推广了连续时间均值-方差投资策略模型,其中风险资产价格受参照因子的影响,而且投资策略终止时间也由参照因子确定.模型化为一个具随机周期的随机最优控制问题,该问题可通过一个辅助的随机LQ模型求解.解决随机LQ模型的关键是导出了两个偏微分的初边值问题,利用Feynman-Kac表示定理得出了这两个偏微分方程问题的解,进而求出了模型的最优投资策略.
This paper generalizes the continuous-time mean-variance investment policy model, where the price of risky asset are affected by a factor of reference and the terminal time of investment policy is determined by the factor of reference. The model is described by a stochastic optimal control problem with random horizon which can be solved through an auxiliary stochastic linear-quadratic (LQ) problems. The key to solve the stochastic LQ is deducing of two partial differential equations for initial bounded problems. Through Feynman-Kac representation theorem, we obtain the solutions of the two partial differential equations. Therefore,we can derive the optimal investment policy. This is a more practicable model. As a result,it meets the need of investors better.