CVaR在信用风险优化中应用及基于遗传算法的解

The Application of Conditional Value-at-Risk Model in the Optimization of Credit Risk of Portfolio and the Solution Based on Genetic Algorithm

条件受险价值是一种能够反映损失分布尾部信息,从而有利于防范小概率极端金融风险的风险度量和优化工具。Fredrik给出了能同时优化组合条件受险价值和受险价值的线性规划模型,该模型存在维数障碍。将其重新变回为一个非线性规划模型,并利用带约束的遗传算法求其近似最优解。通过举例对比原组合与优化后组合的标准差、受险价值、条件受险价值,得出结论:该模型能够在很少损失期望收益的情况下,同时减少标准差、受险价值、条件受险价值等重要风险衡量指标。

CVaR is a new tool for credit risk measurement and optimization, which provides the tail information of loss and is favorable to keeping away the extreme finance risk with very little probability. Fredrik gives a CVaR model which can simultaneously optimize CVaR and VaR of portfolio. But this model has the drawback of dimension obstacle.This paper reverts it to a non-linear program model again and solves it by means of Genetic Algorithm with constraints.Compared with the original portfolios, the optimized portfolios' standard deviation, VaR and CVaR are obviously decreased with nearly the same expected yield rate.