摘要
市场风险值(VaR)是一种常用的度量风险的方法.本文采用极值理论中的阈值模型来计算VaR.基于中国上证指数和深成指数的收盘价,构造超越阈值的极值渐近概率分布,所得到的计算结果与传统方法相比较,有明显的优越性和更好的精度.
Value-at-risk is a commonly used tool to measure risk. In this paper, the POT(peak over threshold) model in extreme value theory is applied to measure VaR. The extreme asymptotic distribution based on excess sample is fitted by using the daily closings of the shanghai shock index and shenzhen development index. After compared the calculation of POT model with that of the classical methodologies, it shows that this approach is more advantaged and accurate than others.
出处
《运筹学学报》
CSCD
北大核心
2005年第1期89-96,共8页
Operations Research Transactions
基金
国家自然科学基金(70171059)资助.