| Summary: | 博士 === 國立中山大學 === 企業管理學系 === 87 === Agent theory starts with the assumption that people act in their own self-interest, and holds that under normal conditions, the goals, interests, and risks of two actors (main actor and agent) are not identical. This means that the agent will not necessarily act according to the interests of the main actor. Thus, a series of mechanisms are required to ensure that the agent will behave according to the main actor''s interests. CEO compensation and CEO replacement are types of control mechanisms that companies employ to reduce the agent problem.
Is it possible for an actor to behave exactly as is described in agent theory, regardless of the context or surrounding relationships? If it is not possible, then are the control mechanisms that agent theory proposes to control the behavior of the agent viable? This question is open for debate. Agent theory - that is, the presumption of self-interest, conflict of interests and goals, and asymmetric information - is unable to grasp the practical reality and special characteristics of Taiwan''s family businesses. This paper will thus discuss whether all companies are suitable to be researched using agent theory.
This paper will divide companies into family enterprises and non-family enterprises, and will research into CEO compensation and termination. We will investigate the influence of company performance, board control, external shareholder control and the possibly for company growth on CEO compensation. In terms of CEO termination, this paper will look at the influence of company performance on the effectiveness of CEO termination for family and non-family enterprises, the role of external board members and external shareholders in the termination process, and the influence of excessive compensation on CEO termination.
This paper will review literature pertaining to the above issues, and will develop a twelve point hypothesis. The samples used in the paper for CEO compensation will come from listed manufacturing companies (not including construction companies) between 1995-1997. The samples were verified using the LISREL model. The samples used in the paper for CEO termination research came from manufacturing companies in Taiwan (not including construction companies) listed between 1996-1997, and determined whether there were any changed in the CEO of companies between these years. The samples were verified using a logistic regressive model.
This paper''s conclusion is as follows:
1. We have confirmed that agent theory is suitable for non-family enterprises in Taiwan, and unsuitable for family enterprises. Agent theory has been found insufficient to understand all companies in Taiwan.
2. Within agent theory, the hypothesis that performance determines CEO compensation and the termination mechanism has been found to be true in non-family enterprises in Taiwan.
3. External board members play an important role in CEO compensation and termination in non-family companies in Taiwan.
4. An accounting basis is more important that a market basis for determining CEO performance.
|
|---|
