Wealth creation and managerial pay: MVA and EVA as determinants of executive compensation

https://doi.org/10.1016/S1044-0283(03)00010-3Get rights and content

Abstract

Designing effective compensation contracts has become increasingly complex due to the globalization of the executive work force and the multitude of incentive schemes. We examine the relationships between managerial pay and firm performance among domestic and global firms using economic value added (EVA) and market value added (MVA) to assess wealth creation. Our work suggests that top managers in domestic- and globally focused firms are not only incented to increase EVA, but also rewarded for past additions to MVA. The results of our research suggest that managers of highly globalized firms tend to be paid at higher levels, reflecting the increased complexity of managing global firms.

Introduction

Previous studies have proposed that optimal executive compensation contracts perfectly align the interests of the executives with those of the firm's shareholders Grossman & Hart, 1983, Harris & Raviv, 1979. In theory, such contracts act as incentive mechanisms for executives to engage in behaviors that maximize the firm's value and reward executives for such behavior Fama, 1980, Jensen & Meckling, 1976. Whether executive compensation contracts meet this test of optimality, ex ante or ex post, is an empirical question subject to ongoing investigation (Tosi, Werner, Katz, & Gomez-Mejia, 2000).

Several studies have examined the relationships between measures of firm performance and top manager pay. For example, Murphy (1985) found a statistically significant relationship between the level of pay and performance, while Mehran (1995) found firm performance is positively related to management's ownership stake and to the percentage of its equity-based compensation. However, Jensen and Murphy (1990) did not find a significant relationship between changes in firm value and changes in executive compensation. Miller (1995) showed no support for a linear relationship between pay and performance, but found strong support for a convex relationship. Hadlock and Lumer (1997) found that pay–performance sensitivities have significantly increased over time for small firms, but not for large firms.

More recently, in a study examining the role of boards in setting managerial pay, Porac, Wade, and Pollock (1999) found evidence that boards make comparisons within and between industries in which the firm competes to support their top management compensation decisions. The authors conclude that boards of directors tend to “anchor their comparability judgments” by examining other firms' performance. This suggests that top manager performance is assessed based on relative measures and with an eye toward the industry environment affecting the firm.

Unfortunately, most of the studies exploring the nature of the relationship between managerial pay and performance have used accounting-based measures of performance (such as return on equity [ROE] or return on assets [ROA]). Such measures may bear little resemblance with the economic return earned by the firm since accounting-based measures do not account for the risk incurred by the firm's managers in their search for growth and profitability (Shiely, 1996). For example, earnings growth which may follow a decision to increase the size of the firm does not automatically lead to a per-share growth in firm value because the former may be achieved at excessive capital costs (Copeland, Koller, & Murrin, 1995). In addition, even studies using measures of performance based on market returns fail to adjust returns for the level of risk exposure (Harris & Raviv, 1979). Thus, the exact relationship between pay and performance can be somewhat different than what the empirical results suggest because the impact of risk is not adequately accounted for in commonly employed measures of performance Lehn & Makhija, 1996, Stewart, 1991.

Our study is designed to further clarify the nature of the pay–performance relationship by adding risk to the equation. Specifically, we seek to investigate the relationship between top management compensation and two measures of risk-adjusted firm performance: economic value added (EVA) and market value added (MVA). EVA and MVA are measures developed and trademarked by the Stern Stewart and Co. First suggested by Stewart (1991), EVA can be thought of as a proxy for the measurement of economic returns. It is the firm's residual profitability in excess of capital costs. A firm's EVA is positive when after-tax operating profits exceed the dollar cost of capital (COC). MVA is a closely related measure in that it is the present value of all expected future EVA and can be thought of as the net present value of the firm.

Variations of these measures have been proposed, and used, by others Copeland et al., 1995, Rappaport, 1986. However, EVA and MVA have received wider attention both in the corporate world and in scholarly research (see for example, Hodak, 1994, Lehn & Makhija, 1996, Spinner, 1995, Tully, 1993, Uyemura et al., 1996). Included among these are studies that have attempted to document the presence (or lack thereof) of a relationship between EVA and measures of stock price performance. Dodd and Chen (1996), for example, report that EVA explains only slightly more than one fifth (20.2%) of the variation in stock returns for a sample of 566 firms, comparing unfavorably with ROA, which explains almost a fourth (24.5%) of the variation. Further, Dodd and Johns (1999) found some differences in performance between the adopters and nonadopters of EVA. However, as documented by Weaver (2001), there are significant differences in how firms measure EVA. Examining a set of 29 EVA adopters who participated in his survey, he found that none of the respondents measured the EVA the same way. Directly relevant to our purpose is Weaver's finding that the adopters' top reasons for implementation of EVA are to “enhance financial management” and to “enhance compensation metrics.” Also of direct interest to this work is Kramer and Peters' (2001) finding that, as a proxy for MVA, EVA does not suffer from any industry-specific bias. However, they also conclude that EVA is consistently outperformed by the “net operating profit after tax” measure.

We believe EVA and MVA are reasonable proxies for the measurement of owner wealth maximization while taking into account the relative risk-based costs of doing so Hodak, 1994, Shiely, 1996. However, the relationships between executive compensation and these measures of firm performance have not yet been explored empirically. In this study, we seek to examine these relationships. Under the pay-for-performance hypothesis, we expect a positive relationship between executive compensation and firm performance. In this study, firm performance is measured in the context of value creation for the owners of the firm using MVA and EVA. We also examine the contribution of these measures in explaining the cross-sectional variation in executive compensation relative to the accounting-based measure of ROA.

Executive compensation generally consists of several components such as salary, bonus, stock options, and long-term incentive payments. It is plausible that certain components, such as bonuses, are used as a reward for past performance, while other components related to firm value are designed to provide the correct incentive for future performance (Murphy, 1985). The complex design of the total compensation package requires that we separately examine the relationship between firm performance and each of these components.

Recently, there has been an increase in the body of research pointing to the global managerial labor market as the basis for better understanding differences in levels of pay, as well as the mix of incentive plan components. That is, there is increasing evidence that top managers in highly global firms have higher proportions of performance-based pay in their total compensation contract (Carpenter, Sanders, & Gregersen, 2001). In addition, Richard (2000) suggests that pay policies that include equity participation practices pioneered in the United States are becoming common throughout the global business community. We hope to assess whether there are differences in the relationships between pay and performance based on the level of international impact of the firm.

We also seek to examine the causal order of the relationship—that is, whether compensation serves as a reward for past performance, or as an incentive for enhanced future performance. Given shareholder wealth maximization as the goal of the firm, is the executive compensation scheme used by the firm incentive-compatible? To answer this question, we use leading and lagged values of firm performance. If compensation is an incentive for top managers to perform in certain ways, then we would expect top manager pay to be unrelated to past performance, but instead would observe a positive relationship between current compensation and future performance. On the other hand, if compensation is a reward for superior performance, we would expect top manager pay to be positively related to past performance, but not to future performance Gomez-Mejia et al., 1987, Tosi et al., 2000.

The rest of our paper is organized as follows. In the next section, we describe our data. We present and discuss our results in Section 3 and our conclusions and implications for further research are presented in the last section.

Section snippets

Data

Executive compensation and firm performance measures used in this study are obtained from three sources: the Standard and Poor's ExecuComp database, Stern Stewart and Co.'s Performance 1000 database, and Standard and Poor's Compustat database.

We define top managers as individuals with the title of Chairman, CEO, President and senior-level Vice President. Compensation data are obtained from the ExecuComp database. In our study, we use three measures of compensation: salary, bonus, and total

Results

We investigate the relationship between executive compensation and firm performance using cross-sectional regression analysis. The dependent variables in all our cross-sectional regressions are the three components of compensation (salary, bonus and TDC) adjusted for firm size. For brevity, we refer to compensation components generically as COMP. In all regressions, the t statistics are based on White's (1980) heteroskedasticity-consistent estimators of the standard errors of the model's

Summary and conclusions

In this study, we examine the relationship between executive compensation and measures of performance capturing the economic profit earned by the firm (EVA and MVA). Consistent with prior studies examining top manager compensation, we desegregate the compensation package of top managers into cash, merit pay (bonus), and TDC, which includes long-term incentives such as stock options. We also investigate the causal direction between the wealth creation activities of the firm and the compensation

References (40)

  • J.R. Deckop

    Determinants of chief executive officer compensation

    Industrial and Labor Relations Review

    (1988)
  • J.S. Demski et al.

    Hierarchical structure and responsibility accounting

    Journal of Accounting Research

    (1989)
  • J.L. Dodd et al.

    EVA: a new panacea?

    Business and Economic Review

    (1996)
  • J.L. Dodd et al.

    EVA reconsidered

    Business and Economic Review

    (1999)
  • E. Fama

    Agency problems and the theory of the firm

    Journal of Political Economy

    (1980)
  • R. Freedman

    Long-term incentive plans: U.S. multinational firms erase the borders

    Journal of Compensation and Benefits

    (2000)
  • L. Gomez-Mejia et al.

    Managerial control, performance and executive compensation

    Academy of Management Journal

    (1987)
  • S.J. Grossman et al.

    An analysis of the principal agent problem

    Econometrica

    (1983)
  • C.J. Hadlock et al.

    Compensation, turnover, and top management incentives: historical evidence

    Journal of Business

    (1997)
  • B. Hall et al.

    Are CEOs really paid like bureaucrats?

    Quarterly Journal of Economics

    (1998)
  • Cited by (0)

    View full text