导　　师： 胡桂武

学科专业： 020208

授予学位： 硕士

作　　者： ;

机构地区： 广东财经大学

摘　　要： 风险度量模型CVaR/(Conditional-value-at-risk,条件在险价值/)有很多优点，它即弥补了VaR（Value-at-risk，在险价值）模型的不足，也满足一致性风险度量的要求，故一度受风险管理者的追捧。但在计算过程中，要求随机变量分布情况已知的前提下进行度量的，而现实中的金融市场常常受到各种复杂因素的影响，尤其我国目前证券市场发展不完善，金融市场波动较大，随机变量分布信息无法完全知道，CVaR风险度量模型度量效率较低。随后，Zhu-FuKushima率先提出了最坏情境下的条件在险价值，简称WCVaR（Worst-case CVaR），它刻画了非完全信息下的风险，在现实中，我们无法预知某件事情的结果时，常常会考虑最坏情况发生时的情况，从而更好预知风险。 本文考虑现实中资产收益率服从混合分布下的WCVaR模型，并在模型中加入比例交易费用函数，使得加入交易费用后的模型研究更贴近现实。然后利用向量自回归构建收益率未来路径，再根据上述回归后残差分布，判别残差可能服从哪几种概率分布情况，结合蒙特卡罗方法随机生成未来资产收益率情景。考虑损失函数为线性的情况下，从而将不确定的线性规划问题转化为确定的线性规划问题，利用Matlab中LP模块，即可求出模型最优解。模型结果证明，加入交易费用后，同等情况下风险相应有一定幅度增加，说明交易费用加入会相应增加风险，对现实中人们投资有一定指导性意义。 Risk measurement model of CVaR /(Conditional value-at-risk and Conditional value at risk/) hasmany advantages, it makes up for the model of VaR /(value-at-risk, value at risk/), and also meets therequirements of Coherent Risk Measurement. Therefore, it is popular among the risk managers. But inthe process of calculation, we must know the distribution of the random variable in advance. However,we can’t know the whole information of the random variable, because of the financial markets is ofteninfluenced by a variety of complicated factors in the reality, especially in our country, the securitiesmarket is imperfect. So the financial market volatile very much. Under that circumstance, the efficiencyof CVaR risk measurement model is low. As for the defects, Zhu-FuKushima put forward theConditional value at risk under the worst-case scenario, here referred to WCVaR /(Worst-case CVaR/),it depicts the risk under incomplete information, in reality, we can't predict the result of something, sowe often consider what the Worst case happens, so as to predict risk better. This paper study the return-on-assets obey the mixed distribution of WCVaR model, and add theproportional transaction costs function to the model,which makes the model much more closer toreality. Then use VAR model yields the return-on-assets future path, so we can get the distribution ofthe residuals after regression, according to the distribution of residuals, and then combined with theMonte Carlo method to generate the future return-on-assets. In this paper, we consider the lossfunction is the linear, in such condition, we can transformed the uncertainty linear programmingproblem into a certain linear programming problem. We can use LP module to solve the problem inMATLAB.The results show that the model of WCVaR with transaction costs is better than the modelwithout that. In short, the new model in my paper have a certain guiding significance for investors inthe real word.

关 键 词： 最坏情况下的条件风险 混合分布 交易费用 蒙特卡罗

分 类 号： [F224 F832.51]

领　　域： [经济管理] [经济管理]