基于SVt-POT模型的上证指数动态VaR测度
|
摘要
如何合理地选取阙值是金融风险测度理论和估计建模的热点前沿问题,阈值选择的合适与否会直接影响模型的拟合效果,进而影响到金融风险测度的精确程度.传统的阙值选取方法是以主观判断为基础,本文则创新性地在SVt-POT模型的阈值选择的问题中,采用基于变点分析的Hill估计方法定量选择阈值,通过SVt-POT模型来描述金融风险测度.进一步,本文以中国股市近30年的上证综指收盘价数据为研究对象,共有6 806个观察样本进行实证模拟.结果表明:基于变点分析的Hill估计方法比传统的方法选择的阈值更合适,该模型可以有效地描述上证指数日收益率的波动特征并计算出动态的风险值,这为股市投资者提供了规避风险的策略选择,在一定程度上具有应用价值.
How to reasonably select the devaluation is the hot topic of financial risk measurement theory and estimation modeling. The appropriate choice of threshold value will directly affect the fitting effect of the model, and thus affect the accuracy of financial risk measurement.The traditional method of devaluation selection is based on subjective judgment. In this paper, the threshold selection method of SVt-POT model is innovatively used. The Hill estimation method based on change point analysis is used to quantitatively select thresholds, which are described by SVt-POT model. Financial risk measurement. Further, this paper takes the data of the closing price of the Shanghai Composite Index in the Chinese stock market for nearly 30 years as the research object, and a total of 6806 observation samples are used for empirical simulation.The results show that the Hill estimation method based on the change point analysis is more suitable than the threshold of the traditional method selection. The model can effectively describe the fluctuation characteristics of the daily yield of the Shanghai Stock Index and calculate the dynamic risk value. This provides the stock market investors with The choice of strategies to avoid risks has application value to a certain extent.
How to reasonably select the devaluation is the hot topic of financial risk measurement theory and estimation modeling. The appropriate choice of threshold value will directly affect the fitting effect of the model, and thus affect the accuracy of financial risk measurement.The traditional method of devaluation selection is based on subjective judgment. In this paper, the threshold selection method of SVt-POT model is innovatively used. The Hill estimation method based on change point analysis is used to quantitatively select thresholds, which are described by SVt-POT model. Financial risk measurement. Further, this paper takes the data of the closing price of the Shanghai Composite Index in the Chinese stock market for nearly 30 years as the research object, and a total of 6806 observation samples are used for empirical simulation.The results show that the Hill estimation method based on the change point analysis is more suitable than the threshold of the traditional method selection. The model can effectively describe the fluctuation characteristics of the daily yield of the Shanghai Stock Index and calculate the dynamic risk value. This provides the stock market investors with The choice of strategies to avoid risks has application value to a certain extent.
引文
[1]Jorion P.Value at Risk:The New Benchmark for Controlling Market Risk[M].New York:The McGraw and Hill ComkpaniesInc,1997.
[2]Hotta L K,Lucas E C,Palaro H P.Estimation of VAR Using Copula and Extreme Value Theory[J].Social Science Electronic Publishing,2008,12(4):205-218.
[3]陈学华,杨辉耀.VaR-APARCH模型与证券投资风险量化分析[J].中国管理科学,2003,11(1):22-27.
[4]徐炜,黄炎龙.GARCH模型与VaR的度量研究[J].数量经济技术经济研究,2008,25(1):120-132.
[5]林宇.基于双曲线记忆HYGARCH模型的动态风险VaR测度能力研究[J].中国管理科学,2011,19(6):15-24.
[6]Wright J H,Hirano K.On the Normal Inverse Gaussian Stochastic Volatility Model[J].Journal of Business&Economic Statistics,2001,19(1):44-54.
[7]Taylor S J.Stock price distribution with stochastic volatility:An analytic approach[J].Review of Financial Studies,1982(4):727-752.
[8]Robinson P M.The Memory of Stochastic Volatility Models[J].Journal of Econometrics,2001,101(2):195-218.
[9]Piterbarg V V.Stochastic Volatility Model with Time-dependent Skew[J].Applied Mathematical Finance,2005,12(2):147-185.
[10]JacquierE,Polson N G,Rossi P E.Bayesian Analysis of Stochastic Volatility Models[J].Journal of Business&Economic Statistics1994,12(4):371-389.
[11]余素红,张世英,宋军.基于GARCH模型和SV模型的VaR比较[J].管理科学学报,2004,7(5):61-66.
[12]董耀武,周孝华,姜婷.基于EVT-POT-SVt模型的风险度量研究[J].北京理工大学学报(社会科学版),2011,13(3):41-45.
[13]苏理云,王杰.基于SV-POT模型的黄金市场的动态VaR预测[J].重庆理工大学学报,2017,31(5):162-168.
[14]叶五一,缪柏其.应用改进Hill估计计算在险价值[J].中国科学院大学学报,2004,21(3):305-309.
[15]李强,邹晓峰.变点分析与极值指标和SV-t-GPD模型在原油极端风险测度中的应用[J].数理统计与管理,2015,34(4):592-602.
[16]Liesenfeld R,Jung R C.Stochastic volatility models:conditional normality versus heavy tailed distributions[J].Journal of Applied Econometrics,2000,15(2):137-160.
[17]王元月,杜希庆,曹圣山.阈值选取的Hill估计方法改进---基于极值理论中POT模型的实证分析[J].中国海洋大学学报(社会科学版),2012(3):42-46.
[18]宋加山,李勇,彭诚,等.极值理论中阈值选取的Hill估计方法改进[J].中国科学技术大学学报,2008,38(9):1104-1108.
[19]陈希孺.变点统计分析简介[J].数理统计与管理,1991(1):52-59.
[20]叶五一,缪柏其,谭常春.基于分位点回归模型变点检测的金融传染分析[J].数量经济技术经济研究,2007,24(10):151-160.
[21]吴鑫育,周海林.基于已实现SV模型的动态VaR测度研究[J].管理工程学报,2018(2):144-149.
[22]Pickands J.Multivariate Negative Exponential and Extreme Value Distributions[J].Lecture Notes in Statistics,1989(51):262-274.
[2]Hotta L K,Lucas E C,Palaro H P.Estimation of VAR Using Copula and Extreme Value Theory[J].Social Science Electronic Publishing,2008,12(4):205-218.
[3]陈学华,杨辉耀.VaR-APARCH模型与证券投资风险量化分析[J].中国管理科学,2003,11(1):22-27.
[4]徐炜,黄炎龙.GARCH模型与VaR的度量研究[J].数量经济技术经济研究,2008,25(1):120-132.
[5]林宇.基于双曲线记忆HYGARCH模型的动态风险VaR测度能力研究[J].中国管理科学,2011,19(6):15-24.
[6]Wright J H,Hirano K.On the Normal Inverse Gaussian Stochastic Volatility Model[J].Journal of Business&Economic Statistics,2001,19(1):44-54.
[7]Taylor S J.Stock price distribution with stochastic volatility:An analytic approach[J].Review of Financial Studies,1982(4):727-752.
[8]Robinson P M.The Memory of Stochastic Volatility Models[J].Journal of Econometrics,2001,101(2):195-218.
[9]Piterbarg V V.Stochastic Volatility Model with Time-dependent Skew[J].Applied Mathematical Finance,2005,12(2):147-185.
[10]JacquierE,Polson N G,Rossi P E.Bayesian Analysis of Stochastic Volatility Models[J].Journal of Business&Economic Statistics1994,12(4):371-389.
[11]余素红,张世英,宋军.基于GARCH模型和SV模型的VaR比较[J].管理科学学报,2004,7(5):61-66.
[12]董耀武,周孝华,姜婷.基于EVT-POT-SVt模型的风险度量研究[J].北京理工大学学报(社会科学版),2011,13(3):41-45.
[13]苏理云,王杰.基于SV-POT模型的黄金市场的动态VaR预测[J].重庆理工大学学报,2017,31(5):162-168.
[14]叶五一,缪柏其.应用改进Hill估计计算在险价值[J].中国科学院大学学报,2004,21(3):305-309.
[15]李强,邹晓峰.变点分析与极值指标和SV-t-GPD模型在原油极端风险测度中的应用[J].数理统计与管理,2015,34(4):592-602.
[16]Liesenfeld R,Jung R C.Stochastic volatility models:conditional normality versus heavy tailed distributions[J].Journal of Applied Econometrics,2000,15(2):137-160.
[17]王元月,杜希庆,曹圣山.阈值选取的Hill估计方法改进---基于极值理论中POT模型的实证分析[J].中国海洋大学学报(社会科学版),2012(3):42-46.
[18]宋加山,李勇,彭诚,等.极值理论中阈值选取的Hill估计方法改进[J].中国科学技术大学学报,2008,38(9):1104-1108.
[19]陈希孺.变点统计分析简介[J].数理统计与管理,1991(1):52-59.
[20]叶五一,缪柏其,谭常春.基于分位点回归模型变点检测的金融传染分析[J].数量经济技术经济研究,2007,24(10):151-160.
[21]吴鑫育,周海林.基于已实现SV模型的动态VaR测度研究[J].管理工程学报,2018(2):144-149.
[22]Pickands J.Multivariate Negative Exponential and Extreme Value Distributions[J].Lecture Notes in Statistics,1989(51):262-274.