摘要
近年来,中美两国贸易摩擦加剧了金融市场之间的相关关系,期货市场作为金融市场重要的一部分,研究期货市场间的相关关系显得尤为重要。本文将GARCH模型与Copula模型相结合,建立二元金融时间序列的Copula-GARCH模型以研究中美大豆期货市场的相关关系。首先利用GARCH-t模型刻画中美两国大豆期货市场收益率的边缘分布,然后利用极大似然估计法和K-S检验进行参数估计并拟合优度检验,接着利用极大似然法估计常用Copula参数,利用AIC,BIC准则和Log likelihood最小准则选择合适的Copula函数来描述中美大豆期货市场间的相关性。实证研究表明,大连商品期货交易所大豆期货价格与芝加哥商品交易所大豆期货价格有较大相关性,t-Copula函数能较优地描述两市场的相关性,两个市场之间大豆期货的上下尾相关系数皆为0.087 019,存在同涨同跌现象。
In recent years,trade frictions between China and the United States have strengthened the correlation between financial markets around the world,so it is particularly important to study the correlation between them,as futures markets are the important parts of the financial markets.Combining GARCH model with Copula model,the Copula-GARCH model of binary financial time series is established to study the correlation between soybean futures markets.Firstly,GARCH-t model is used to depict the edge distribution of the soybean futures market yields in China and the United States,then the method of maximum likelihood estimation and K-S test are used for parameter estimation and testing the goodness of fit,the maximum likelihood method is used to estimate parameters of copulas,AIC and BIC criterion,the Log likelihood criterion is used to choose the appropriate copula function to describe the correlation between soybean futures market of China and the United States.The empirical study shows that the soybean futures price of Dalian Commodity Futures Exchange has a great correlation with the soybean futures price of Chicago Mercantile Exchange,and the T-Copula function can better describe the correlation between the two markets.The correlation coefficients of the upper and lower end of soybean futures between the two markets are both 0.087019,and there is a phenomenon of both rising and falling at the same time.
作者
高瑞
卢俊香
GAO Rui;LU Junxiang(School of Science,Xi'an Polytechnic University,Xi'an 710000,China)
出处
《四川轻化工大学学报（自然科学版）》
CAS
2021年第2期95-100,共6页
Journal of Sichuan University of Science & Engineering(Natural Science Edition)
基金
国家自然科学基金项目(11601410)
中国博士后科学基金项目(2017M613169)。
