基于EVT的谱风险测度及其在风险管理中的应用  

Extremal spectral risk measures and their applications in financial risk management

作　　者：刁训娣[1] 童斌[2] 吴冲锋[3] 

机构地区：[1]上海交通大学电子信息与电气工程学院,上海200240 [2]东华大学旭日工商管理学院,上海200051 [3]上海交通大学安泰经济与管理学院,上海200052

出　　处：《系统工程学报》2015年第3期354-369,405,共17页Journal of Systems Engineering

基　　金：国家自然科学基金资助项目(71201100;71320107002);上海交通大学新进青年教师启动计划资助项目(13X1000-40056)

摘　　要：金融资产收益率的分布往往呈现厚尾特性,忽略此现象往往会造成极值风险的低估.基于极值理论(EVT),引入一类测度尾部损失风险的谱风险测度方法(TSRMs).与传统的在险价值(VaR)和期望尾损(ES)相比,新的谱风险测度(TSRMs)赋予大的尾部损失以大的权重,非尾部损失权重为0,因而更能反映投资者的风险规避心理及其风险偏好程度.最后,以标准普尔500指数的日对数收益率为例,分析比较了TSRMs与VaR,ES在正态分布和极值分布下的估计结果,并进一步解释不同的风险偏好在投资者对风险的心里预期中起的作用.It has been well recognized that the distribution of financial asset returns are fat-tailed. The neglect of the fact can lead to unfavourable extremal tail risk estimates. Based on extreme value theory a new type of tail-related spectral risk measures（TSRMs） is proposed. Traditional risk measures such as value at risk（VaR） and expected shortfall（ES）, do not make any allowance for the investors to include their own risk attitudes. The extremal spectral risk measure proposed in this paper has this useful feature in that it gives higher weights to higher tail losses and a weight of 0 to the other quantiles. The applications of the proposed extremal TSRMs in estimating the tail risk of Standard and Poor 500（S＆P 500） stock index are illustrated, and the impact of the risk attitude of investors on the value of extremal TSRMs is investigated.

关 键 词：谱风险测度 极值风险 GEV分布 GPD分布 一致性风险测度 

分 类 号：F830[经济管理—金融学]