基于条件风险价值法的标普500指数期货风险预警研究
摘要
随着中国首个股指期货的推出,中国结束了长达28年的股指期货市场空白,走向了更加成熟的资本市场建设阶段。但股指期货由于其自身的投机性及高杠杆性,导致股指期货市场的风险更复杂,影响更大。而且美国次贷危机之后,通过金融渠道传染的金融危机越来越受到大家的广泛关注,中国在这个复杂的国际环境下推出股指期货,其风险的预警与防范对中国股指期货市场以至于整个资本市场的健康发展都具有非常重要的现实意义。
以具有代表性的标普500指数期货为研究对象,首先,对标普500指数期货的内涵进行了简单的分析,包括基本统计量分析、风险的类型、成因及特点;然后进行定量分析,通过建立非对称的成分自回归条件异方差-广义误差分布(非对称的CARCH-GED)模型消除标普500指数期货收益率序列存在的有偏、尖峰厚尾、波动聚集性以及杠杆效应,并且采用更加审慎的风险度量方法——条件风险价值法(CVaR),利用Matlab软件对标普500指数期货的风险进行准确的度量,从而达到风险预警的效果。最后,通过对比分析证明了沪深300指数期货与标普500指数期货的收益率具有共同的形态特征,基于非对称的CARCH-GED-CVaR模型的标普500指数期货风险预警模型也适用于中国股指期货的风险预警并通过具体数据加以验证。通过对两国股指期货的比较,可以发现中国应该逐渐强化期货业协会和期货交易所在股指期货风险监管中的作用,通过完善股指期货法律制度逐步解决我国股指期货存在的自身缺陷问题,从而加快我国股指期货市场的健康发展。
As China has introduced the first stock index futures, China ends its 28-year blank in the stock index futures market and come towards a more mature capital market construction phase. Because of its high leverage and speculative, the risk of stock index futures is more complex and influential. After the subprime mortgage crisis of the USA, the financial crisis through the financial channel gets more concern. In this complex international environment, stock index futures early warning has very important practical significance in the healthy development of stock index futures and the capital market.
Standard & Poor’s 500 (S&P 500) index futures is taken as the research object. First of all, analyze the connotation of S&P 500 index futures, including the basic statistics, the types, causes and characteristics of the risk. Then, make a quantitative analysis through building a Asymmetric Component Auto-Regression Conditions Heteroscedastic-General error distribution (Asymmetric CARCH-GED) model, eliminating the biased, thick tail of peak, fluctuation of agglomeration and leverage characteristics of the rate of return sequence of the S&P 500 index futures. And we adopt a more prudent risk measurement method conditional value at risk (CVaR) to accutately measure the risk of S&P 500 index futures with Matlab, so as to achieve the effect of early warning. Finally, we prove that the CSI 300 index futures and the S&P 500 index futures have common characteristics through the constast analysis of them. The early warning model of S&P 500 index futures based on CARCH-GED-CVaR is also applicable for the CSI 300 index futures. Through the comparison of the two index futures, it can be found that China should gradually strengthen the role of the futures industry association and futures exchange in risk supervision of stock index futures, solve the defects existing in the stock index futures through improving the stock index futures legal system, thus speeding up the healthy development of the stock index futures.
以具有代表性的标普500指数期货为研究对象,首先,对标普500指数期货的内涵进行了简单的分析,包括基本统计量分析、风险的类型、成因及特点;然后进行定量分析,通过建立非对称的成分自回归条件异方差-广义误差分布(非对称的CARCH-GED)模型消除标普500指数期货收益率序列存在的有偏、尖峰厚尾、波动聚集性以及杠杆效应,并且采用更加审慎的风险度量方法——条件风险价值法(CVaR),利用Matlab软件对标普500指数期货的风险进行准确的度量,从而达到风险预警的效果。最后,通过对比分析证明了沪深300指数期货与标普500指数期货的收益率具有共同的形态特征,基于非对称的CARCH-GED-CVaR模型的标普500指数期货风险预警模型也适用于中国股指期货的风险预警并通过具体数据加以验证。通过对两国股指期货的比较,可以发现中国应该逐渐强化期货业协会和期货交易所在股指期货风险监管中的作用,通过完善股指期货法律制度逐步解决我国股指期货存在的自身缺陷问题,从而加快我国股指期货市场的健康发展。
As China has introduced the first stock index futures, China ends its 28-year blank in the stock index futures market and come towards a more mature capital market construction phase. Because of its high leverage and speculative, the risk of stock index futures is more complex and influential. After the subprime mortgage crisis of the USA, the financial crisis through the financial channel gets more concern. In this complex international environment, stock index futures early warning has very important practical significance in the healthy development of stock index futures and the capital market.
Standard & Poor’s 500 (S&P 500) index futures is taken as the research object. First of all, analyze the connotation of S&P 500 index futures, including the basic statistics, the types, causes and characteristics of the risk. Then, make a quantitative analysis through building a Asymmetric Component Auto-Regression Conditions Heteroscedastic-General error distribution (Asymmetric CARCH-GED) model, eliminating the biased, thick tail of peak, fluctuation of agglomeration and leverage characteristics of the rate of return sequence of the S&P 500 index futures. And we adopt a more prudent risk measurement method conditional value at risk (CVaR) to accutately measure the risk of S&P 500 index futures with Matlab, so as to achieve the effect of early warning. Finally, we prove that the CSI 300 index futures and the S&P 500 index futures have common characteristics through the constast analysis of them. The early warning model of S&P 500 index futures based on CARCH-GED-CVaR is also applicable for the CSI 300 index futures. Through the comparison of the two index futures, it can be found that China should gradually strengthen the role of the futures industry association and futures exchange in risk supervision of stock index futures, solve the defects existing in the stock index futures through improving the stock index futures legal system, thus speeding up the healthy development of the stock index futures.
引文
[1] Paul Krugman. The Third Depression[N]. New York Times,2010-06-27
[2] John F O, Bilson. Leading Indicators of Currency Devaluations[J]. The International Executive,1980(22):21-23
[3] Guillermo A Calvo, Morris Goldstein, Eduard Hochreiter. Private Capital Flows to Emerging Markets After the Mexican Crisis[M]. Washington DC,Institute for International Economics, 1996:233-282.
[4] Frankel, Jeffery A, Andrew K Rose. Currency Crashes in Emerging Markets: An Empirical Treatment[J]. Journal of International Economics,1996(41):351-368.
[5] Sachs J, A Tornell, A Velasco. Financial Crises in Emerging Markets: The Lessons from 1995[J]. Brookings Papers on Economic Activity,1996(1):147-198.
[6] Kaminsky, Lizondo, M Reinhart. Leading Indicators of Currency Crises[R]. IMF WORKING PAPERS,1998,45(1):1-48.
[7] Glyn A Holton. Value-at-risk: theory and practice[M]. 2003:18~21.
[8] Robert F Engle. Autoregressive conditional heteroscedasticity with estimates of the variance of United Kingdom inflation[J]. Econometrics,1982(50):987-1008.
[9] P Artzner, F Delbaen. Thinking coherently[J]. Risk,1997(10):33-49.
[10] Frittelli, Rosazza Gianin. Putting order in risk measures[J]. Journal of Banking & Finance,2002(26):1473-1486.
[11] Giorgio S. Measures of risk[J]. Journal of Banking & Finance,2002(26):1253-1272.
[12] Beder T S. VaR: Seductive but dangerous[J]. Financial Analysts Journal,1995(51):12-24.
[13] P Artzner, F Delbaen. Coherent Measure of Risk[J]. Mathematical Finance,1999(3):203-228.
[14] R McKay, T E Keefer. VaR is a dangerous technique[J]. Corporate Finance Searching for Systems Integration Supplement,1996(9):30.
[15] Rockafellar R, Uryasev S. Optimization of Conditional Value-at-Risk[J]. Journal of risk,2000(2):21-41.
[16] Pavlo Krokhmal, Jonas Palmquist. Portfolio Optimization with Conditional Value-at-Risk Objective and Constraints [2001-9-25]. http://www.ise.ufl.edu/ uryasev.
[17] Topaloglou N, Vladimirou H, Zenios S. CVaR models with selective hedging for international asset allocation[J]. Journal of Banking & Finance,2002(26):1535-1561.
[18] Alexander G J, Baptista A M. Economic Implications of Using a Mean-VaR Model for Portfolio Selection: A Comparison with Mean-VaRiance Analysis[J]. Journal of Economic Dynamics and Control,2002(26):1159-1193.
[19] Fredrik Andersson, Helmut Mausser. Credit Risk Optimization with Conditional Value-at-Risk Criterion[OL].(2000-10-1) [2000-12-15]. http:// www. ise.ufl.edu/ uryasev /Credit_risk optimization .pdf.
[20] Renata Mansini, Wlodzimierz Ogryczak. Conditional value at risk and related linear programming models for portfolio optimization[J]. Annals of operations research,2006(152):227-256.
[21]迟国泰,余方平,李洪江,等.单个期货合约市场风险VaR-GARCH评估模型及其应用研究[J].大连理工大学学报,2006(1):127-134.
[22]印凡成,王晶,茹正亮. GARCH-M模型在股指预测中的应用[J].贵州大学学报(自然科学版),2010(2):14-17.
[23]李基梅,刘青青. VaR-GARCH模型在我国股指期货风险管理中的应用[J].山东理工大学(自然科学版),2009(4):73-76.
[24]佘传奇,汤益民.基于Bp网络的股指期货风险预警体系研究[N].期货日报,2008-04-16(003).
[25]蒋敏.条件风险值(CVaR)模型的理论研究[D].西安:西安电子科技大学学位论文,2005:21-74.
[26]包峰,俞金平,李胜宏. CVaR对VaR的改进与发展[J].山东师范大学学报,2005,20(4):95-96.
[27]刘小茂,田立. CVaR与VaR的对比研究及实证分析[J].华中科技大学学报,2005,33(10):112-114.
[28]蒋敏,蒋宝珍,孟之青,等.基于多目标CVaR模型的证券组合投资的风险度量和策略[J].经济数学,2007(24):385-397.
[29]方楚旭.多期投资组合选择的均值-CVaR模型[J].金融经济,2007(l0):94-95.
[30]彭正宇. VaR及其改进模型CVaR[J].科技情报开发与经济,2008(24):168-169.
[31]郑玉仙.风险测量的VaR及CVaR方法的对比研究[J].生产力研究,2010(4):146-147.
[32]王树娟,黄渝祥.基于GARCH-CVaR模型的我国股票市场风险分析[J].同济人学学报,2005,33(2):260-262.
[33]李海涛.基于CVaR的衍生证券投资组合优化模型研究[D].武汉:武汉理工大学学位论文,2005:20-40.
[34]杨琦峰,任方,杨恩宁. VaR与CVaR在商业银行风险度量方面的比较研究[J].当代经济,2006(7):92-93.
[35]王雨飞.基于遗传算法的CVaR模型研究[D].西安:西安电子科技大学学位论文,2007:16-30.
[36]林财超. VaR和CVaR在证券投资组合决策中的应用[D].长沙:长沙理工大学学位论文,2008:13-36.
[37]王宝森,梁奉.基于CVaR的投资组合优化模型及实证[J].重庆工商大学学报(自然科学版),2010(3):213-222.
[38]魏丹,单锋.基于CVaR风险度量方法的投资组合模型研究[J].沈阳航空工业学院学报,2010(3):80-82.
[39]王霞.资产组合椭球分布下的CVaR及均值-CVaR有效前沿[D].上海:华东师范大学学位论文,2006:8-24.
[40]蒋春福,尤川川,彭红毅.偏尾Laplace分布下的条件风险价值探讨[J].统计与决策,2008(5):27-29.
[41]林孝贵,聂永红.正态分布下期货套期保值CVaR风险的敏感度[J].中国管理信息化,2009(12):63-65.
[42]周健.稳定分布下的CVaR分析[J].重庆工学院学报,2009(1):165-168.
[43]姚海祥.基于均值和CVaR的效用最大化模型研究[J].数理统计与管理,2010(5):913-920.
[44]王晶,张裕生,王玉玲.均值-CVaR优化模型在投资组合中的应用[J].枣庄学院学报,2010(2):41-44.
[45]姚慧丽.均值-CVaR投资组合有效边缘的灵敏度分析[J].统计与决策,2010(19):65-67.
[46]向华,周伟峰.协方差矩阵为奇异的均值-CVaR模型的研究[J].重庆工商大学学报(自然科学版),2010(27):327-320.
[47]姚海洋.奇异协方差矩阵和任意收益率分布下均值-CVaR模型的有效边界特征[J].中国管理科学,2008(16):273-277.
[48]王丽娜,张丽娟.基于CVaR-SV-N模型的股指期货风险度量[J].金融市场,2010(10):60-64.
[49] Zhou Ying, Zhang Hong-xi, Wu Hui-shuo. Market Risk Evaluation on Single Futrues Contract: SV-CVaR Model and Its Application on Cu00 Data[J]. Journal of Beijing Institute of Techonlogy,2009,18(3):365-369.
[50]王树娟,黄渝祥.基于GARCH-CVaR模型的我国股票市场风险分析[J].同济大学学报(自然科学版).2005(2):260-263.
[51]周小敏.基于GARCH模型的CVaR金融风险测度研究[D].长沙:湖南大学学位论文,2007:40-60.
[52]刘琦铀,张能福,刘铁生.基于GARCH模型的CVaR信贷风险度量方法研究[J].统计与决策,2010(10):26-29.
[53] Timmerman A. Structural breaks,incomplete information,and stock prices[J]. Journal of Business and Economic Statistics,2001(3):299-314.
[54]李鹏立. S&P500指数研究.证券市场导报[J].证券市场导报,2000(2):44-46.
[55] Bernard Baumohl.The Secrets of Economic Indicators (Second Edition)[M]. New Jersey:Wharton School Publishing,2008:13.
[56] R Tyrell Rockafellar, Stanislav Uryasev. Conditional value-at-risk for general loss distributions[J]. Journal of Banking & Finance,2002(26):1443-1470.
[57] P Artzner, F Delbaen, J-M Eber, et al. Coherent measures of risk[J]. Mathematical Finance,1999(9):203-228.
[58]胡辉.我国金融风险预警机制研究[D].江苏:江苏大学学位论文,2008:28-40.
[59]中国金融期货交易所.中国金融期货交易所交易规则及实施细则修订说明[EB],(2010-1-19):5-6.
[2] John F O, Bilson. Leading Indicators of Currency Devaluations[J]. The International Executive,1980(22):21-23
[3] Guillermo A Calvo, Morris Goldstein, Eduard Hochreiter. Private Capital Flows to Emerging Markets After the Mexican Crisis[M]. Washington DC,Institute for International Economics, 1996:233-282.
[4] Frankel, Jeffery A, Andrew K Rose. Currency Crashes in Emerging Markets: An Empirical Treatment[J]. Journal of International Economics,1996(41):351-368.
[5] Sachs J, A Tornell, A Velasco. Financial Crises in Emerging Markets: The Lessons from 1995[J]. Brookings Papers on Economic Activity,1996(1):147-198.
[6] Kaminsky, Lizondo, M Reinhart. Leading Indicators of Currency Crises[R]. IMF WORKING PAPERS,1998,45(1):1-48.
[7] Glyn A Holton. Value-at-risk: theory and practice[M]. 2003:18~21.
[8] Robert F Engle. Autoregressive conditional heteroscedasticity with estimates of the variance of United Kingdom inflation[J]. Econometrics,1982(50):987-1008.
[9] P Artzner, F Delbaen. Thinking coherently[J]. Risk,1997(10):33-49.
[10] Frittelli, Rosazza Gianin. Putting order in risk measures[J]. Journal of Banking & Finance,2002(26):1473-1486.
[11] Giorgio S. Measures of risk[J]. Journal of Banking & Finance,2002(26):1253-1272.
[12] Beder T S. VaR: Seductive but dangerous[J]. Financial Analysts Journal,1995(51):12-24.
[13] P Artzner, F Delbaen. Coherent Measure of Risk[J]. Mathematical Finance,1999(3):203-228.
[14] R McKay, T E Keefer. VaR is a dangerous technique[J]. Corporate Finance Searching for Systems Integration Supplement,1996(9):30.
[15] Rockafellar R, Uryasev S. Optimization of Conditional Value-at-Risk[J]. Journal of risk,2000(2):21-41.
[16] Pavlo Krokhmal, Jonas Palmquist. Portfolio Optimization with Conditional Value-at-Risk Objective and Constraints [2001-9-25]. http://www.ise.ufl.edu/ uryasev.
[17] Topaloglou N, Vladimirou H, Zenios S. CVaR models with selective hedging for international asset allocation[J]. Journal of Banking & Finance,2002(26):1535-1561.
[18] Alexander G J, Baptista A M. Economic Implications of Using a Mean-VaR Model for Portfolio Selection: A Comparison with Mean-VaRiance Analysis[J]. Journal of Economic Dynamics and Control,2002(26):1159-1193.
[19] Fredrik Andersson, Helmut Mausser. Credit Risk Optimization with Conditional Value-at-Risk Criterion[OL].(2000-10-1) [2000-12-15]. http:// www. ise.ufl.edu/ uryasev /Credit_risk optimization .pdf.
[20] Renata Mansini, Wlodzimierz Ogryczak. Conditional value at risk and related linear programming models for portfolio optimization[J]. Annals of operations research,2006(152):227-256.
[21]迟国泰,余方平,李洪江,等.单个期货合约市场风险VaR-GARCH评估模型及其应用研究[J].大连理工大学学报,2006(1):127-134.
[22]印凡成,王晶,茹正亮. GARCH-M模型在股指预测中的应用[J].贵州大学学报(自然科学版),2010(2):14-17.
[23]李基梅,刘青青. VaR-GARCH模型在我国股指期货风险管理中的应用[J].山东理工大学(自然科学版),2009(4):73-76.
[24]佘传奇,汤益民.基于Bp网络的股指期货风险预警体系研究[N].期货日报,2008-04-16(003).
[25]蒋敏.条件风险值(CVaR)模型的理论研究[D].西安:西安电子科技大学学位论文,2005:21-74.
[26]包峰,俞金平,李胜宏. CVaR对VaR的改进与发展[J].山东师范大学学报,2005,20(4):95-96.
[27]刘小茂,田立. CVaR与VaR的对比研究及实证分析[J].华中科技大学学报,2005,33(10):112-114.
[28]蒋敏,蒋宝珍,孟之青,等.基于多目标CVaR模型的证券组合投资的风险度量和策略[J].经济数学,2007(24):385-397.
[29]方楚旭.多期投资组合选择的均值-CVaR模型[J].金融经济,2007(l0):94-95.
[30]彭正宇. VaR及其改进模型CVaR[J].科技情报开发与经济,2008(24):168-169.
[31]郑玉仙.风险测量的VaR及CVaR方法的对比研究[J].生产力研究,2010(4):146-147.
[32]王树娟,黄渝祥.基于GARCH-CVaR模型的我国股票市场风险分析[J].同济人学学报,2005,33(2):260-262.
[33]李海涛.基于CVaR的衍生证券投资组合优化模型研究[D].武汉:武汉理工大学学位论文,2005:20-40.
[34]杨琦峰,任方,杨恩宁. VaR与CVaR在商业银行风险度量方面的比较研究[J].当代经济,2006(7):92-93.
[35]王雨飞.基于遗传算法的CVaR模型研究[D].西安:西安电子科技大学学位论文,2007:16-30.
[36]林财超. VaR和CVaR在证券投资组合决策中的应用[D].长沙:长沙理工大学学位论文,2008:13-36.
[37]王宝森,梁奉.基于CVaR的投资组合优化模型及实证[J].重庆工商大学学报(自然科学版),2010(3):213-222.
[38]魏丹,单锋.基于CVaR风险度量方法的投资组合模型研究[J].沈阳航空工业学院学报,2010(3):80-82.
[39]王霞.资产组合椭球分布下的CVaR及均值-CVaR有效前沿[D].上海:华东师范大学学位论文,2006:8-24.
[40]蒋春福,尤川川,彭红毅.偏尾Laplace分布下的条件风险价值探讨[J].统计与决策,2008(5):27-29.
[41]林孝贵,聂永红.正态分布下期货套期保值CVaR风险的敏感度[J].中国管理信息化,2009(12):63-65.
[42]周健.稳定分布下的CVaR分析[J].重庆工学院学报,2009(1):165-168.
[43]姚海祥.基于均值和CVaR的效用最大化模型研究[J].数理统计与管理,2010(5):913-920.
[44]王晶,张裕生,王玉玲.均值-CVaR优化模型在投资组合中的应用[J].枣庄学院学报,2010(2):41-44.
[45]姚慧丽.均值-CVaR投资组合有效边缘的灵敏度分析[J].统计与决策,2010(19):65-67.
[46]向华,周伟峰.协方差矩阵为奇异的均值-CVaR模型的研究[J].重庆工商大学学报(自然科学版),2010(27):327-320.
[47]姚海洋.奇异协方差矩阵和任意收益率分布下均值-CVaR模型的有效边界特征[J].中国管理科学,2008(16):273-277.
[48]王丽娜,张丽娟.基于CVaR-SV-N模型的股指期货风险度量[J].金融市场,2010(10):60-64.
[49] Zhou Ying, Zhang Hong-xi, Wu Hui-shuo. Market Risk Evaluation on Single Futrues Contract: SV-CVaR Model and Its Application on Cu00 Data[J]. Journal of Beijing Institute of Techonlogy,2009,18(3):365-369.
[50]王树娟,黄渝祥.基于GARCH-CVaR模型的我国股票市场风险分析[J].同济大学学报(自然科学版).2005(2):260-263.
[51]周小敏.基于GARCH模型的CVaR金融风险测度研究[D].长沙:湖南大学学位论文,2007:40-60.
[52]刘琦铀,张能福,刘铁生.基于GARCH模型的CVaR信贷风险度量方法研究[J].统计与决策,2010(10):26-29.
[53] Timmerman A. Structural breaks,incomplete information,and stock prices[J]. Journal of Business and Economic Statistics,2001(3):299-314.
[54]李鹏立. S&P500指数研究.证券市场导报[J].证券市场导报,2000(2):44-46.
[55] Bernard Baumohl.The Secrets of Economic Indicators (Second Edition)[M]. New Jersey:Wharton School Publishing,2008:13.
[56] R Tyrell Rockafellar, Stanislav Uryasev. Conditional value-at-risk for general loss distributions[J]. Journal of Banking & Finance,2002(26):1443-1470.
[57] P Artzner, F Delbaen, J-M Eber, et al. Coherent measures of risk[J]. Mathematical Finance,1999(9):203-228.
[58]胡辉.我国金融风险预警机制研究[D].江苏:江苏大学学位论文,2008:28-40.
[59]中国金融期货交易所.中国金融期货交易所交易规则及实施细则修订说明[EB],(2010-1-19):5-6.