中国——东盟金融市场的结构相依与极值风险:基于“一带一路”的背景
|
摘要
为了研究"一带一路"战略实施前后中国与东盟主要国家股市之间的结构相依特征与极值风险,文章首先基于GJR-GARCH模型过滤收益率序列,然后采用极值理论对边缘分布建模,最后运用regular vine copula理论构建联合分布。同时,文章对比分析"一带一路"战略实施前后股市之间相依程度与极值风险程度的变化。研究发现:(1) GJR-GARCH模型和极值理论可以拟合中国与东盟主要国家股市收益率序列的边缘分布特征,且采用同一种类型的规则藤Copula模型可以刻画股市之间的相依性结构。(2)二元t-copula函数可以刻画中国与东盟主要国家股市之间对称的无条件尾部相依特征,而非对称的条件尾部相依特征可以运用其它Copula函数来度量。(3)"一带一路"战略实施期间,股市之间的相依程度有所下降,市场极值风险有所增加,但仍具有一定的风险规避功能。
Since the establishment of the China-ASEAN Free Trade Area, the relationship of trade and cross-border finance in China and the ASEAN nations has been increasing. In particular, the implementation of "The Belt and Road" strategy further emphasizes the enhancement of economic and financial links among China-ASEAN FTA nations. However, due to the differences of the economic characteristics in these nations and the regulations of capital markets,there are complex non-linear relationships between stock markets of China and the ASEAN nations. These relationships are very useful and significant for measuring the volatility spillover effects and extreme risks among these stock markets.With the deepening of economic and financial integration, financial markets are characterized by more complex dependence and volatility spillover effects,and the transmission of financial risks is gradually strengthened, especially during the financial crisis. The regular vine copula model can capture more complex structural dependence among capital markets. Also, when the extreme events occur, the distributions of return on financial assets are not satisfied with the normal distribution, but the asymmetric features of leptokurtosis and fat-tail. The tail risks among capital markets can be measured based on the extreme value theory. Therefore, this paper constructs the EVT-regular vine copula model by combining the extreme value theory with regular vine copula model and calculates the Value-at-Risk of the portfolio by Monte Carlo simulation method. The proposed model is used to study the dependency structure and spillover effects of extreme risks of the stock markets in the China-ASEAN Free Trade Area during the implementation of "The Belt and Road" strategy. This study provides investors with some reference for investment strategy and risk management in the stock markets of the China-ASEAN Free Trade Area.The paper mainly includes five parts. The first part introduces the background of this paper. The second part is a literature review. The third part is mainly to describe the basic theorem and methodology. The fourth part describes the data and empirical study, and the last part concludes and proposes some future research.This study firstly filters the logarithm returns based on the GJR-GARCH model, then models the marginal distributions and joint distribution by using extreme value theory and regular vine copula theorem respectively. Furthermore, this paper studies the structural dependences among China and the ASEAN nations before the implementation of the strategy of "The Belt and Road" strategy. Based on the comparative study, it gives a clear understanding of the changes of volatility spillover and extreme risk among the stock markets of China and the ASEAN nations before and during the implementation of "The Belt and Road"strategy. The main results are as follows. Firstly, the GJR-GARCH model and extreme value theory can fit the marginal distributions of the logarithm returns of the stock markets in China and the ASEAN nations, and the dependencies can be captured by the same kind of regular vine copula framework before and during the period of "The Belt and Road" strategy. Secondly, there are symmetric unconditional tail dependences among the stock markets of China and the ASEAN nations before and after the implementation of "The Belt and Road" strategy. And the symmetric unconditional tail dependences can be described by the bivariate Student t-copula, while other Copulas measure the asymmetric conditional tail dependences. Thirdly, the dependences decrease during the implementation of "The Belt and Road" strategy, and there are still some barriers to some extent among the stock markets in China and the ASEAN nations.Fourthly, the VaR of the portfolio is larger during the implementation of "The Belt and Road" strategy than that before the implementation of "The Belt and Road" strategy, indicating that the extreme risks increase in the markets, while it still has a certain function of risk aversion among the stock markets in China and the ASEAN nations.
Since the establishment of the China-ASEAN Free Trade Area, the relationship of trade and cross-border finance in China and the ASEAN nations has been increasing. In particular, the implementation of "The Belt and Road" strategy further emphasizes the enhancement of economic and financial links among China-ASEAN FTA nations. However, due to the differences of the economic characteristics in these nations and the regulations of capital markets,there are complex non-linear relationships between stock markets of China and the ASEAN nations. These relationships are very useful and significant for measuring the volatility spillover effects and extreme risks among these stock markets.With the deepening of economic and financial integration, financial markets are characterized by more complex dependence and volatility spillover effects,and the transmission of financial risks is gradually strengthened, especially during the financial crisis. The regular vine copula model can capture more complex structural dependence among capital markets. Also, when the extreme events occur, the distributions of return on financial assets are not satisfied with the normal distribution, but the asymmetric features of leptokurtosis and fat-tail. The tail risks among capital markets can be measured based on the extreme value theory. Therefore, this paper constructs the EVT-regular vine copula model by combining the extreme value theory with regular vine copula model and calculates the Value-at-Risk of the portfolio by Monte Carlo simulation method. The proposed model is used to study the dependency structure and spillover effects of extreme risks of the stock markets in the China-ASEAN Free Trade Area during the implementation of "The Belt and Road" strategy. This study provides investors with some reference for investment strategy and risk management in the stock markets of the China-ASEAN Free Trade Area.The paper mainly includes five parts. The first part introduces the background of this paper. The second part is a literature review. The third part is mainly to describe the basic theorem and methodology. The fourth part describes the data and empirical study, and the last part concludes and proposes some future research.This study firstly filters the logarithm returns based on the GJR-GARCH model, then models the marginal distributions and joint distribution by using extreme value theory and regular vine copula theorem respectively. Furthermore, this paper studies the structural dependences among China and the ASEAN nations before the implementation of the strategy of "The Belt and Road" strategy. Based on the comparative study, it gives a clear understanding of the changes of volatility spillover and extreme risk among the stock markets of China and the ASEAN nations before and during the implementation of "The Belt and Road"strategy. The main results are as follows. Firstly, the GJR-GARCH model and extreme value theory can fit the marginal distributions of the logarithm returns of the stock markets in China and the ASEAN nations, and the dependencies can be captured by the same kind of regular vine copula framework before and during the period of "The Belt and Road" strategy. Secondly, there are symmetric unconditional tail dependences among the stock markets of China and the ASEAN nations before and after the implementation of "The Belt and Road" strategy. And the symmetric unconditional tail dependences can be described by the bivariate Student t-copula, while other Copulas measure the asymmetric conditional tail dependences. Thirdly, the dependences decrease during the implementation of "The Belt and Road" strategy, and there are still some barriers to some extent among the stock markets in China and the ASEAN nations.Fourthly, the VaR of the portfolio is larger during the implementation of "The Belt and Road" strategy than that before the implementation of "The Belt and Road" strategy, indicating that the extreme risks increase in the markets, while it still has a certain function of risk aversion among the stock markets in China and the ASEAN nations.
引文
[1]Morgenstern O.International financial transactions and business cycles[M].Princeton University Press,1959.
[2]Hamao Y,Masulis RW,Ng V.Correlations in price changes and volatility across international stock markets[J].Review of Financial Studies,1990,3(2):281-307.
[3]Lin W,Engle RF,Ito T.Do bulls and bears move across borders?international transmission of stock returns and volatility[J].Review of Financial Studies,1994,7(3):507-538.
[4]赵留彦,王一鸣.A、B股之间的信息流动与波动溢出[J].金融研究,2003,(10):37-52.
[5]洪永淼,成思危,刘艳辉,等.中国股市与世界其他股市之间的大风险溢出效应[J].经济学(季刊),2004,3(3):703-726.
[6]Hartman P,Straetmans S,de Vries CG.Asset market linkages in crisis periods[J].The Review of Economics and Statistics,2004,86(1):313-326.
[7]Longin F,Solnik B.Extreme correlation of international equity markets[J].The Journal of Finance,2001,56(2):649-676.
[8]McNeil AJ,Frey R,Embrechts P.Quantitative risk management:concepts,techniques and tools[M].Princeton University Press,2005.
[9]Longin F.From VaR to stress testing:the extreme value approach[J].Journal of Banking and Finance,2000,24:1097-1130.
[10]魏宇.金融市场的收益分布与EVT风险测度[J].数量经济技术经济研究,2006,23(4):101-110.
[11]林宇,黄登仕,魏宇.胖尾分布及长记忆下的动态EVT-VaR测度研究[J].管理科学学报,2011,14(7):71-82.
[12]Sklar A.Fonctions de répartitionàn Dimensions et leurs marges[J].Publications de l'Institut de Statistique de l'Universitéde Paris,1959,8:229-231.
[13]Wang Z,Chen X,Jin Y,et al.Estimating risk of foreign exchange portfolio:using VaR and CVaR based on GARCH-EVT-Copula model[J].Physica A:Statistical Mechanics and its Applications,2010,389(21):4918-4928.
[14]Berger T.Forecasting value-at-risk using time varying copulas and EVT return distributions[J].International Economics,2013,133:93-106.
[15]刘晓星,邱桂华.基于Copula-EVT模型的我国股票市场流动性调整的VaR和ES研究[J].数理统计与管理,2010,29(1):150-161.
[16]刘晓星,段斌,谢福座.股票市场风险溢出效应研究:基于EVT-Copula-CoVaR模型的分析[J].世界经济,2011,(11):145-159.
[17]于文华,魏宇,康明惠.不同时变Copula-EVT-ES模型精度比较研究[J].管理科学学报,2015,18(5):32-45.
[18]Joe H.Families of m-variate distributions with given margins and m(m-1)/2 bivariate dependence parameters[A].in:L.R?uschendorf,B.Schweitzer,and M.Taylor(Eds.),Distributions with Fixed Marginals and Related Topics[C],Lecture Notes-Monograph Series,1996,28:120-141.
[19]Bedford T,Cooke RM.Probability density decomposition for conditionally dependent random variables modeled by vines[J].Annals of Mathematics and Artificial Intelligence,2001,32(1):245-268.
[20]Bedford T,Cooke RM.Vines-a new graphical model for dependent random variables[J].The Annals of Statistics,2002,30(4):1031-1068.
[21]Kurowicka D,Cooke R.Uncertainty analysis with high dimensional dependence Modelling[M].Wiley,2006.
[22]Kurowicka D,Joe H.Dependence modeling:vine copula handbook[M].World Scientific,2011.
[23]Di?mann J,Brechmann EC,Czado C,et al.Selecting and estimating regular vine copulae and application to financial returns[J].Computational Statistics and Data Analysis,2013,59(1):52-69.
[24]Koliai L.Extreme risk modeling:an EVT-Pair-copulas approach for financial stress tests[J].Journal of Banking and Finance,2016,Forthcoming.
[25]张帮正,魏宇,余江,等.基于EVT-Vine-copula的多市场相关性及投资组合选择研究[J].管理科学,2014,27(3):133-144.
[26]SesaY RB.使用GARCH-EVT和藤式Copula进行极端值依赖性建模和在险价值估计[D].成都:西南财经大学博士学位论文,2014.
[27]苟红军,陈迅,花拥军.基于GARCH-EVT-COPULA模型的外汇投资组合风险度量研究[J].管理工程学报,2015,29(1):183-193.
[28]何娟,王建,蒋祥林,等.基于Copula-CVaR-EVT方法的供应链金融质物组合优化[J].系统工程理论与实践,2015,35(1):1-16.
[29]李丽梅,胡华.基于SVt-EVT-Vine Copula的投资组合Va R研究[J].数学的实践与认识,2016,(9):104-112.
[30]Engle RF.Autoregressive conditional heteroscedasticity with estimates of the variance of United Kingdom inflation[J].Econometrica,1982,50(4):987-1007.
[31]Bollerslev,T.Generalized autoregressive conditional heteroscedasticity[J].Journal of Econometrics,1986,31(3):307-327.
[32]Glosten LR,Jagannathan R,Runkle DE.On the Relation between the expected value and the volatility of the nominal excess return on stocks[J].The Journal of Finance,1993,48(5):1779-1801.
[33]Scarrott C,Mac Donald A.A review of extreme value threshold estimation and uncertainty quantification[J].Statistical Journal,2012,10(1):33-60.
[34]DuMouchel WH.Estimating the table indexαin order to measure tail thickness:a critique[J].Annals of Statistics,1983,11(4):1019-1031.
[35]Rosenberg J,Schuermann T.A general approach to integrated risk management with skewed,fat-tailed risks[Z].Federal Reserve Bank of New York,2004.
(1)《习近平主持召开中央财经领导小组第八次会议》,http://politics.people.com.cn/n/2014/1106/c70731-25989646. html
(2)中共中央宣传部:《习近平总书记系列重要讲话读本》(2016年版),学习出版社、人民出版社,2016年第1版,第266-268页。
(1)图2中,原始指数经正态化处理,以便比较各指数之间的相对变化趋势。
(1)在对数据检验的过程中,文中仅展示中国的情形。如有读者感兴趣,其它股市的检验结果可向作者索取。
(1)此处仅展示中国的情形,如有读者感兴趣,其它股市的检验结果可向作者索取。
(2)对于其它树的结构图,如读者感兴趣,可以向作者索取。
(1)2015年11月22日,中华人民共和国商务部部长高虎城与东盟十国部长在马来西亚吉隆坡正是签署《中华人民共和国与东南亚国家联盟关于修订〈中国—东盟全面经济合作框架协议〉及项下部分协议的议定书》。
(2)尽管所得到的规则藤结构相同,但在不同“树”上,用来描述股票市场之间的条件Copula函数略有差别,且股票市场之间的相依程度不大。如读者感兴趣,可以向作者索取详细结果。
[2]Hamao Y,Masulis RW,Ng V.Correlations in price changes and volatility across international stock markets[J].Review of Financial Studies,1990,3(2):281-307.
[3]Lin W,Engle RF,Ito T.Do bulls and bears move across borders?international transmission of stock returns and volatility[J].Review of Financial Studies,1994,7(3):507-538.
[4]赵留彦,王一鸣.A、B股之间的信息流动与波动溢出[J].金融研究,2003,(10):37-52.
[5]洪永淼,成思危,刘艳辉,等.中国股市与世界其他股市之间的大风险溢出效应[J].经济学(季刊),2004,3(3):703-726.
[6]Hartman P,Straetmans S,de Vries CG.Asset market linkages in crisis periods[J].The Review of Economics and Statistics,2004,86(1):313-326.
[7]Longin F,Solnik B.Extreme correlation of international equity markets[J].The Journal of Finance,2001,56(2):649-676.
[8]McNeil AJ,Frey R,Embrechts P.Quantitative risk management:concepts,techniques and tools[M].Princeton University Press,2005.
[9]Longin F.From VaR to stress testing:the extreme value approach[J].Journal of Banking and Finance,2000,24:1097-1130.
[10]魏宇.金融市场的收益分布与EVT风险测度[J].数量经济技术经济研究,2006,23(4):101-110.
[11]林宇,黄登仕,魏宇.胖尾分布及长记忆下的动态EVT-VaR测度研究[J].管理科学学报,2011,14(7):71-82.
[12]Sklar A.Fonctions de répartitionàn Dimensions et leurs marges[J].Publications de l'Institut de Statistique de l'Universitéde Paris,1959,8:229-231.
[13]Wang Z,Chen X,Jin Y,et al.Estimating risk of foreign exchange portfolio:using VaR and CVaR based on GARCH-EVT-Copula model[J].Physica A:Statistical Mechanics and its Applications,2010,389(21):4918-4928.
[14]Berger T.Forecasting value-at-risk using time varying copulas and EVT return distributions[J].International Economics,2013,133:93-106.
[15]刘晓星,邱桂华.基于Copula-EVT模型的我国股票市场流动性调整的VaR和ES研究[J].数理统计与管理,2010,29(1):150-161.
[16]刘晓星,段斌,谢福座.股票市场风险溢出效应研究:基于EVT-Copula-CoVaR模型的分析[J].世界经济,2011,(11):145-159.
[17]于文华,魏宇,康明惠.不同时变Copula-EVT-ES模型精度比较研究[J].管理科学学报,2015,18(5):32-45.
[18]Joe H.Families of m-variate distributions with given margins and m(m-1)/2 bivariate dependence parameters[A].in:L.R?uschendorf,B.Schweitzer,and M.Taylor(Eds.),Distributions with Fixed Marginals and Related Topics[C],Lecture Notes-Monograph Series,1996,28:120-141.
[19]Bedford T,Cooke RM.Probability density decomposition for conditionally dependent random variables modeled by vines[J].Annals of Mathematics and Artificial Intelligence,2001,32(1):245-268.
[20]Bedford T,Cooke RM.Vines-a new graphical model for dependent random variables[J].The Annals of Statistics,2002,30(4):1031-1068.
[21]Kurowicka D,Cooke R.Uncertainty analysis with high dimensional dependence Modelling[M].Wiley,2006.
[22]Kurowicka D,Joe H.Dependence modeling:vine copula handbook[M].World Scientific,2011.
[23]Di?mann J,Brechmann EC,Czado C,et al.Selecting and estimating regular vine copulae and application to financial returns[J].Computational Statistics and Data Analysis,2013,59(1):52-69.
[24]Koliai L.Extreme risk modeling:an EVT-Pair-copulas approach for financial stress tests[J].Journal of Banking and Finance,2016,Forthcoming.
[25]张帮正,魏宇,余江,等.基于EVT-Vine-copula的多市场相关性及投资组合选择研究[J].管理科学,2014,27(3):133-144.
[26]SesaY RB.使用GARCH-EVT和藤式Copula进行极端值依赖性建模和在险价值估计[D].成都:西南财经大学博士学位论文,2014.
[27]苟红军,陈迅,花拥军.基于GARCH-EVT-COPULA模型的外汇投资组合风险度量研究[J].管理工程学报,2015,29(1):183-193.
[28]何娟,王建,蒋祥林,等.基于Copula-CVaR-EVT方法的供应链金融质物组合优化[J].系统工程理论与实践,2015,35(1):1-16.
[29]李丽梅,胡华.基于SVt-EVT-Vine Copula的投资组合Va R研究[J].数学的实践与认识,2016,(9):104-112.
[30]Engle RF.Autoregressive conditional heteroscedasticity with estimates of the variance of United Kingdom inflation[J].Econometrica,1982,50(4):987-1007.
[31]Bollerslev,T.Generalized autoregressive conditional heteroscedasticity[J].Journal of Econometrics,1986,31(3):307-327.
[32]Glosten LR,Jagannathan R,Runkle DE.On the Relation between the expected value and the volatility of the nominal excess return on stocks[J].The Journal of Finance,1993,48(5):1779-1801.
[33]Scarrott C,Mac Donald A.A review of extreme value threshold estimation and uncertainty quantification[J].Statistical Journal,2012,10(1):33-60.
[34]DuMouchel WH.Estimating the table indexαin order to measure tail thickness:a critique[J].Annals of Statistics,1983,11(4):1019-1031.
[35]Rosenberg J,Schuermann T.A general approach to integrated risk management with skewed,fat-tailed risks[Z].Federal Reserve Bank of New York,2004.
(1)《习近平主持召开中央财经领导小组第八次会议》,http://politics.people.com.cn/n/2014/1106/c70731-25989646. html
(2)中共中央宣传部:《习近平总书记系列重要讲话读本》(2016年版),学习出版社、人民出版社,2016年第1版,第266-268页。
(1)图2中,原始指数经正态化处理,以便比较各指数之间的相对变化趋势。
(1)在对数据检验的过程中,文中仅展示中国的情形。如有读者感兴趣,其它股市的检验结果可向作者索取。
(1)此处仅展示中国的情形,如有读者感兴趣,其它股市的检验结果可向作者索取。
(2)对于其它树的结构图,如读者感兴趣,可以向作者索取。
(1)2015年11月22日,中华人民共和国商务部部长高虎城与东盟十国部长在马来西亚吉隆坡正是签署《中华人民共和国与东南亚国家联盟关于修订〈中国—东盟全面经济合作框架协议〉及项下部分协议的议定书》。
(2)尽管所得到的规则藤结构相同,但在不同“树”上,用来描述股票市场之间的条件Copula函数略有差别,且股票市场之间的相依程度不大。如读者感兴趣,可以向作者索取详细结果。