基于CVaR的相对鲁棒投资组合问题研究

Relative Robust Portfolio Selection under Conditional Value at Risk

鲁棒优化方法是处理不确定环境下决策问题的有效技术,已在众多领域得到广泛应用.为降低现有鲁棒投资组合选择模型的鲁棒性成本,避免结果过于保守,本文提出了具有优良特性的相对鲁棒CVaR风险度量,探讨了其计算等问题.由其所导出的鲁棒投资组合选择模型的转化、简约与求解等问题,为求解实际的金融投资决策问题奠定了基础.

Robust optimization is an efficient technique for coping with decision problems under uncertainty and has been applied to many fields. To reduce the robustness cost of the existing robust portfolio selection models and to avoid the too-conservative-problem of the se- lected portfolio, we propose a relative robust CVaR risk measure with good properties, discuss its computation and establish the corresponding portfolio selection model. Then we investigate the transformation, simplification and efficient solution to the established relative robust port- folio selection problem. The obtained results establish the foundation for solving real financial investment problems.