外汇风险的极值相关模型  

Extremal Dependence Model of Foreign Exchange Risk

作　　者：李胜朋[1] 王洪礼[1] 李栋[2] 

机构地区：[1]天津大学机械工程学院,天津300072 [2]天津大学管理学院,天津300072

出　　处：《系统工程理论与实践》2007年第9期82-86,共5页Systems Engineering-Theory & Practice

基　　金：博士学科点专项科研基金(20040056041)

摘　　要：研究在已知某一分量极值出现的情况下,通过多维随机向量的联合条件分布的近似分布,利用半参数方法,给出一个极值相关模型,估计多维随机向量的任意尾部事件概率.应用该方法估计加元和日元资产组合一天的99.9%,99.99%,99.999%VaR值分别为0.0321,0.0439,0.0598.Joint tail distribution of financial assets portfolios is necessarily considered in market risk analysis. Existing approaches can only estimate the probability of tail event whose components are all very large. A semi-parametric approach was developed. It was motivated by the asymptotic form of the joint conditional distribution conditioning on it having an extreme component. It can distinguish different extremal dependence and estimate any joint tail probability of random mnltivariable. By our conditional approach, the 99.9% ,99.99% and 99.999% VaR of jen and canadian dollar portfolio for one day is 0.0321,0.0439,0.0598.

关 键 词：渐近独立 条件分布 高斯估计 多元极值理论 半参数模型 VaR(Value at Risk) 

分 类 号：F830.91[经济管理—金融学]