Hibiki在文[13]中利用模拟路径提出一种Hybrid模型.文章在该模型的基础上作了一定的改进,利用GARCH模型得到相应的股票的时间序列价格;风险度量方法CVaR来控制风险.另外,考虑了交易费用和不允许卖空等市场客观因素.并且将模型转化为一种较易求解的线性规划进行求解,并利用模拟路径方法对本文模型与Hybrid模型进行的一些比较分析,数值实验表明了文中的模型可以更好的控制风险.
Recently, an alternative stochastic programming model with simulated paths was proposed by Hibiki,and it is called a hybrid model. We do some improvement of the Hybrid model in this paper as follows :apply the GARCH (generalized autoregressive conditional heteroskedasticity)model to derive the corresponding time sequence price of assets; control risk using CVaR (conditional value-at-risk). In addition, we also consider some market objective factors such as transaction costs and restricted short selling. And then we transfer the resulting model to a more tractable linear programming and use the simulated paths method to compare the new model with the Hybrid model. The numerical tests show that our model can control risk better than the Hybrid model.