基于非对称GARCH-GH的期权定价公式及实证分析

摘要

In this paper,the Eerser measurement transformation technique is used to construct the asymmetric GARCH-GH option pricing formula.In this formula,the non normal distribution and the asymmetric GARCH model are used to describe the peak,the thick tail,the skewness distribution and the random fluctuation characteristics of the financial assets.The empirical results show that compared with the HN model (Heston and Nandi(2000)),which is called the new standard of option pricing,the pricing error of EGARCH-NIG model is small,and the pricing error of the EGARCH-NIG model is reduced by 26.24％ compared with the HN model.For the real valued option,the pricing errors of EGARCH-GH,GJR-GH and GJR-NIG model are less than the HN model.The estimation error and the real valued degree is negative correlation:the real valued degree is stronger,and the error is smaller.%本文应用Eerser测度转换技术构建了非对称GARCH-GH期权定价闭形公式,该公式通过非正态分布和非对称GARCH模型刻画金融资产的尖峰、厚尾、偏态分布的和随机波动特征.通过金融市场交易数据实证发现:与被称为期权定价工业新标准的Heston和Nandi(2001)的HN模型相比,本文构建的EGARCH-NIG模型定价误差更小,比HN模型的定价误差减小了26.24％,对于实值程度较强的期权,本文的EGARCH-GH、GJR-GH和GJR-NIG模型的定价误差均小于HN模型.模型的估计误差与期权实值程度成反比,即实值程度越强,估计误差越小,所估计的结果越准确.