极端金融风险的有效测度与非线性传染
|
摘要
本文采用预期损失指标(expected shortfall, ES)来衡量中国金融市场及各金融部门的极端风险,并结合回溯测试方法进行后验分析,发现ES指标能够对极端风险进行有效测度。在此基础上,本文基于非线性的研究视角,进一步考察了各部门间极端风险的非线性特征与金融风险的跨部门传染效应,并应用相关的网络关联指标,对金融系统整体以及单个金融机构的极端风险的非线性关联展开分析,研究发现房地产等部门是中国金融风险的重要来源。此外,本文还从动态分析的角度考察金融风险跨部门传染的渐进演变。最后,进一步引入中国经济政策不确定性指数及其细化指标,并结合最新发展的混频因果检验等方法,深入考察政策不确定性与极端金融风险的联动效应,研究发现中国股市整体金融风险与经济政策不确定性之间存在双向因果关系。在此基础上,对完善金融风险防范体系及其监管机制提出了若干建议,从而使得本研究具有重要的学术价值与现实意义。
Preventing the cross-sectoral transmission of systemic financial risks is a strategic challenge with implications for China's overall economic and social development. It is thus important to measure extreme risk accurately and examine the impacts of extreme across-markets risk and China's economic policy uncertainty. Doing so will help us not only curb risk contagion, but also help develop a systemic financial risk measurement system, improve the regulatory framework's double pillars of "monetary policy and macro-prudential policy", and provide an appropriate reference for theoretically and empirically analyzing financial system supervision in China.The literature on systemic risk seldom discusses the contagion of extreme risks across financial sectors. As a result, the risk spillover effect in the financial system may not be measured correctly(Hautsch et al., 2014). Research on cross-sectoral risk contagion tends to focus on the effects within traditional financial sectors rather than the contagion between these sectors and the real estate sector. Moreover, studies often use the traditional VaR index to measure the tail risk of an individual institution. However, the expected shortfall(ES) index, which is more sensitive to tail risk, better describes extreme risks(Kratz et al., 2018). Moreover, existing studies pay little attention to the co-movement between the economic policy uncertainty(EPU) index, along with its policy-specific indices, and extreme risks in China's financial market. Finally, most methods commonly used are based on a linear framework(Brana et al., 2018), but ignoring non-linear traits may tend to bias conclusions.(De Vita et al., 2018).In this study, we use the ES index to measure the extreme risks of the A-share stock market and five financial sectors in China: banks, securities, insurance, diversified financials, and real estate. In addition, we apply the newly developed backtesting method(Du & Escanciano, 2016). Then, from a non-linear perspective, this paper examines the risk features of each sector and the contagion effects among them. Moreover, we use network connectivity measures to more accurately describe the transmission of tail risk. Based on a dynamic view, the rolling estimation method is also used to investigate the gradual evolution of cross-sectoral risk contagion. Finally, we introduce China's EPU index(Baker et al., 2016) and its policy-specific indices(Huang & Luk, 2019) and analyze the co-movement between uncertainty and extreme financial risk by using the nonlinear and mixed-frequency causality test(Ghysels et al., 2016). The main conclusions of this paper are as follows.(1) The ES index serves as a precise tail risk indicator. The stock market crash in June 2015 and subsequent use of the circuit breaker in January 2016 exposed sectors such as real estate to significant risks, turning these sectors into latent dangers.(2) There is a significant non-linear risk transmission mechanism between different financial sectors. The network connectivity measures show a continuous and stable spread of extreme risks among different sectors and the entire financial system, while the banking and real estate sectors are the main sources of systemic financial risks. In addition, based on dynamic analysis, we show that after the real estate sector came under intensive regulation in 2016, it continued to produce varying degrees of risk contagion to the banking, insurance, and securities industries in the short term. The results also show that extreme risk events sharply increase the banking industry's risk contagion to the securities and insurance sectors.(3) Finally, there is mutual causality between extreme risk in China's stock market and economic policy uncertainty, suggesting that the Chinese stock market is a typically policy-oriented one where unsuitable policies may spawn extreme risks.The findings of this study have three policy implications.(1) To construct an appropriate measure of systemic risk in China, we need to consider the ES index as an important reference indicator.(2) In regulating the real estate sector, the government should better manage expectations to avoid the excessive volatility associated with inconsistent policies.(3) To avoid the negative effects of bidirectional infection between economic policy uncertainty and extreme risks in China's stock market, economic policies should be consistent.
Preventing the cross-sectoral transmission of systemic financial risks is a strategic challenge with implications for China's overall economic and social development. It is thus important to measure extreme risk accurately and examine the impacts of extreme across-markets risk and China's economic policy uncertainty. Doing so will help us not only curb risk contagion, but also help develop a systemic financial risk measurement system, improve the regulatory framework's double pillars of "monetary policy and macro-prudential policy", and provide an appropriate reference for theoretically and empirically analyzing financial system supervision in China.The literature on systemic risk seldom discusses the contagion of extreme risks across financial sectors. As a result, the risk spillover effect in the financial system may not be measured correctly(Hautsch et al., 2014). Research on cross-sectoral risk contagion tends to focus on the effects within traditional financial sectors rather than the contagion between these sectors and the real estate sector. Moreover, studies often use the traditional VaR index to measure the tail risk of an individual institution. However, the expected shortfall(ES) index, which is more sensitive to tail risk, better describes extreme risks(Kratz et al., 2018). Moreover, existing studies pay little attention to the co-movement between the economic policy uncertainty(EPU) index, along with its policy-specific indices, and extreme risks in China's financial market. Finally, most methods commonly used are based on a linear framework(Brana et al., 2018), but ignoring non-linear traits may tend to bias conclusions.(De Vita et al., 2018).In this study, we use the ES index to measure the extreme risks of the A-share stock market and five financial sectors in China: banks, securities, insurance, diversified financials, and real estate. In addition, we apply the newly developed backtesting method(Du & Escanciano, 2016). Then, from a non-linear perspective, this paper examines the risk features of each sector and the contagion effects among them. Moreover, we use network connectivity measures to more accurately describe the transmission of tail risk. Based on a dynamic view, the rolling estimation method is also used to investigate the gradual evolution of cross-sectoral risk contagion. Finally, we introduce China's EPU index(Baker et al., 2016) and its policy-specific indices(Huang & Luk, 2019) and analyze the co-movement between uncertainty and extreme financial risk by using the nonlinear and mixed-frequency causality test(Ghysels et al., 2016). The main conclusions of this paper are as follows.(1) The ES index serves as a precise tail risk indicator. The stock market crash in June 2015 and subsequent use of the circuit breaker in January 2016 exposed sectors such as real estate to significant risks, turning these sectors into latent dangers.(2) There is a significant non-linear risk transmission mechanism between different financial sectors. The network connectivity measures show a continuous and stable spread of extreme risks among different sectors and the entire financial system, while the banking and real estate sectors are the main sources of systemic financial risks. In addition, based on dynamic analysis, we show that after the real estate sector came under intensive regulation in 2016, it continued to produce varying degrees of risk contagion to the banking, insurance, and securities industries in the short term. The results also show that extreme risk events sharply increase the banking industry's risk contagion to the securities and insurance sectors.(3) Finally, there is mutual causality between extreme risk in China's stock market and economic policy uncertainty, suggesting that the Chinese stock market is a typically policy-oriented one where unsuitable policies may spawn extreme risks.The findings of this study have three policy implications.(1) To construct an appropriate measure of systemic risk in China, we need to consider the ES index as an important reference indicator.(2) In regulating the real estate sector, the government should better manage expectations to avoid the excessive volatility associated with inconsistent policies.(3) To avoid the negative effects of bidirectional infection between economic policy uncertainty and extreme risks in China's stock market, economic policies should be consistent.
引文
方意,2016:《系统性风险的传染渠道与度量研究——兼论宏观审慎政策实施》,《管理世界》第8期。
郭晔、赵静,2017:《存款竞争、影子银行与银行系统风险——基于中国上市银行微观数据的实证研究》,《金融研究》第6期。
梁琪、李政、郝项超,2015:《中国股票市场国际化研究:基于信息溢出的视角》,《经济研究》第4期。
杨子晖、周颖刚,2018:《全球系统性金融风险溢出与外部冲击》,《中国社会科学》第12期。
Acerbi,C.,and D.Tasche,2002,“On the Coherence of Expected Shortfall”,Journal of Banking & Finance,26,1487—1503.
Adrian,T.,and M.K.Brunnermeier,2016,“CoVaR”,American Economic Review,106,1705—1741.
Atik,J.,2010,“Basel II and Extreme Risk Analysis”,Restoring Financial Stability:The Legal Response,6.Available at SSRN:https://ssrn.com/abstract=1677682.
Bai,J.,and P.Perron,2003,“Critical Values for Multiple Structural Change Tests”,Econometrics Journal,6,72—78.
Baker,S.R.,N.Bloom,and S.J.Davis,2016,“Measuring Economic Policy Uncertainty”,Quarterly Journal of Economics,131,1593—1636.
Basel Committee on Banking Supervision,2010,“Basel III:A Global Regulatory Framework for More Resilient Banks and Banking Systems”,Bank for International Settlements.
Benati,L.,2013,“Economic Policy Uncertainty and the Great Recession”,Working paper,University of Bern,Mimeo.
Berkowitz,J.,P.Christoffersen,and D.Pelletier,2011,“Evaluating Value-at-risk Models with Desk-level Data”,Management Science,57,2213—2227.
Brana,S.,A.Campmas,and I.Lapteacru,2018,“(Un) Conventional Monetary Policy and Bank Risk-taking:A Nonlinear Relationship ”,Economic Modelling,In Press.
Brock,W.A.,W.D.Dechert,J.A.Scheinkman,and B.LeBaron,1996,“A Test for Independence Based on the Correlation Dimension”,Econometric Reviews,15,197—235.
Christoffersen,P.F.,1998,“Evaluating Interval Forecasts”,International Economic Review,39,841—862
Cont,R.,R.Deguest,and G.Scandolo,2010,“Robustness and Sensitivity Analysis of Risk Measurement Procedures”,Quantitative Finance,10,593—606.
Corsi,F.,F.Lillo,D.Pirino,and L.Trapin,2018,“Measuring the Propagation of Financial Distress with Granger-causality Tail Risk Networks”,Journal of Financial Stability,38,18—36.
Costanzino,N.,and M.Curran,2015,“Backtesting General Spectral Risk Measures with Application to Expected Shortfall”,Journal of Risk Model Validation,9,21—31.
Costanzino,N.,and M.Curran,2018,“A Simple Traffic Light Approach to Backtesting Expected Shortfall”,Risks,6,2—8.
Cotter,J.,M.Hallam,and K.Yilmaz,2017,“Mixed-frequency Macro-financial Spillovers”,Working Paper,KO? University,Turkey.
De Vita,G.,E.Trachanas,and Y.Luo,2018,“Revisiting the Bi-directional Causality between Debt and Growth:Evidence from Linear and Nonlinear Tests”,Journal of International Money & Finance,83,55—74.
Diebold,F.X.,and K.Yilmaz,2014,“On the Network Topology of Variance Decompositions:Measuring the Connectedness of Financial Firms”,Journal of Econometrics,182,119—134.
Diks,C.,and V.Panchenko,2006,“A New Statistic and Practical Guidelines for Nonparametric Granger Causality Testing”,Journal of Economic Dynamics & Control,30,1647—1669.
Du,Z.,and J.C.Escanciano,2016,“Backtesting Expected Shortfall:Accounting for Tail Risk”,Management Science,63,940—958.
Ghulam,Y.,and J.Doering,2018,“Spillover Effects among Financial Institutions within Germany and the United Kingdom”,Research in International Business & Finance,44,49—63.
Ghysels,E.,J.B.Hill,and K.Motegi,2016,“Testing for Granger Causality with Mixed Frequency Data”,Journal of Econometrics,192,207—230.
G?tz,T.B.,A.Hecq,and S.Smeekes,2016,“Testing for Granger Causality in Large Mixed-frequency VARs”,Journal of Econometrics,193,418—432.
H?rdle,W.K.,W.Wang,and L.Yu,2016,“Tenet:Tail-event Driven Network Risk”,Journal of Econometrics,192,499—513.
Hautsch,N.,J.Schaumburg,and M.Schienle,2014,“Financial Network Systemic Risk Contributions”,Review of Finance,19,685—738.
Hiemstra,C.,and J.D.Jones,1994,“Testing for Linear and Nonlinear Granger Causality in the Stock Price‐volume Relation”,Journal of Finance,49,1639—1664.
Huang,Y.,and P.Luk,2019,“Measuring Economic Policy Uncertainty in China”,Working Paper,Hong Kong Baptist University,Hong Kong.
Kratz,M.,Y.H.Lok,and A.J.McNeil,2018,“Multinomial VaR Backtests:A Simple Implicit Approach to Backtesting Expected Shortfall”,Journal of Banking & Finance,88,393—407.
Lee,T.H.,H.White,and C.W.Granger,1993,“Testing for Neglected Nonlinearity in Time Series Models:A Comparison of Neural Network Methods and Alternative Tests”,Journal of Econometrics,56,269—290.
Mcleod,A.I.,and W.K.Li,1981,“Distribution of the Residual Autocorrelations in Multivariate ARMA Time Series Models”,Journal of Times Series Analysis,43,231—239.
Ng,S.,and P.Perron,2001,“Lag Length Selection and the Construction of Unit Root Tests with Good Size and Power”,Econometrica,69,1519—1554.
Ramsey,J.B.,1969,“Tests for Specification Errors in Classical Linear Least-squares Regression Analysis”,Journal of the Royal Statistical Society,31,350—371.
Schwert,G.W.,1989,“Why Does Stock Market Volatility Change over Time?”,Journal of Finance,44,1115—1153.
Wang,G.J.,C.Xie,K.He,and H.E.Stanley,2017,“Extreme Risk Spillover Network:Application to Financial Institutions”,Quantitative Finance,17,1—23.
(1)由于篇幅限制,未报告回溯测试的检验结果,以备索的方式提供。
(2)本文采用Ng & Perron(2001)基于GLS退势原理构造的单位根检验,以及协整迹检验方法(Johansen,1991)对各部门ES测度指标进行检验,结果表明它们存在着显著协整关系。为节省篇幅,本文未报告相关的检验结果,有兴趣的读者可向作者索取。
(3)TVAL非参检验方法以及非参数的Tn检验方法的相关细节参见Hiemstra & Jones(1994)以及Diks & Panchenko(2006)。
(4)使用GARCH(1,1)模型对过滤后的残差序列再次过滤进行非线性检验,得到了一致的基本结论。本文基于非线性Granger因果检验的结论是稳健和可靠的。感兴趣的读者可以向作者索取。
(5)基于动态分析方法的基本原理(Diebold & Yilmaz,2014),本文以90天作为滚动分析的窗口长度,并采用10%显著性水平的临界值对统计检验量进行正则化,因此大于1的统计检验值表明在该时点拒绝原假设。
(6)为了节省篇幅,这里没有报告所有部门的动态非线性因果检验结果,以备索的方式提供。
(7)但长期来看,提供流动性的市场调整缓解了同年12月美联储宣布加息给中国资本市场带来的冲击。
(8)根据Cotter et al.(2017),本文将每个月内的交易日划分为4周,并在此基础上构造相应的极端风险周指标。
(9)本文同样基于非线性Granger因果检验方法,对经济政策不确定性的细化指标进行分析,得到了一致的结论,结果备索。
郭晔、赵静,2017:《存款竞争、影子银行与银行系统风险——基于中国上市银行微观数据的实证研究》,《金融研究》第6期。
梁琪、李政、郝项超,2015:《中国股票市场国际化研究:基于信息溢出的视角》,《经济研究》第4期。
杨子晖、周颖刚,2018:《全球系统性金融风险溢出与外部冲击》,《中国社会科学》第12期。
Acerbi,C.,and D.Tasche,2002,“On the Coherence of Expected Shortfall”,Journal of Banking & Finance,26,1487—1503.
Adrian,T.,and M.K.Brunnermeier,2016,“CoVaR”,American Economic Review,106,1705—1741.
Atik,J.,2010,“Basel II and Extreme Risk Analysis”,Restoring Financial Stability:The Legal Response,6.Available at SSRN:https://ssrn.com/abstract=1677682.
Bai,J.,and P.Perron,2003,“Critical Values for Multiple Structural Change Tests”,Econometrics Journal,6,72—78.
Baker,S.R.,N.Bloom,and S.J.Davis,2016,“Measuring Economic Policy Uncertainty”,Quarterly Journal of Economics,131,1593—1636.
Basel Committee on Banking Supervision,2010,“Basel III:A Global Regulatory Framework for More Resilient Banks and Banking Systems”,Bank for International Settlements.
Benati,L.,2013,“Economic Policy Uncertainty and the Great Recession”,Working paper,University of Bern,Mimeo.
Berkowitz,J.,P.Christoffersen,and D.Pelletier,2011,“Evaluating Value-at-risk Models with Desk-level Data”,Management Science,57,2213—2227.
Brana,S.,A.Campmas,and I.Lapteacru,2018,“(Un) Conventional Monetary Policy and Bank Risk-taking:A Nonlinear Relationship ”,Economic Modelling,In Press.
Brock,W.A.,W.D.Dechert,J.A.Scheinkman,and B.LeBaron,1996,“A Test for Independence Based on the Correlation Dimension”,Econometric Reviews,15,197—235.
Christoffersen,P.F.,1998,“Evaluating Interval Forecasts”,International Economic Review,39,841—862
Cont,R.,R.Deguest,and G.Scandolo,2010,“Robustness and Sensitivity Analysis of Risk Measurement Procedures”,Quantitative Finance,10,593—606.
Corsi,F.,F.Lillo,D.Pirino,and L.Trapin,2018,“Measuring the Propagation of Financial Distress with Granger-causality Tail Risk Networks”,Journal of Financial Stability,38,18—36.
Costanzino,N.,and M.Curran,2015,“Backtesting General Spectral Risk Measures with Application to Expected Shortfall”,Journal of Risk Model Validation,9,21—31.
Costanzino,N.,and M.Curran,2018,“A Simple Traffic Light Approach to Backtesting Expected Shortfall”,Risks,6,2—8.
Cotter,J.,M.Hallam,and K.Yilmaz,2017,“Mixed-frequency Macro-financial Spillovers”,Working Paper,KO? University,Turkey.
De Vita,G.,E.Trachanas,and Y.Luo,2018,“Revisiting the Bi-directional Causality between Debt and Growth:Evidence from Linear and Nonlinear Tests”,Journal of International Money & Finance,83,55—74.
Diebold,F.X.,and K.Yilmaz,2014,“On the Network Topology of Variance Decompositions:Measuring the Connectedness of Financial Firms”,Journal of Econometrics,182,119—134.
Diks,C.,and V.Panchenko,2006,“A New Statistic and Practical Guidelines for Nonparametric Granger Causality Testing”,Journal of Economic Dynamics & Control,30,1647—1669.
Du,Z.,and J.C.Escanciano,2016,“Backtesting Expected Shortfall:Accounting for Tail Risk”,Management Science,63,940—958.
Ghulam,Y.,and J.Doering,2018,“Spillover Effects among Financial Institutions within Germany and the United Kingdom”,Research in International Business & Finance,44,49—63.
Ghysels,E.,J.B.Hill,and K.Motegi,2016,“Testing for Granger Causality with Mixed Frequency Data”,Journal of Econometrics,192,207—230.
G?tz,T.B.,A.Hecq,and S.Smeekes,2016,“Testing for Granger Causality in Large Mixed-frequency VARs”,Journal of Econometrics,193,418—432.
H?rdle,W.K.,W.Wang,and L.Yu,2016,“Tenet:Tail-event Driven Network Risk”,Journal of Econometrics,192,499—513.
Hautsch,N.,J.Schaumburg,and M.Schienle,2014,“Financial Network Systemic Risk Contributions”,Review of Finance,19,685—738.
Hiemstra,C.,and J.D.Jones,1994,“Testing for Linear and Nonlinear Granger Causality in the Stock Price‐volume Relation”,Journal of Finance,49,1639—1664.
Huang,Y.,and P.Luk,2019,“Measuring Economic Policy Uncertainty in China”,Working Paper,Hong Kong Baptist University,Hong Kong.
Kratz,M.,Y.H.Lok,and A.J.McNeil,2018,“Multinomial VaR Backtests:A Simple Implicit Approach to Backtesting Expected Shortfall”,Journal of Banking & Finance,88,393—407.
Lee,T.H.,H.White,and C.W.Granger,1993,“Testing for Neglected Nonlinearity in Time Series Models:A Comparison of Neural Network Methods and Alternative Tests”,Journal of Econometrics,56,269—290.
Mcleod,A.I.,and W.K.Li,1981,“Distribution of the Residual Autocorrelations in Multivariate ARMA Time Series Models”,Journal of Times Series Analysis,43,231—239.
Ng,S.,and P.Perron,2001,“Lag Length Selection and the Construction of Unit Root Tests with Good Size and Power”,Econometrica,69,1519—1554.
Ramsey,J.B.,1969,“Tests for Specification Errors in Classical Linear Least-squares Regression Analysis”,Journal of the Royal Statistical Society,31,350—371.
Schwert,G.W.,1989,“Why Does Stock Market Volatility Change over Time?”,Journal of Finance,44,1115—1153.
Wang,G.J.,C.Xie,K.He,and H.E.Stanley,2017,“Extreme Risk Spillover Network:Application to Financial Institutions”,Quantitative Finance,17,1—23.
(1)由于篇幅限制,未报告回溯测试的检验结果,以备索的方式提供。
(2)本文采用Ng & Perron(2001)基于GLS退势原理构造的单位根检验,以及协整迹检验方法(Johansen,1991)对各部门ES测度指标进行检验,结果表明它们存在着显著协整关系。为节省篇幅,本文未报告相关的检验结果,有兴趣的读者可向作者索取。
(3)TVAL非参检验方法以及非参数的Tn检验方法的相关细节参见Hiemstra & Jones(1994)以及Diks & Panchenko(2006)。
(4)使用GARCH(1,1)模型对过滤后的残差序列再次过滤进行非线性检验,得到了一致的基本结论。本文基于非线性Granger因果检验的结论是稳健和可靠的。感兴趣的读者可以向作者索取。
(5)基于动态分析方法的基本原理(Diebold & Yilmaz,2014),本文以90天作为滚动分析的窗口长度,并采用10%显著性水平的临界值对统计检验量进行正则化,因此大于1的统计检验值表明在该时点拒绝原假设。
(6)为了节省篇幅,这里没有报告所有部门的动态非线性因果检验结果,以备索的方式提供。
(7)但长期来看,提供流动性的市场调整缓解了同年12月美联储宣布加息给中国资本市场带来的冲击。
(8)根据Cotter et al.(2017),本文将每个月内的交易日划分为4周,并在此基础上构造相应的极端风险周指标。
(9)本文同样基于非线性Granger因果检验方法,对经济政策不确定性的细化指标进行分析,得到了一致的结论,结果备索。