基于MSGARCH-Mixture Copula模型的离、在岸人民币利率风险溢出研究  

A Study on Risk Spillover Effects between Offshore and Onshore RMB Interest Rates Based on MSGARCH-Mixture Copula Model

作　　者：李强 何妃婷 董耀武[3] 王扬 LI Qiang;HE Fei-ting;DONG Yao-wu;WANG Yang(School of Big Data Applications and Economics,Guizhou University of Finance and Economics,Guiyang 550025,China;School of Finance,Tongling University,Tongling 244061,China;School of Finance,Guizhou University of Commerce,Guiyang 550014,China;School of Computer Science and Technology,Guizhou University,Guiyang 550025,China)

机构地区：[1]贵州财经大学大数据应用与经济学院,贵州贵阳550025 [2]铜陵学院金融学院,安徽铜陵244061 [3]贵州商学院金融学院,贵州贵阳550014 [4]贵州大学计算机科学与技术学院,贵州贵阳550025

出　　处：《统计与信息论坛》2022年第6期75-86,共12页Journal of Statistics and Information

基　　金：国家社会科学基金一般项目“基于混合Copula的中国系统性金融风险测度及稳定研究”(18XTJ004)。

摘　　要：人民币国际化进程不断推进,沪港通等相关举措的实施对提高中国金融市场开放程度和完善离、在岸利率传导机制等做出了贡献,因此,探究两岸人民币利率市场间的风险溢出效应对促进利率市场化和维护金融稳定具有重要意义。以沪港通正式启动为节点,将研究样本划分为资本单向开放和资本双向开放两个阶段,运用单(双)状态结合的马尔科夫状态转换GARCH(MSGARCH)模型对收益率序列进行边缘拟合,在此基础上构建混合Copula模型分析离、在岸人民币同业拆借利率之间的相关性,最后通过CoES和CoVaR实现两者之间风险溢出的有效测度。实证结果表明:人民币在岸与离岸利率间整体呈正相关性,其中对于短期期限品种两者相关性较弱,且其在资本双向开放阶段相关性小幅增强,中长期期限的利率与之相反;各利率对的ΔCoVaR绝对值均低于ΔCoES绝对值,传统的CoVaR模型相较于CoES模型可能会低估各利率对之间的风险溢出;分析%CoES和%CoVaR发现,各期限的Shibor与CNH Hibor间在两个资本开放阶段均具有双向的风险溢出效应,对于两岸短期利率市场间的风险溢出而言,在岸较离岸在资本单向开放时期具有明显优势;资本的双向开放在一定程度上削弱了在岸人民币利率对离岸人民币利率的风险溢出,离岸利率对在岸利率的风险溢出在资本双向开放阶段逐步显现,且对于大部分利率期限品种而言,离岸对在岸的风险溢出强度远大于二者反向的风险溢出。因此,文末就如何推动两岸市场双向发展和防范系统性风险溢出提出相关政策建议。Regarding the official launch of the Shanghai-Hong Kong Stock Connect on November 17,2014 as a node, the selected samples are divided into two stages of capital unidirectional opening and capital bidirectional opening.The single(double) state-combined Markov switching GARCH(MSGARCH) model is used to edge the yield sequence, and on the basis, a Mixture Copula is constructed to reveal the correlation between offshore and onshore RMB interbank offered rates.Finally, the effective measurement of risk overflow between the two is realized by CoES and CoVaR.The empirical results are as follows: There is an overall positive correlation between the onshore and offshore interest rates of RMB,among which the correlation is weak for short-term maturity varieties, and the correlation is slightly enhanced in the two-way capital opening phase, whereas for the interest rate of medium and long-term maturity the situation is the opposite.The absolute value of ΔCoVaR for each interest rate pair is lower than the absolute value of ΔCoES,and the traditional CoVaR model may underestimate the risk spillover between interest rate pairs compared to the CoES model.Analyzing from %CoES and %CoVaR,there is a two-way risk spillover effect between Shibor and CNH Hibor for each maturity in both capital opening phases, and for the risk spillover between the short-term interest rate markets on both sides of the Taiwan Strait, onshore has a significant advantage over offshore in the period of one-way capital opening.The two-way opening of capital has to some extent weakened the risk spillover of onshore RMB interest rate to offshore RMB interest rate.The risk spillover of offshore interest rate to onshore interest rate gradually emerges during the two-way capital opening phase, and the intensity of the offshore to onshore risk spillover is much stronger than the reverse risk spillover of the two for most interest rate maturity varieties.Consequently, relevant policy suggestions are put forward on how to promote the two-way development of cross-

关 键 词：MSGARCH-Mixture Copula模型 CoES CoVaR 利率 资本开放 

分 类 号：F726[经济管理—产业经济]