经济附加值与高管薪酬的实证研究【外文翻译】.doc

本科毕业设计（论文）文 献 综 述原文二：An Empirical Investigation of EVA and Executive CompensationABSTRACTDespite a growing literature, the relationship between the structure of executive compensation and firm performance is not fully understood. Furthermore, little work has been done on the link between Economic Value Added as a measure of firm performance and the form of executive compensation. An examination of the compensation structure and economic value added of 209 companies in 1995 1998 provides evidence supporting incentive compensation. Economic Value Added is found to be positively and significantly related to incentive based compensation. Cash based remuneration, was found to be unrelated to EVA performance.1.0 INTRODUCTIONThe objective of this paper is to evaluate the alignment of EVA with management pay incentives in an agency theory context. Agency theory has been widely applied to the study of executive compensation in an endeavor to establish whether executive rewards are being set in a manner consistent with the theory. The theory posits that the sensitivity of executive pay to firm performance will be dependent upon risk sharing considerations, the level of monitoring of the executive by the owners and the degree to which their respective incentives are aligned. Despite extensive study, Finklestein and Ham brick (1996, p.285) suggest that in sum, a significant gap seems to exist between economic predictions of the consequences of executive compensation and actual empirical results. Results for models of CEO incentives to both market and accounting return measures have produced only small explanatory power (Rosen, 1990). This study broadens prior research in the area of executive compensation by using EVA as a measure of value delivered to shareholders and reporting its alignment with CEO pay outcomes.2.0 BACKGROUNDAgency theory requires the separation of ownership (principal) from the management (agent) of a firm. This division leads to agency costs, where managers exploit their superior information (information asymmetry) to maximize their own utility. Rosen (1990), in distilling empirical executive compensation research, finds that the semi log elasticity of executive compensation with respect to accounting rates of return is around 1.0, compared to a stock market return of approximately 0.1. There has been much argument over the normative level of this elasticity with no consensus emerging; however, there does appear to be general agreement that the correlation is weaker than expected and that accounting returns are a more reliable predictor than market-based measures. (Refer to Rosen 1990; Pelvic, Scott and Tisane 1993 and Hillock and Murphy 1999, for reviews.) With respect to the magnitude of the correlation, Jensen and Murphy (1990) argue that the relationship of executive pay relative to shareholder stock value has reduced in the past 50 years. They believe the upper and lower ends of executive pay distributions have been truncated resulting in an absence of management incentives. Some authors have posited that market returns are the only correct measure, as these represent the real return to owners and are not at the discretion of management (Coughlin and Schmidt 1985, p.48). Jensen and Murphy (1990) point out that paying executives according to accounting profits provides incentives to both manipulate accounting policy choice and to favor projects with short term profit returns over those that add long-run value. This view, held by many financial economists, has led to a concentration of research on market returns at the expense of the more closely correlated accounting returns. However, despite stock price performance being a popular base for analyzing management compensation plans this base is not always a good indicator of whether management have created or destroyed value during a given year. Individual share prices will fluctuate and even within the context of an efficient market will not necessarily always reflect the true worth of a firm.Jarrell (1993) notes that positive accounting theory based on the premise that accounting methods develop to provide more cost-effective measures of performance in the absence of regulation or tax effectssupports the use of accounting measures of earnings. That is, the almost universal use of this measure in compensation contracts suggest that it is the most efficient available. Jarrell (1993, p. 80) further argues that the use of a profit measure shields executive compensation from market-wide fluctuations in equity values that are not caused by expected changes in fundamentals. This view is supported by Jarrell and Donkey (1992) and Sloan (1991), who both find strong correlations between accounting returns and market-based returns. The former discovered a stronger relationship between accounting returns and market returns of individual companies than the market returns of a company and the market index over a five-year period. EVA provides a potential alternative measure of performance that goes beyond the traditional market and accounting measures (albeit rooted in them) mentioned above. It therefore provides the opportunity to expand our knowledge of the determinants of executive pay. A number of studies attest to the efficacy of EVA as a measure of company performance. Talley (1999) summarizes a study that reveals superior stock market performance of companies which have adopted EVA compared to competitors using other valuation methods. The study, conducted by Stern Stewart, comprised of 67 publicly owned US EVA clients were compared to ten firms with similar Standard Industrial Classification Index (SIC) codes over a five year period. The findings suggest that EVA adopting companies consistently outperformed their competitors in terms of total returns to shareholders. Chen and Dodd (1997) compared EVA to traditional accounting measures of valuation, EPS, ROE and ROA. A regression model of both EVA and accounting variables revealed that EVA contained significant information beyond traditional accounting measures. Research by Lean and Makita (1997) on 452 firms for the period 1985 1994 compared ROA, ROE, ROS (return on sales), RET (stock performance), EVA and MVA. Among their findings were that stock returns and EVA had a correlation coefficient of .59. Other accounting measures, ROE, ROA and ROS had coefficients of .46, .46 and .39 respectively, indicating a stronger correlation between EVA and stock return than the accounting measures.3.0 DEVELOPMENT OF HYPOTHESESAs outlined above numerous studies have identified a positive relationship between executive compensation and firm performance, although debate continues as to the exact size of this function. Using Australian data, Evans and Stomach (1994) find lower (and statistically insignificant) elasticity when compared with US studies. Their results for executive pay with respect to accounting rates of return was statistically significant but with very small coefficients for the profit variable. This result is supported by a more recent Australian study that found that “consistent with prior allegations, it appears that Australian CEOs have had, by international standards, a relatively small proportion of total compensation at risk” (Iran, Sedum and Taylor 1998, p.46). Despite the relative size of the effect, both studies supported a positive pay-performance relationship. No prior Australian study could be located which tested the effect of executive remuneration on EVA returns. A recent US study using 1996 cross-sectional survey based data found a modest positive relationship between EVA and CEO base salary, cash bonus and long term compensation (Sheikholeslami, 2001). Interestingly, they also find no relationship with CEO option holdings, proffering that perhaps compensation committees hope that granting stock options will produce (future) results not currently reflected in (or anticipated by) current earnings and/or stock prices (p. 16). Nevertheless, the general finding of the study lends support to the contention that CEO cash compensation and shareholdings are positively related to EVA. Indirect support for this finding may also be gleaned from the many studies reporting a positive pay-performance relationship (Rosen, 1990) when the performance measures used, both market and accounting based, contain elements which are used in the computation of EVA. Given the above, it would be expected that EVA, which is based on accounting returns but includes market elements (market value of equity and the capital charge) would similarly be positively related to executive compensation.Hence:H1: Ceteris paribus, CEO cash and bonus pay will be positively related to compute EVA.Agency costs arise where managers exploit their superior information to maximize their own utility. Where the CEO has significant investment in the company, the division between owners and managers is minimized and should, in theory lead to a reduction in agency costs. To the extent that companies have CEOs with significant share holdings they essentially become management-controlled-and-owned companies, and therefore are less subject to moral hazard problems (Antler and Smith, 1985).Support for this is found in Lambert and Larker (1987) in a study of executive compensation effects of large corporate acquisitions. In establishing that CEO compensation could not be increased through the selection of acquisitions that reduced shareholder wealth, they noted: the executives of the firms whose acquisitions were associated with a negative stock price reaction typically owned a smaller percentage of their firms stock than did the executives of the firms whose acquisitions were associated with a positive stock price reaction (p. 238). Other evidence of the mitigating effect of stock ownership on aberrant management behavior can be found in Dahlia, Salomon and Smith (1982) where it is shown that managers with significant shareholdings in their companys stock were less likely to manipulate earnings. This is consistent with Decor and Sloans (1991) finding with respect to opportunistic reduction in research and development expenditures. In their investigation of the horizon problem in firms, they find that executives in their final years spend less on research and development; the implication being that short-term accounting performance is managed to maximize the executives eventual retirement benefit. They also establish that this effect was reduced in firms where executives held shares or options. Similarly, Kaplan and Atkinson (1989, p. 724), in promoting the use of equity linked performance measures over accounting results, provide an extensive list of actions managers may take to improve reported earnings and yet provide no long-run benefit to the firm. They further argue that executives could decline investments that increase the net present value of the firm but penalize short-run earnings. Taken together, these studies lead to the proposition that the remuneration of managers with significant share holdings will be more related to performance effects and less to scale effects than for those with insignificant share holdings. Hence:H2: Ceteris paribus, the proportion of firm equity held by the CEO will be positively related to compute EVA.Under agency theory, the agent (CEO) is attempting to maximize their utility within the constraints imposed by the principal (owner). In the situation where only the agents output is observable, but his/her effort level is not, Hailstorms (1979) model solution includes an incentive to the agent based on output. In these terms, we could view the cash component as largely equating to insurance (not completely as it may include an incentive based bonus) and equity returns as incentives. The CEOs exposed to the greatest incentive, will be those with the largest incentive-based payments relative to insurance, all else being equal. Llewellyn, Launderer and Rosenfeld (1985) found that executives with personal share holdings that were relatively small when compared with their salary package were more likely to force their firms into acquisitions that reduced shareholder wealth. In this way the executives wealth would be maximized through the firm-size effect on salary. It may be inferred from this that executives with relatively high personal share holdings may be less interested in scale effects and more concerned with the factors that influence the market value of the firm. As such, the absolute value of the output-linked incentive (i.e. equity returns) is more meaningful when compared to the insured portion of their remuneration (i.e. base salary). In simple terms, a CEO with a salary of $1 million and equity holdings of $2 million would be considered to have an equivalent incentive to a CEO with salary of $5 million and equity holdings returns of $10million. In this way, an attempt is made to gauge the relative importance of the equity incentive. Hence:H3: Ceteris paribus, Incentive based pay as a proportion of total CEO pay will be positively related to compute EVA.6.0 CONCLUSIONAn examination of the relationship between executive compensation structure and firm performance finds a positive relationship between incentives based compensation and EVA. The evidence presented in the paper lends further support to those advocating incentive based compensation packages and is consistent with the theoretical work on Agency. It illustrates an additional dimension to the study of executive compensation beyond the traditional accounting and marketing measures, however in common with them, and consistent with the finding as of Sheikholeslami (2001) the explanatory power is relatively modest.Source: John Evans, Robert Evans, 1999. “An Empirical Investigation of EVA and Executive Compensation”. Journal of management Accounting Research ,vol.20, no.6, pp.23-25.译文：经济附加值与高管薪酬的实证研究摘要尽管文献越来越多，但其中阐述的中高层管理人员的薪酬结构与公司经营绩效之间关系并不完全清晰。此外，关于经济附加值作为公司业绩评价措施和高管薪酬结构之间的衔接工作还没有完全完成。1995年至1998年209个公司的薪酬结构和经济附加值的调查为支持激励补偿提供了证据。经济附加值与高管薪酬具有显著的正相关关系。现金的报酬，被认为是与经济附加值业绩无关的。1.0简介本文的目的是评价在代理理论背景下的基于管理人员的薪酬激励的经济增加值。代理理论已被广泛应用在对高管薪酬的研究，不管奖金制度在某种程度是否正在被设置，都试图达到与该理论相一致。该理论认为与公司业绩有关的高管薪酬的敏感性依赖于风险分担的注意事项，所有者对高管的监管水平和他们的各自的奖励是一致的。尽管进行了广泛研究，菲尼克斯特和汉姆布瑞克（1996年， 285页）认为，“总之，高管薪酬预测的结果与实际的实证结果之间似乎存在着一个巨大的缺口。”关于市场和会计回报措施的首席执行官的激励模型的结果已经产生，但只有很小的解释力（罗森，1990）。本研究扩大关于高管薪酬的研究领域，使用了经济附加值，经济附加值作为一种向股东交付价值和报告首席执行官的薪酬的测量手段。2.0背景代理理论要求一个公司的管理层（代理人）的所有权（委托人）分离。这种分离产生了代理成本，管理人员利用其优越的信息（信息不对称），使自身的效用达到最大化。罗森（1990年），在高管薪酬的实证研究中，发现了股票市场回报率大约在0.1，关于高管薪酬的投资报酬率半对数弹性在1.0左右。人们关于这些弹性一直没有形成一致的标准，产生很大的争议，但又普遍认为以市场为基础的措施的相关性弱于预期会计收益的相关性，后者更为可靠。（参照1990年詹森帕夫利克、斯科特谭森 1993年、哈洛克墨菲1999年的评价）关于规模的相关性，詹森和墨菲（1990）认为，与股东股票价值相关的高管薪酬的相关性在过去50年里已经减少。他们认为高管薪酬分布的上下两端已被截断导致缺乏管理层激励机制。有些作者已经假定，市场回报率是唯一正确的度量措施，因为这些代表所有权人的真正回报而不是取决于在管理人员的判断（考夫兰和施密特1985年，第48页）。詹森和墨菲（1990）指出，根据会计利润支付管理人员薪酬为操纵会计政策的选择以及偏好那些短期的利润回报的项目，增加长期价值提供了诱因。这种观点，被许多财务经济学家提出，已经形成了对以牺牲更为关联性的会计回报为代价的市场回报的研究热潮。然而，尽管股票价格表现作为分析管理人员薪酬分配计划是一个流行的指标，但不管是管理人员在被授予的一年里是创造了还是减少了公司价值它都不是好的指示器。个别股票价格将出现波动，甚至在一个有效市场背景下，不一定总是反映一个企业的真正价值。贾雷尔（1993）指出，积极会计理论 - 基于会计方法的发展在不考虑法规及税收影响下能提供更具成本效益的绩效评价方式的前提下支持的会计收入度量措施的使用。也就是说，当这一度量方式在薪酬合同中几乎被普遍的使用时表明它是最有效的。贾雷尔（1993年，第80页）进一步指出，一个利润测量的用途“防止高管薪酬的变动由于股票价值在市场上广泛的波动而导致的而不是由于预计的理论上的变动导致的。贾雷尔和东瑞克（1992）和斯隆（1991）支持这种观点，他们两个都发现了会计收益和市场为基础的回报有很强的相关性。前者发现会计回报和市场回报之间存在很强的相关性，个别的公司比一个公司的市场回报更相关，而且这个市场指数是超过五年期。经济附加值提供一个潜在可选择的业绩评价措施，超越了如上所述的传统的市场和会计核算方法（尽管是以它们为基础）。因此，它为扩大了我们关于高管薪酬的决定因素的知识面创造了机会。大量研究证实了经济附加值衡量公司绩效的效果。塔利（1999）总结了一项研究，揭示了采用经济附加值的公司与使用其它评价方式的竞争对手相比拥有更加卓越的股票市场绩效。这项研究由斯特恩斯图尔特进行，将美国的由公众持股的用经济附加值考核的67家企业与存续五年以上在类似的标准产业分类指数内的十家公司进行比较。调查结果表明，采用经济附加值的公司一贯的总的回报表现优于其竞争对手。陈和多德（1997年）将经济附加值和传统的会计核算办法、每股收益、净资产收益率和资产回报率进行了比较。关于经济附加值和会计变量的回归模型显示，经济附加值比传统的会计核算办法容纳更多的重要信息。莱恩和马克哈甲（1997）对1985年到1994年起期间的452家公司的资产回报率、净资产收益率、销售回报率、股票表现、经济附加值和市场增加值进行了比较，他们的研究结果中发现，股票收益和经济附加值的相关系数为0.59。其他会计指标数据，净资产收益率、资产报酬率和销售回报率的相关系数分别为0.46，0.46和0.39，表明经济附加值与股票回报的相关性比与会计核算方法的相关性更强。3.0假设的发展正如上文所述许多研究已经发现了经理人报酬与公司业绩的正相关关系，尽管关于这个函数的确切大小，争论仍在继续。埃文斯和斯达发敕（1994）利用澳大利亚的数据，在与美国的研究数据比较时，发现了一个低的弹性关系（统计学意义上是无关紧要的）。他们的结果表明，基于会计收益率的管理层薪酬支付在统计上对利润变量具有显著性但是却是一个非常小的系数。澳大利亚的一项最近研究成果支持这一结论，他们得出 “与以前的主张一致，按照国际标准，澳大利亚的首席执行官们只是获得总的风险补偿金的一个相当小的比例”（威山，西胡和泰勒1998年，46页）。暂且不管研究结果的作用的相对大小，这两项研究都支持了薪酬绩效的正相关关系。没有先前的位于澳大利亚的研究就不可以测试出基于经济附加值的薪酬回报的影响。美国最近的一项研究使用了1996年具有代表性的调查研究的原始数据，发现经济附加值与首席执行官基本工资、现金红利和长期现金补偿之间存在适度的的正相关关系（沙克好勒拉米，2001年）。有趣的是，他们也发现经济附加值与首席执行官的股份期权没有关系，指出也许是薪酬委员会希望授予股票期权会创造（未来的）业绩而不是当前就反映在当期（或预期的）收入和/或股价上（第16页）