基于QR-MS(2)-EGARCH(1,1)-st模型的互联网金融指数风险度量

Risk measurement of internet finance index based on QR-MS(2)-EGARCH(1,1)-st model

基于2012—2021年互联网金融指数的日收盘价数据,采用二区制MS-GARCH(1,1)类模型刻画互联网金融指数收益率的波动过程,通过分析选出较优的模型MS(2)-EGARCH(1,1)-st,结果显示,互联网金融指数收益率存在两种划分明显的波动状态:平缓波动状态比剧烈波动状态的持续性更强,且剧烈波动存在非对称效应。将MS-EGARCH模型与分位数回归(QR)模型的组合模型进行互联网金融指数收益率的风险测度,并通过Kupiec回测检验方法计算拟合成功率,结果表明,QR-MS(2)-EGARCH(1,1)-st求解得到的风险价值(VaR)具有较高拟合成功率。

Based on the daily closing price data of the internet finance index from 2012 to 2021,the two-zone MS-GARCH(1,1)model is firstly used to describe the fluctuation process of the internet finance index,and the optimal model MS(2)-EGARCH(1,1)-st is selected through analysis.The results show that the return rate of the internet finance index has two clearly divided states:the mild fluctuation state is more persistent than the shape fluctuation state,and the shape fluctuation state has asymmetric effects.Secondly,the combined model of MS-EGARCH model and quantile regression(QR)model are used to measure the risk of internet finance return series,and the success rate is calculated by Kupiec backtracking test method.The results show that the success rate of value at risk(VaR)obtained by QR-MS(2)-EGARCH(1,1)-st is higher.