摘要
考虑金融时间序列的厚尾特性,讨论了应用极值理论中的广义Pareto分布模型度量风险的问题。利用Bootstrap和MLE方法对参数进行点估计和区间估计,得出E-VaR的估计值,并对深证综指收益进行实证分析,探讨与尾部相关的极值风险,结果令人满意。
In view that the fat tail of financial time series, we will discuss how to measure risk with GPD （general Pareto distribution） models in extreme value theory. Bootstrap and likelihood-based methods are used to estimate the parameters（point estimate and confidence interval）,and give the result of E-VaR. Besides, Shenzhen Stock Index is analyzed to predict the tail related extreme risk, we can find the result is satisfying.
出处
《价值工程》
2007年第3期102-106,共5页
Value Engineering
