极值理论与VaR计算

Extreme Value Theory and VaR Computation

介绍了利用极值理论研究随机变量尾部性质的方法,并将其应用于金融产品在险价值(VaR)的计算.实证分析表明,与传统的计算VaR的方法相比,极值理论方法能更好地利用已知历史数据,并能在计算高置信度VaR时克服传统方法中误差较大的缺点.

The extreme value approach is introduced for analyzing the tail performances of random variables. Then it is used to estimate value at risk （VaR） of financial products. Demonstrative analysis shows that the extreme value method has a higher efficiency in using few historical data than the classical method to estimate the VaR.