论文部分内容阅读
近年来的研究表明,CEO报酬不仅受公司业绩的影响,还受到其他诸多因素的影响,而且,CEO报酬与其决定因素之间往往存在着非线性关系。本文以2003-2005年沪深股市的A股上市公司为样本,采用BP神经网络对CEO总报酬、CEO年薪、CEO持股价值及其决定因素分别进行训练和学习,结果表明:(1)网络训练输出值与实际值的拟合度分别达到91.09%、97.23%和78.44%;(2)网络的预测能力相对于传统的线性回归模型分别提高了92.72%、92.08%和53.89%。因此,本文认为在分析和确定CEO报酬水平时引入神经网络模型是可行的。
Recent studies have shown that CEO compensation is not only affected by company performance, but also by many other factors. Moreover, there is often a nonlinear relationship between CEO compensation and its determinants. This article takes the A-share listed companies in the Shanghai and Shenzhen stock markets from 2003 to 2005 as samples, adopts BP neural network to train and learn CEO total reward, CEO annual salary, CEO shareholding value and its determinants. The results show that: (1) Network The fitting degrees of training output and actual values reached 91.09%, 97.23% and 78.44%, respectively; (2) The predictive power of the network was 92.72%, 92.08% and 53.89% higher than the traditional linear regression model respectively. Therefore, this paper believes that it is feasible to introduce the neural network model when analyzing and determining the CEO compensation level.