基于马尔科夫状态转换下的CAPM实证研究  

The Empirical Research of CAPM based on Markov Regime Switching

作　　者：苏涛[1] 詹原瑞[1] 刘家鹏[1] 

机构地区：[1]天津大学管理学院

出　　处：《系统工程理论与实践》2007年第6期21-26,共6页Systems Engineering-Theory & Practice

基　　金：国家自然科学基金(70573076);高等学校博士学科点专项科研基金(20050056057)

摘　　要：将一阶马尔科夫过程引入到经典资本资产定价模型(CAPM)中,从动态的角度提出基于马尔科夫状态转换下的资本资产定价模型,即允许反映市场风险载荷的β系数和条件方差呈现时变状况.实证显示该模型优于经典资本资产定价模型,并能够刻画动态变化过程,经典CAPM为该模型的特例.The first-order Markov process is introduced into traditional Capital Asset Price Model （CAPM）. The CAPM based on Markov regime switching is proposed from the dynamic view. That is, the market rist load coefficient fl and the condition variance parameters are allowed to time-vary. The empirical research shows that this model is better than traditional CAPM and can capture the switching process of different state .The traditional CAPM is only one of the special cases of this model.

关 键 词：马尔科夫过程 状态 EM算法 资本资产定价模型 

分 类 号：F830.91[经济管理—金融学]