基于非参数和半参数CARR模型的上海股票市场波动性研究

Nonparametric and Semiparametric CARR Models for Shanghai Stock Market Volatility

以上证综指日对数价格的极差为研究对象,分别建立参数、半参数和非参数CARR(1,1)模型来研究上海股票市场的波动性。采用MSE、MAE两种误差度量指标比较参数、非参数、半参数CARR(1,1)模型的拟合能力。结果表明:半参数CARR(1,1)模型在对上海股市波动性的拟合方面表现最优,非参数CARR(1,1)模型次之,GCARR(1,1)模型最差。

Previous empirical results reveal that CARR significantly outperforms GARCH in the prediction of volatility.As we all know,the estimation of CARR is based on the function form and the residual＇s distribution.It is because the estimation of the nonparametric and semi-paremetric CARR ignores the hypothesis,and the two models can reduce the error obviously.By using the daily range data of Shanghai composite index,we establish the parametric,nonparametric and semi-paremetric CARR to study Shanghai stock market＇s volatility.We select MSE and MAE to compare the fitting ability of the three models.The results show that among the three models,the best one to feature shanghai stock market＇s volatility is semi-parametric CARR,and the nonparametric CARR is inferior and the weak one is parametric CARR.