基于隐Markov模型的最优资产组合选择  

Optimal Portfolio Selection under Hidden Markov Model

作　　者：张玲[1] 

机构地区：[1]广东金融学院经济贸易系,广东广州510521

出　　处：《经济数学》2014年第2期23-28,共6页Journal of Quantitative Economics

基　　金：教育部人文社会科学基金资助项目(13YJCZH247);广东省哲学社会科学基金资助项目(GD12XYJ06);广东金融学院资助项目(12XJ02-10)

摘　　要：在具有可观测和不可观测状态的金融市场中,利用隐马尔可夫链描述不可观测状态的动态过程,研究了不完全信息市场中的多阶段最优投资组合选择问题.通过构造充分统计量,不完全信息下的投资组合优化问题转化为完全信息下的投资组合优化问题,利用动态规划方法求得了最优投资组合策略和最优值函数的解析解.作为特例,还给出了市场状态完全可观测时的最优投资组合策略和最优值函数.This paper studied a multi-period optimal portfolio selection problem in the financial market consisting of the observable and the unobservable market states where the dynamics of the unobservable market states is described by a hidden Markov chain.By using the sufficient statistic method,the portfolio optimization problem with incomplete information was converted into the one with complete information.The optimal investment strategy and the optimal value function were derived in closed-form by adopting the dynamic programming approach.The optimal portfolio strategy and the optimal value function in the special case where the market states are completely observable were also presented.

关 键 词：不完全信息 隐马尔可夫链 充分统计量 基准准则 动态规划 

分 类 号：F830.59[经济管理—金融学] F224.3