密级
硕士学位论文
题目 影子薪酬模型在 CEO 薪酬粘性治理的应用
（英文） The application of shadow pay model in the governance
of CEO compensation stickiness
研 究 生 学 号:19052202008
研 究 生 姓 名:叶信坚
指导教师姓名、职称:王凌峰教授
申 请 学 位门 类:管理学硕士
学 科、专 业:企业管理
论 文 答 辩日 期:2022 年 6 月 5 日
独创性（或创新性）声明
独创性（或创新性）声明
本人声明所呈交的论文是我个人在导师指导下进行的研究工作及取得的研究成
果。尽我所知，除了文中特别加以标注和致谢中所罗列的内容以外，论文中不包含
其他人已经发表或撰写过的研究成果；也不包含为获得桂林电子科技大学或其它教
育机构的学位或证书而使用过的材料。与我一同工作的同志对本研究所做的任何贡
献均已在论文中做了明确的说明并表示了谢意。
申请学位论文与资料若有不实之处，本人承担一切相关责任。
本人签名： 日期：
关于论文使用授权的说明
本人完全了解桂林电子科技大学有关保留和使用学位论文的规定，即：研究生
在校攻读学位期间论文工作的知识产权单位属桂林电子科技大学。本人保证毕业离
校后，发表论文或使用论文工作成果时署名单位仍然为桂林电子科技大学。学校有
权保留送交论文的复印件和电子档，允许查阅和借阅论文；学校可以公布论文的全
部或部分内容，可以允许采用影印、缩印或其它复制手段保存论文。（保密的论文在
解密后遵守此规定）
本学位论文属于保密在____年解密后适用本授权书。
本人签名： 日期：
导师签名： 日期：
摘要
摘要
CEO 业绩型薪酬被用于解决代理问题，但股东与 CEO 之间信息不对称，使业绩
型薪酬无法完全规避CEO 的道德风险。普遍存在的 CEO 薪酬粘性现象不仅体现了业
绩型薪酬机制的缺陷，薪酬激励的失效，同时不利于公司长期发展，严重的甚至会引
发系统性风险。以 A 股市场2010—2019 年的制造业民营上市公司为样本设计 CEO
薪酬定价公式，解决CEO 薪酬粘性问题。CEO 薪酬定价公式的设计思路是：将行业
内的公司分组，并把薪酬业绩敏感性最高，也即薪酬粘性最低的一组作为影子市场，
选择影响 CEO 薪酬的前三个主要因素作为经验回归的变量，以影子市场数据作为样
本，拟合 CEO 薪酬定价的经验公式。
通过门限回归模型发现公司规模存在双门限效应，用该门限值将制造业样本划分
为小规模公司、中等规模公司和大规模公司三组，使用双固定效应面板回归模型研究
公司竞争压力对 CEO 薪酬业绩敏感性的影响，得到以下结论：（1）小规模公司中普
遍存在“重奖轻罚”现象，容易产生 CEO 薪酬粘性问题，使得以业绩为导向的薪酬
机制失效。（2）大规模公司中存在竞争信号“传递失真”问题，公司竞争压力增大不
容易从业绩上观察到，CEO 更容易获得超额薪酬，从而产生薪酬粘性问题。（3）中等
规模公司中，业绩型薪酬最能发挥激励作用，股东与CEO 之间信息不对称程度最低。
因此，选择中等规模公司分组作为构建影子薪酬模型的影子市场。
将影子薪酬公式应用于制造业全样本，回归分析发现，当实际薪酬-影子薪酬偏
离度越小，即CEO 实际薪酬越接近影子薪酬，公司业绩越好，表明影子薪酬能够实
现有效激励目的，从而可以治理 CEO 薪酬粘性。为了进一步检验对 CEO 薪酬粘性的
治理效果，选取了有10 年连续数据的313 家公司，分别计算 CEO 实际薪酬粘性和影
子薪酬粘性并进行对比。结果发现，影子薪酬粘性在[-5,5]区间内波动，实际薪酬粘性
在[-20,20]区间内波动，且离散程度远高于影子薪酬粘性。
本文提出影子薪酬模型，为CEO 薪酬粘性提供一种新的理论视角，同时，为确
定 CEO 薪酬提供一个操作性强的公式，可作为上市公司董事会及薪酬委员会制定
CEO 薪酬时的参考，以解决 CEO 薪酬粘性问题。
关键词：CEO 薪酬粘性；公司竞争压力；CEO 薪酬业绩敏感性；薪酬偏离度；
影子薪酬模型
I Abstract
Abstract
The principal-agent problem is often solved by using CEO performance-based
compensation, however, the information asymmetry between shareholders and the CEO
makes performance-based compensation unable to completely avoid the CEO;s moral hazard.
It’s not only reflects the flaw of the performance-based compensation, but also is harmful to
the long-term development of the company, or even leads to systemic risks in serious cases.
This paper tries to design a formula for determining CEO compensation that is CEO shadow
pay model which is by using samples from manufacturing private listed companies in the A-
share market in China from 2010 to 2019.The CEO shadow pay model design steps as
follows: first, group the companies and select the group as shadow market which with the
highest pay-performance sensitivity, that is, the lowest compensation stickiness, and then
use the top three main factors that affect CEO pay as the variables of empirical regression,
and finally use the shadow market samples to fit empirical formula for determining CEO
compensation.
The threshold regression model shows there is double-threshold effect when take
company size as a threshold variable. Manufacturing company samples could be divided into
three groups named small-scale, medium-scale and large-scale by two thresholds of
company size. Double fixed effects panel regression model suitable for study the relationship
between firm competitive pressure and CEO pay-performance sensitivity. We got the
conclusion as below: (1) “Heavy rewards and light punishments” is common in small-scale
companies, which is prone to the problem of CEO pay stickiness, making the performance-
based compensation mechanism ineffective. (2) “distortion of competitive signals” in large-
scale companies, which makes it company hard to observe the increase of the firm
competitive pressure from the performance. (3) Among medium-scale companies,
performance-based compensation can play the most motivating role, and the degree of
information asymmetry is the lowest between shareholders and CEOs. Therefore, the
grouping of medium-scale companies is selected as the shadow market for constructing the
shadow pay model.
Applying the shadow pay model to the full sample of the manufacturing industry,
regression analysis found that when the pay deviation is smaller, that is, the closer the CEO;s
real salary is to the shadow salary, the better the company;s performance, indicating that
shadow salary could achieve the purpose of effective incentives, thus could control CEO pay
II
Abstract
stickiness. In order to further test the control effect of CEO salary stickiness, 313 companies
with 10-year continuous data were selected to calculate and compare the CEO;s real salary
stickiness and shadow salary stickiness respectively. It turns out that the shadow salary
stickiness fluctuates in the [-5,5] interval, and the real salary stickiness fluctuates in the [-
20,20] interval, and the dispersion degree of the real salary stickiness is much higher than
the shadow salary stickiness.
This paper proposes a shadow pay model, which provides a new theoretical perspective
for CEO compensation stickiness. At the same time, it provides a operational formula for
determining CEO compensation. The board of directors and compensation committees of
listed companies can use it as a reference to formulate CEO compensation and solve the
problem of CEO compensation stickiness.
Keywords: CEO compensation stickiness; firm competitive pressure; CEO pay-
performance sensitivity; pay deviation; shadow pay model
III