Executive Pay And Company Performance Analysis Endogenous

Roberts and Whited proposed that endogeneity is the pivotal problem in empirical research of corporate finance. Executive compensation theory, a typical research subject of corporate finance, will inevitably encounter this problem. Endogeneity is defined as the relevance between explanatory variables and the error term. If a model is endogenous, the coefficient estimated by ordinary least squares is biased. There are three causes of introducing endogeneity:omitted variables, measurement error and interaction. The omitted variables and measurement error can be solved by fixed effect model, while the interaction can be either solved by instrumental variables model or simultaneous equation model. In this paper, a model of executive compensation and company performance is constructed from the aspect of endogenous analysis. The relationship between executive compensation and company performance is discussed. This paper also provides an approach to solve endogeneity in other research topics of corporate finance.Executive compensation theory, deriving from moral hazard theory, aims to solve the agency problem between shareholders and top management. Theoretically, better company performance leads to higher executive compensation and vice versa. Such compensation mechanism can motivate executives to make decisions for the benefit of shareholders. Since there are omitted variables and interaction, executive compensation endogenously influence company performance, this paper assumes that executive compensation and company performance are endogenous variables. Corporate governance variables and the corporate characteristic variables are assumed to be exogenous variables. Based on the above theories, this paper utilizes fixed effect model to deal with the endogeneity caused by omitted variables and instrumental variables model and simultaneous equations model to deal with endogeneity caused by interaction.Results of empirical study reveal a positive relationship between executive compensation and company performance. Compared with the pay performance coefficient estimated by OLS, the one estimated by fixed effect model, the instrumental variables model and simultaneous equations model has a much bigger value. That is to say, company performance and executive compensation are more positively related. In addition, corporate governance variables have a significant impact on executive compensation and board of directors cannot effectively supervise executive compensation. Corporate characteristic variables also have a significant effect on executive compensation. The thesis provides an approach to solve the endogeneity in empirical studies of corporate finance. The first and foremost step for solving endogeneity problems is to clarify the endogenous variables and verify the causes of having endogeneity. The next step is to construct an appropriate model to eliminate endogeneity. The endogeneity caused by omitted variables can be solved by fixed effect model while that caused by interaction can be solved by instrumental variables model or simultaneous equation model.