基于收益序列相关的动态投资组合选择——动态均值-方差模型  

Dynamic portfolio selection based on serially correlated return-Dynamic mean-variance formulation

作　　者：许云辉[1] 李仲飞[1] 

机构地区：[1]中山大学岭南学院,广州510275

出　　处：《系统工程理论与实践》2008年第8期123-131,共9页Systems Engineering-Theory & Practice

基　　金：教育部人文社会科学研究规划基金(07JA630031);国家自然科学基金(70518001);高等学校全国优秀博士学位论文作者专项资金(200267)

摘　　要：以动态均值-方差模型研究基于收益序列相关的投资组合选择.利用嵌入法将动态均值-方差模型嵌入到二次效用模型当中,通过动态规划方法求得最优投资策略和有效边界的解析形式.还得到一个反映风险资产动态投资价值的指标,它能包涵序列相关性的影响,在静态情形下,该指标是夏普指数绝对值的单调递增光滑变换.风险资产投资价值影响最优投资,并决定有效边界斜率.最优投资策略使期望终期财富向目标值趋近,该目标值随风险资产投资价值递增.最后以收益服从AR(1)过程为例对最优投资策略进行数值计算,发现序列相关可以对最优策略造成显著影响.This paper investigates dynamic portfolio selection based on serially correlated returns for an investor who optimizes the mean and variance of the terminal wealth, Through embedding the mean-variance model into the quadratic utility model and then adopting dynamic programming, we get an analytical expression for the optimal investment strategy. The efficient frontier is analyzed. We also get an index measuring dynamic investment value of the risky asset, which can reflect the impact of serial correlations of returns on the investment value, and which is equivalent to the Sharpe Ratio in the static case. The investment value affects the optimal invest strategy and determines the slop of the efficient frontier. Under the optimal invest strategy, the expected terminal wealth tends to a target value which increases with the investment value of the risky asset. At last an example of AR（1） process is used to illustrate the impact of serial correlations on the optimal strategy, which is shown to be significant.

关 键 词：资产配置 均值-方差模型 序列相关 投资价值 动态规划 

分 类 号：F830.91[经济管理—金融学]