基于GARCH-EVT模型的我国创业板市场风险分析

Risk analysis of Chinese gem market based on GARCH-EVT model

选取了2015年10月至2021年9月创业板综指的日收盘价数据,使用GARCH部分拟合创业板的动态波动的基础上,利用极值理论拟合尾部极端数据,并用POT方法计算出残差序列不同分位数下的VaR和ES估计,最后将其与历史模拟法、普通GARCH模型同时进行回测分析,比较其估计效果。实证结果表明,利用GARCH-EVT组合模型,在较高置信水平下计算得到ES能够更加精确地度量我国创业板市场的风险。

This paper selects daily closing price data of gem composite index from October 2015 to September 2021.On the basis of using GARCH to partially fit the dynamic fluctuation of Chinese GEM,the extreme value theory is used to fit the extreme data of the tail.The VaR and ES estimates under different quantiles of the residual series were calculated by the POT method.Finally,they were back-tested with the historical simulation method and the ordinary GARCH model to compare their estimation effects.The empirical results show that using the GARCH-EVT combination model and calculating ES at a higher confidence level can more accurately measure the risk of Chinese GEM.