多期马尔科夫链动态投资组合模型

Markov Chain Dynamic Investment Portfolio Model

我们对模型参数随随机市场改变的多期最优投资组合问题，提出了一个新的均值一方差模型。它度量的是终端财富的增长倍数，这方便不同投资策略的比较以及不同投资个体的业绩评价。本文引入马尔科夫链描述跨期相关性。在构造了一个辅助问题并建立它们与原问题之间的关系之后，利用动态规划的方法求出了辅助问题的最优解的解析表达式，从而推导出我们模型的最优投资组合策略和有效前沿的解析表达式。

The authors give a new multi period portfolio optimization mean-variance model where the model parameters change according to a stochastic market. The model measured the growth multiplier of the terminal wealth. It is extremely convenience for comparation of different invests policy and performance evaluation. In our paper, we use Markov chain to describe the Cross-Correlation. We create an auxiliary problem which is solved using dynamic programming. The solution to the auxiliary problem is then manipulated to obtain the optimal mean variance portfolio management policy and the corresponding efficient frontier.