论文部分内容阅读
由于风险价值、条件风险价值等下方风险度量没有考虑尾部数据的变异性,因此在刻画极端金融风险方面存在一定的缺陷。为了更好地控制尾部极端损失的发生概率,我们选择用尾部条件方差来刻画这种极端风险,即超过风险价值的那部分损失的方差。考虑到混合椭球分布在金融数据建模中的重要性,本文在这类分布下研究了证券组合的尾部条件方差,得到了证券组合尾部条件方差风险的精确表达式,为了验证本文的结果,我们也进行了一些数值计算及在最优投资组合方面的应用研究。
Since the underlying risk measures such as risk value and conditional VaR do not consider the variability of tail data, there are some shortcomings in characterizing extreme financial risks. In order to better control the probability of occurrence of a tail extreme loss, we choose to use the tail conditional variance to characterize this extreme risk, that is, the variance of the portion of the loss that exceeds the VaR. Considering the importance of mixed ellipsoidal distribution in the modeling of financial data, this paper studies the variance of the tail of the portfolio under such distribution and obtains the exact expression of the variance risk of the tail of the portfolio. To verify the result of this paper, We also conducted some numerical calculations and applied research on the optimal portfolio.