基于Copula分位数回归原油期货市场套保模型及效率研究  

Research on Hedging Model and Efficiency Based on Copula Quantile Regression in Crude Oil Futures Market

作　　者：任仙玲[1] 邓磊 REN Xian-ling;DENG Lei(School of Economics,Ocean University of China,Qingdao 266100,China)

机构地区：[1]中国海洋大学经济学院,山东青岛266100

出　　处：《数理统计与管理》2020年第4期746-760,共15页Journal of Applied Statistics and Management

基　　金：国家自然科学基金资助项目(71671056)。

摘　　要：伴随中国原油期货的上市,作为商品期货最大交易单品的原油期货,其套期保值功能必将成为新的研究热点。本文采用skew-t-GARCH(1,1)模型捕捉原油期现货收益率的"波动集聚"和"尖峰厚尾"特性,在此基础上通过构造Copula分位数套保模型研究不同原油市场状态下(牛市、熊市)的套保比率及效率。利用蒙特卡洛模拟对线性、Normal Copula及T-Copula分位数回归模型进行效率比较,并对英国Brent和美国WTI原油期货收益率进行实证研究,结果表明:①不同市场状态下,原油期货的最优套保比率具有非对称性;②T-Copula分位数回归模型的尾部套期保值效率更稳定。因此,利用原油期货进行规避风险时,要根据市场行情合理运用套期保值模型。With the listing of crude oil futures in China,the hedging function of crude oil futures,which is the largest trading unit of commodity futures,will become a new research hotspot.In this paper,the skew-t-GARCH(1,1)model is used to capture the"volatility agglomeration"and"peak and thick tail"of the spot yield during the crude oil period.On this basis,the Copula quantile hedging model is constructed to study the different crude oil market conditions(Bull market,Bear market)hedging ratio and efficiency.Monte Carlo simulation is used to compare the efficiency of linear,Normal Copula and T-Copula quartile regression models.And the futures yield of crude oil in British Brent and America WIT markers conducts empirical studies.The results show that:1)the optimal hedging ratio of crude oil futures is asymmetric under different market conditions;2)the tail hedging efficiency of the T-Copula quantile regression model is more stable.Therefore,the hedging model should be used reasonably according to the market conditions using crude oil futures to avoid risks.

关 键 词：原油期货 套期保值 线性分位数回归 Copula分位数回归 GARCH模型 

分 类 号：O212[理学—概率论与数理统计]