance or lagged long-term performance.

Overall, our analysis provides strong support for RPE theory. We observe from our findings that each element of pay is related to a different measure of peer group performance. We conclude from these results that it is imperative to divide chief executive compensation into its individual components in order to understand how relative firm performance influences chief executive pay.

Table 6.

CEO Total Pay on Firm and Peer Group Performance.

	ln(Total pay)
	(1)	(2)	(3)	(4)	(5)	(6)
Short-term performance variables
Stock return	0.15* (2.21)	0.18** (2.99)
Industry return	−0.24* (−2.18)
Market return		−0.48*** (−3.72)
Lagged short-term performance variables
Stock return			0.06 (1.11)	0.07 (1.34)
Industry return			0.04 (0.40)
Market return				−0.01 (−0.08)
Lagged long-term performance variables
Stock return					0.23*** (4.70)	0.21*** (4.41)
Industry return					−0.04 (−0.55)
Market return						0.01 (0.11)

Note. t-Statistics in parentheses.

p < .05. **p < .01. ***p < .001. ^†p < .1.

Conclusion

The global financial crisis has prompted renewed calls for executive pay to be more closely tied to firm performance. At the moment, barely a week goes by without the media highlighting yet another example of so-called, unjustifiable executive rewards. This renewed attack, provoked by the banking crisis, may not be entirely justified according to the findings of this study.

The existence of RPE in the determination of executive compensation is an important theoretical question with important practical implications given that the U.K. Combined Code on Corporate Governance recommended that executive pay should be linked to firm peer group performance. The most important contribution of this study is the evidence to support the use of RPE in chief executive basic pay, bonus pay, long-term pay and total pay. Unlike many previous studies and in contrast to media and public perceptions, this study finds a significant relationship between firm performance and the compensation earned by CEOs.

We are critical of previous academic research for not correctly specifying the compensation variables or the performance periods over which they relate and suggest a failure to do so is a reason for often inclusive results. Our findings confirm that U.K. remuneration committees consider own firm performance relative to peer group performance when making compensation decisions such as changes to basic pay and incentive plan design.

The chief executive compensation data were hand collected exclusively for this study and provides a level of detail not previously explored in the literature. The original data set enables this study to empirically address the construct of pay. We contribute to the executive pay literature by distinguishing between potential and realized CEO pay, and consequently we are the first study to examine the relationship between firm performance and realized long-term pay. We show that it is important to separately analyze the individual components of total compensation in order to understand how remuneration committees employ RPE in U.K. chief executive pay and, in particular, how firm performance is related to the payouts from bonus and long-term incentive schemes.

A further contribution is that different elements of CEO pay are based on different firm performance outcomes. First, lagged short-term stock return and market peer group performance determine basic pay. Second, contemporaneous short-term stock return and market peer group performance determine bonus pay. Third, lagged long-term stock return and industry peer group performance determine long-term pay.

This study shows that to comprehend the CEO pay-for-performance relationship we must do the following:

Measure realized pay outcomes and not pay opportunity. Previous studies frequently ignore the performance conditions attached to the vesting of long-term incentives.

Separately analyze individual elements of pay because they are related to firm performance in different ways. Many earlier studies only measure total pay.

Consider the duration over which performance is measured and the timing of bonus/long-term payouts. The majority of studies do not use a performance period over more than one year.

Our study spans a period of 5 years commencing immediately after the introduction of the DRR Regulations. Therefore, an important caveat is that it does not capture the recent downturn in the economy and the impact of the corresponding financial crisis on the relationship between relative firm performance and chief executive pay. However, although this is an empirical question, RPE ought to be equally as important in a recessionary period since theoretically the use of RPE in executive pay should shield the executive from the common uncertainty associated with the market benchmark. A further caveat is that this study focuses on the compensation of the CEO, but as Liu and Stark identify in their study of RPE and executive board compensation, the CEO does not act alone but collectively with the board of directors. Further research might include realized total board pay. A final point of caution is that the results of this study are applicable to large nonfinancial U.K. publicly listed companies and cannot necessarily be generalized beyond. It would therefore be interesting to replicate the study in other countries, particularly where performance vesting conditions are widely used in LTIPs.

Overall, the findings provide an original insight to the structure of chief executive pay and confirm that changes to improve corporate governance practice in the field of executive pay are working to the benefit of shareholders. Various groups, including institutional investors, the government and the media, require that chief executive pay is linked to firm performance. Our study shows that a positive relationship exists between firm performance and realized CEO pay. These findings will be of particular practical interest to investors who expect the interests of executives to be aligned with those of the company’s shareholders, via an incentive contract that rewards executives for enhanced firm performance. Our findings provide robust evidence that is consistent with the principal-agent framework of executive pay and corporate performance.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

Mark Farmer is a lecturer in reward management and accounting at Kingston University and has recently completed a PhD examining the use of relative performance evaluation in UK chief executive compensation. He has a professional background in industry, including as a remuneration consultant for Towers Perrin and remuneration manager for Legal and General Investment Management.

Stuart Archbold is Deputy Dean and Head of Kingston Business School at Kingston University. He joined Kingston University in 2001 from the University of North London. Prior to his academic career, he worked in various management and business planning roles in the construction, steel and manufacturing industries for a number of national and multinational companies. As well as teaching and researching in UK universities, he has also taught on programs in France, Greece, Portugal and Spain.

George Alexandrou is a principle lecturer accounting and finance at Kingston University. He has presented papers in a number of international conferences and has reviewed papers for conferences and journals. At Kingston he teaches postgraduates and undergraduate courses in corporate finance and finance and investments.