椭球不确定集下的投资组合鲁棒优化模型

Robust Portfolio Selection under Ellipsoidal Uncertainty

对于含有不确定参数的采用CVaR风险度量的投资组合模型,基于鲁棒优化理论的最新进展,结合统计或时间序列,构造形式较为简单的椭球不确定集作为对参数不确定性的近似,把原问题转化为易于求解的确定型最优化问题,解决了该模型由于参数具有不确定性所造成的缺陷,得到鲁棒性与最优性都较为满意的解.通过市场数据对模型的可操作性和实用性进行验证.

A study of the Ways to formulate and solve portfolio selection using CVaR strategy with data u robust folio selection problems based on recent zation were proposed. By using statistic theory and time series techniques, we uncertainty set which contained the most possible realizations of uncertain para problem was converted to a deterministic problem which could obtain a solution realizations of uncertainty parameters. To demonstrate our model and method, ments with real market data. ncertainty was conducted. progress in robust optimiconstructed an ellipsoidal meters. Then the original good for the most possible we did numerical experiments with real market data.