多分形波动率(Multifractal Volatility,MFV)是一种最近提出的金融市场波动率测度方法。以上证综指和标准普尔500指数的高频价格数据为例,构造了多分形波动率测度的lnMFV-ARMA动力学模型,并运用基于Bootstrap方法的后验分析过程,实证对比了lnMFV-ARMA模型与其他6种常用波动模型对ES(Excepted Shortfall)风险测度的估计精度差异。实证结果表明:在所考察的大多数分位数水平下,lnMFV-ARMA模型对ES风险测度的估计精度都优于许多现有常用波动模型,特别是对标准普尔500指数的极端价格波动风险具有最优的刻画能力。
Multifractal Volatility(MFV) is a recently proposed volatility measurement.We take the high-frequency datasets of SSEC and SP500 as sample and construct an lnMFV-ARMA model to estimate excepted shortfall.We further compare lnMFV-ARMA model with other volatility models based on the ES estimation precision.Empirical results show that model based on multifractal volatility performs better than many GARCH models and lnMFV-ARMA performs best in ES estimation to SP500 index.