【正文】 s ownership is lower than the median CEO ownership for 1993 and equal to zero otherwise), and (ii) firms that have been targeted by shareholder activists using various targeting definitions with data from Strickland et al. (1996) and Wahal (1996). The sensitivity to stock returns increases both for low ownership firms and for both definitions of milliondollar firms after 1993 pared to firms with high ownership or firms that are not milliondollar firms with high ownership. A similar yet weaker pattern is also observable for the sensitivity to EPS. Overall, these results suggest that sensitivities to performance measures increased after 1993 both for milliondollar and low ownership firms. This result is consistent with the notion that 162(m) affected the pay for performance sensitivity and that boards of firms with low CEO ownership increased the pay for performance sensitivity of pensation contracts to align CEO and shareholder interests. We do not find such a result for interactions with the shareholder activist targeting dummy. Overall, our results show that although the governance variables matter, they do not affect our prior results, namely that milliondollar firms have increased the sensitivity of pensation to performance beyond what other firms have done following the enactment of 162(m). These results are not tabulated for brevity. 7. Concluding ments The objectives behind the adoption of the new SEC Compensation Disclosure Rules and Section 162(m) were to reduce excessive CEO pensation levels and strengthen the pay for performance relation. We formulate five propositions to test the effect of these pensation regulations using 1991–1997 CEO pensation data. We find that while the regulations have not achieved the objective of reducing CEO pensation growth, they appear to have had a statistically and economically significant impact on the pensation structure of firms most likely to be affected by the regulations. We also document an impact on the level and structure of pensation for an easily identifiable subset of firms that reduce CEO salaries from more than one million dollars to one million dollars or below. Twentythree of the 25 firms doing so refer to 162(m), however, total pensation still increases for a majority of these firms. We also find an increasing relation between stock returns and both bonus payments and total pensation after 1993 for all firms. Consistent with the law39。 and (3) the regulations were directly targeting this annual flow, especially cash pensation. We recognize, however, as documented by Jensen and Murphy (1990) and Hall and Liebman (1998), that the explicit relation between CEO stock and option holdings and shareholder wealth provides most of the CEO incentives and we focus on this aspect in the second part of the paper. We follow Hall and Liebman by examining three measures of panyrelated CEO wealth changes and find that the sensitivity of CEO wealth to shareholder wealth is higher than the one reported by Jensen and Murphy for the 1970s and 1980s. For instance, using the JensenMurphy statistic or sharing rate, defined as the dollar change in CEO wealth for a $1,000 change in firm value, we document median CEO sharing rates in 1996 from $7 to about $25 per $1,000 shareholder gain or loss, depending on the subgroup. This is in line with the Hall and Liebman finding that pay to performance sensitivity in the 1990s is substantially higher than the sensitivity documented by Jensen and Murphy for earlier periods. More importantly, in a regression framework we find that there is an increase in the wealth to performance sensitivity from 1993 to 1996 for firms approaching the milliondollar benchmark, . firms affected by 162(m). The paper is organized as follows. In Section 2, we discuss the background for the changes in the SEC Compensation Disclosure Rules and 162(m). In Section 3, we discuss our data collection procedure. Section 4 presents our empirical results on the effects of the regulation on pensation structure, salary levels and growth rates, and pensation to performance sensitivity. Section 5 evaluates the effects of the regulations on the sharing rate and other measures of the sensitivity of CEO wealth on changes in shareholder wealth. We examine the robustness of our results in Section 6 and offer concluding ments in Section . 2. The SEC39。s first release on July 2, 1992, was to make disclosure of pensation paid or awarded to executive officers clearer and more concise, and of greater utility to shareholders. Murphy (1995) explains that the SEC39。再有，我們通過對百萬美元企業(yè)的員工的薪酬研究時發(fā)現(xiàn)：相對其它企業(yè)而言 , 從1993 年到 1996 年間，百萬美元企業(yè)的 CEO 的薪酬敏感性對其他股東財富所產(chǎn)生的影響力增加了。我們通過使用 1991～ 1997 的 CEO 的 薪酬數(shù)據(jù)來說明這個問題。而 Core 和其他人一起 (1999)，研究了政策的出臺對 CEO 薪酬的影響 。這個數(shù)據(jù)局限了我們測試的可信度。薪酬數(shù)據(jù)被分為七個類別，在薪酬摘要表中包括薪 金、獎金以及其它的每年的報酬 (包括額外補貼和對稅收的補償 ), 有限的股票獎勵、 SARs，和長期刺激性計劃的支出 (LTIP)以及其它報酬。因而 ,他們討論的結(jié)果是：管理者傾向于選擇披露夸大公司效益的數(shù)據(jù)來獲得更高的薪酬。在第六部分我們檢驗我們研究的結(jié)果是否正確。我們和霍爾、 Liebman 研究 CEO 薪酬的變動與公 司的效益相關(guān)時發(fā)現(xiàn)： CEO 對薪酬的敏感性高于股東對其股權(quán)的敏感性，這與 Jensen 和墨菲在 70 年代和 80 年代所得出的研究成果的相似。 (4)所有企業(yè)和特別是那些受到 162(m)影響的企業(yè) , 我們發(fā)現(xiàn)在 1993 年后這些企業(yè)的獎金和總薪酬對股票收益的敏感度增加了。這個廣泛的定義囊括了我們進行研究的樣本企業(yè)中相當大的一部分。 在本文，我們求證那些章程的出臺和 162 (m)是否真的影響到薪酬的支付。[部分國內(nèi)稅收代碼為 162(m), 從此該部分定為 162(m) ]?！? Greenspan 的擔心不是最近才有的。我們進一步發(fā)現(xiàn)，在 1993 年后，許多企業(yè)的獎金和總薪酬對股票的收益越來越敏感了。我們還發(fā)現(xiàn)：我們對其他的控制因素一旦影響了 CEO 們的工作熱情， CEO 對其財富的擔心也會對企業(yè)的其他股東的財富產(chǎn)生影響。例如 Jensen 和墨菲 (1990), 分析和討論了在 70 年代和 80 年代低水平的薪酬都直接導(dǎo)致了員工工作的低效率。 SEC 組織 透露這個政策的目的是 ,依照 SEC 在 1992 年 7 月 2 日的發(fā)布內(nèi)容，首先薪酬將以一個更為清楚和更為簡明的方式來發(fā)放，并且也將會將以一個對股東更有利的形式來發(fā)放。首先，我們區(qū)分那些存在于 ExecuComp 數(shù)據(jù)庫的、并且可能受 162 (m)章程的影響的企業(yè)，并從此將其稱為百萬美元企業(yè)?？傊?, 我們的結(jié)果顯示這三個定義對我們的研究是很適用的。薪酬增加對股票收益的敏感性其 實不應(yīng)該歸因于：因為股票持有價值的上升而導(dǎo)致薪酬價值的減少。例如 , 使用 Jensen－ Murphy 統(tǒng)計或分享率 , 被認為在 CEO 財富上的一個美元的變化可以對公司產(chǎn)生 $1,000 上的變化 , 1996 年我們通過對小量樣本企業(yè)的研究，得悉了 CEO 從大約 $7 到 $25 的股票分享率都會對每 $1,000 個股東利益產(chǎn)生影響。在第七部分是結(jié)論。 在 1993 年成