基于矩信息的不确定投资组合优化问题

Uncertain portfolio optimization problem based on moment information

在已知随机变量的部分矩信息的条件下,对带有机会约束的投资组合优化问题进行研究.首先,对初始机会约束进行鲁棒性处理,得到分布鲁棒机会约束优化问题;然后,通过引入条件风险价值(CVaR)来近似地表示分布鲁棒机会约束,给出最坏情况下的CVaR近似且表明其精确性.

Under the condition that the partial moment information of random variables is known,the portfolio optimization problem with chance constraint is discussed.Firstly,the initial chance constraint is handled by robust method,which obtained distributionally robust chance constraint optimization problem.Then,the distributionally robust chance constraint is approximately represented by introducing the Conditional Value at Risk(CVaR).Finally,the Worst-Case CVaR approximation is given and its accuracy is shown.