基于多元金融资产收益率分布模型的高阶动态投资组合研究  

Dynamic Portfolio with Higher Order Moments Based on Multivariate Distribution of Financial Assets Return

作　　者：彭胜志[1] 王福胜[2] 

机构地区：[1]东北林业大学经济管理学院,黑龙江哈尔滨150040 [2]哈尔滨工业大学管理学院,黑龙江哈尔滨150001

出　　处：《金融经济学研究》2013年第2期54-65,共12页Financial Economics Research

基　　金：国家自然科学基金项目(70972097);教育部人文社会科学青年基金项目(12YJC790150);中央高校基本科研业务费专项资金资助项目(DL11CC07)

摘　　要：针对金融资产收益率分布的非正态性和时变性问题,构建了多元风险资产收益率分布的动态模型,提出了模型识别、参数估计、模型检验方法,并运用所构建的模型,解决了一直以来限制高阶动态投资组合优化研究深入进行的难题:条件协偏度阵和条件协峰度阵的估计问题;同时根据所构建的时变模型,进行了高阶动态投资组合优化研究,将目前只考虑前两阶矩时变特征的动态投资组合优化问题扩展到考虑高阶矩时变特征的层面。With non-normality and time-variance of distribution of financial asset re-turn, the issue of dynamic portfolio selection with higher order moments is studied. The method of dynamic portfolio selection with higher order moments is deduced through Taylor expansion of utility function;the dynamic model of multivariate distribution of financial as-sets return is proposed to estimate conditional higher moments of portfolio. Through the mod-el, the time-variant characteristics of multivariate distribution of financial assets return could be described and conditional higher moments of portfolio could be estimated easily. In the empirical analysis of Chinese stock market, the theoretical results are verified.

关 键 词：动态投资组合 高阶矩风险 多元金融资产 效用函数 

分 类 号：F830.91[经济管理—金融学]