关键词: 金融工程, 投资组合模型, 鲁棒优化, 安全准则

Abstract: Considering the uncertainty in the real stock market, the paper regards the security return as an interval random variable and develops a robust mean-CVaR portfolio model based on the robust theory. Following the duality theory, the proposed model can be transformed as a linear programming problem, which reduces the computational complexity and contributes to solve a large-scale portfolio model. To consider investors’ safety requirement, a concept of the most violated probability is introduced, which can be used to adjust the conservatism of the proposed model and reflect investors’ safety requirement intuitively. The performance of the proposed model is empirically studied. The result shows that the proposed model can be used to construct portfolios that exhibit robustness against the expected return, meet investors’ safety requirement and are viable in real investment.

Key words: financial engineering, portfolio model, robust optimization, safety criterion