Risk Dependence Structure, Contagion Network and Spillover Effect among Commercial Banks

doi:10.12005/orms.2022.0059

Abstract

Abstract: To reveal the risk spillover effect among commercial banks in China more scientifically, the research idea of ‘dependent structure-contagion network-risk measurement' is proposed, and the risk spillover effect of 14 commercial banks in China from 2008 to 2018 is systematically analyzed by using the bayesian network and the R-Vine-Copula-CAViaR-CoVaR model. The empirical study shows that the risk-dependent structure of banks has the characteristics of aggregation and distribution of state-owned commercial banks and joint-stock commercial banks respectively, and urban commercial banks are distributed around joint-stock commercial banks, among which joint-stock commercial banks play a pivotal role in connection. It is found that the ability of managing, resisting and spreading risks of state-owned commercial banks is greater than that of joint-stock commercial banks. There is a two-way spillover effect between banks and asymmetry, in which the size of risk spillover of joint-stock banks is larger than that of urban commercial banks than that of state-owned commercial banks, and there is a negative spillover effect between banks.

Key words: dependent structure, Bayesian network, CoVaR, R-Vine, CAViaR

摘要： 为更加科学地揭示我国商业银行间的风险溢出效应,提出“相依结构-传染网络-风险测度”的研究思路,并使用贝叶斯网络与R藤Copula-CAViaR-CoVaR模型对我国14家商业银行在2008年至2018年区间内的风险溢出效应进行了系统分析。实证研究表明:银行风险相依结构具有国有商业银行、股份制商业银行各自聚集,城市商业银行分布于股份制商业银行周边的经济性质聚集分布特征,其中股份制商业银行起到了枢纽连接作用;国有商业银行管理、抵御及分散风险的能力大于股份制商业银行大于城市商业银行;银行间存在双向溢出效应且呈现非对称性,其中股份制银行的风险溢出大小大于城市商业银行大于国有商业银行,且银行之间存在负向溢出效应。

关键词: 相依结构, 贝叶斯网络, CoVaR, R-Vine, CAViaR

CLC Number:

F832.33

LIU Chao, QIAN Cun. Risk Dependence Structure, Contagion Network and Spillover Effect among Commercial Banks[J]. Operations Research and Management Science, 2022, 31(2): 166-172.

刘超, 钱存. 商业银行间风险相依结构、传染网络与溢出效应[J]. 运筹与管理, 2022, 31(2): 166-172.

References

[1] 陈守东,王妍.我国商业银行的系统性金融风险评估——基于极端分位数回归技术的风险度量[J].中国管理科学,2014,22(7):10-17.
[2] 陈建青,王擎,许韶辉.金融行业间的系统性金融风险溢出效应研究[J].数量经济技术经济研究,2015,32(9):89-100.
[3] 王国刚.新常态下的金融风险防范机制[J].金融研究,2015(2):16-22.
[4] 刘超,马玉洁.货币政策多目标交互行为协调、优化与控制[M].北京:科学出版社,2016
[5] 杨子晖,李东承.我国银行系统性金融风险研究——基于“去一法”的应用分析[J].经济研究,2018,53(8):36-51.
[6] 肖璞,刘轶,杨苏梅.相互关联性、风险溢出与系统重要性银行识别[J].金融研究,2012(12):96-106.
[7] 欧阳资生,莫廷程.基于广义CoVaR模型的系统重要性银行的风险溢出效应研究[J].统计研究,2017,34(9):36-43.
[8] 陆静,胡晓红.基于条件在险价值法的商业银行系统性风险研究[J].中国软科学,2014(4):25-42.
[9] 李丛文,闫世军.我国影子银行对商业银行的风险溢出效应——基于GARCH-时变Copula-CoVaR模型的分析[J].国际金融研究,2015(10):64-75.
[10] 方先明,权威.影子银行规模变动的金融资产价格效应[J].经济理论与经济管理,2018(2):39-50.
[11] Pozsar Z, Adrian T, Ashcraft A B, et al. Shadow banking[J]. Staff Reports, 2010. 458(7):447-457.
[12] 陈建青,王擎,许韶辉.金融行业间的系统性金融风险溢出效应研究[J].数量经济技术经济研究,2015,32(9):89-100.
[13] 李政,梁琪,方意.中国金融部门间系统性风险溢出的监测预警研究——基于下行和上行ΔCoES指标的实现与优化[J].金融研究,2019,464(2):40-58.
[14] 刘璐,韩浩.我国保险市场与银行市场间的风险溢出效应研究——基于上市银行和保险公司的实证分析[J].保险研究,2016(12):3-14.
[15] 刘向丽,顾舒婷.房地产对金融体系风险溢出效应研究——基于AR-GARCH-CoVaR方法[J].系统工程理论与实践,2014,34(S1):106-111.
[16] 杨继平,袁璐,张春会.基于结构转换非参数GARCH模型的VaR估计[J].管理科学学报,2014,17(02):69-80.
[17] 许启发,张金秀,蒋翠侠.基于非线性分位数回归模型的多期VaR风险测度[J].中国管理科学,2015,23(03):56-65.
[18] Karmakar M, Paul S. Intraday portfolio risk management using VaR and CVaR: A CGARCH-EVT-Copula approach[J]. International Journal of Forecasting, 2019, 35(2): 699-709.
[19] Adrian T, Brunnermeier M K. CoVaR[J]. Admerican Economic Review, 2016, 106(7): 1705-1741.
[20] Girardi G, Tolga Ergün, A. Systemic risk measurement: Multivariate GARCH estimation of CoVaR[J]. Journal of Banking & Finance, 2013, 37(8): 3169-3180.
[21] Mainik G, Shaanning E. Ondependence consistency of CoVaR and some other sysmatic risk measures[J]. Statistics & Finance, 2013, 37(8): 3169-3180
[22] 吴恒煜,赵平,严武,王辉,吕江林.运用Student T-Copula的极值理论度量我国商业银行的操作风险[J].运筹与管理,2011,20(1):157-163.
[23] 周爱民,韩菲.股票市场和外汇市场间风险溢出效应研究——基于GARCH-时变Copula-CoVaR模型的分析[J].国际金融研究,2017(11):54-64.
[24] Bedford T, Cooke R M. Probability density decomposition for conditionally dependent random variables modeled by vines[J]. Annals of Mathematics and Artificial Intelligence, 2001, 32(1): 245-268.
[25] Kurowicka D, Cooke R M. Distribution-free continuous bayesian belief nets[J]. Quality and Reliability Engineering Statistics, 2005(2): 309-322.
[26] 林宇,李福兴,陈粘,汪巍.基于R-vine-copula-CoVaR模型的金融市场风险溢出效应研究[J].运筹与管理,2017,26(9):148-156.
[27] Riadh A, Mohamed S B A. Relationship between oil, stock prices and exchange rates: a vine copula based GARCH method[J]. The North American Journal of Economics and Finance, 2016, 37(7): 458-471
[28] Karimalis E N , Nomikos N K . Measuring systemic risk in the european banking sector: a copula CoVaR approach[J]. European Journal of Finance, 2017, 24(1): 1-38.
[29] 陈国进,钟灵,张宇.我国银行体系的系统性关联度分析:基于不对称CoVaR[J].系统工程理论与实践,2017,37(1):61-79.
[30] Manganelli E S . CAViaR: conditional autoregressive value at risk by regression quantiles[J]. Journal of Business & Economic Statistics, 2004, 22(4): 367-381.
[31] 王新宇,宋学锋.间接TARCH-CAViaR模型及其MCMC参数估计与应用[J].系统工程理论与实践,2008(9):46-51.
[32] Laporta A G, Luca M, Lea P. Selection of value at risk models for energy commodities[J]. Energy Economics, 2018, 74(8): 628-643.
[33] 曾裕峰,张晗.基于CAViaR和GARCH模型的沪深300股指期货动态风险测度[J].系统工程,2017,35(3):29-35.
[34] 许启发,王侠英,蒋翠侠,熊熊.基于藤Copula-CAViaR方法的股市风险溢出效应研究[J].系统工程理论与实践,2018,38(11):2738-2749.