沪深股市的相关结构分析与投资组合风险度量——基于ARFIMA-GARCH-Copula模型  

Shanghai and Shenzhen Stock Market Related Structure Aanalysis and Portfolio Risk Measurement Based on ARFIMA-GARCH-Copula Model

作　　者：吴玉宝[1] 汪金菊[1] 

机构地区：[1]合肥工业大学数学学院,安徽合肥230009

出　　处：《运筹与管理》2016年第2期220-225,共6页Operations Research and Management Science

基　　金：国家重大科研装备研制项目(ZDYZ2012-1);安徽省自然科学基金资助项目(1208085MF91);中央高校基本科研业务费专项资金项目(J2014HGXJ0072)

摘　　要：金融资产收益率不仅具有尖峰厚尾性、异方差性,还具有长记忆性。基于此,本文建立ARFIMAGARCH-Copula模型来研究沪深股市的相关结构和等权重投资组合风险值VaR,利用上证指数和深成指数收益率的组合来进行实证研究。首先采用经典R/S分析法检验各个资产收益率的长记忆性,经过分数阶差分后选用GARCH模型建模得到边缘分布。然后选择Copula函数来刻画两资产之间的相关结构,建立联合分布模型。进而采用Monte Carlo方法模拟产生各资产的收益率序列,计算出投资组合的风险值VaR。实证研究表明:沪深股市具有长记忆性,且两者具有对称的尾部相关性;Kupiec检验说明ARFIMA-GARCH-Copula模型较之于GARCHCopula模型能更准确地度量投资组合风险。It is well known that financial return has sharp-peaks, fat-tails, heteroscedasticity and long memory. Considering three features, the paper constructs a risk measure model based on the ARFIMA-GARCH-Copula for financial portfolio, which is composed by Shanghai Stock index return and Shenzhen Component Index return equal weight. First the classical R/S analysis is adopted to test the long memory of a single asset. Second, the paper adopts different GARCH models to fit each asset return series. Third, it selects Copula function to describe the relational structure between each asset. Fourth, it uses Monte Carlo method to produce each return sequence of the assets. And then it calculates the VaR of financial portfolio. The empirical results show that there is appar- ent long memory property in the Shanghai and Shenzhen stock market which has symmetrical tail correlation. Kupiec test results show that the model of ARFIMA-GARCH-Copula is more efficient than GARCH-Copula model in measuring the portfolio risk.

关 键 词：ARFIMA GARCH COPULA函数 VaR风险值 Kupiec检验 

分 类 号：F830.9[经济管理—金融学]