基于稳定分布的PARCH模型

PARCH Model with Stable Distribution

本文首先介绍了稳定分布和基于正态分布、稳定分布的PARCH模型,并通过股票指数收益率的稳定化PP图和直方图发现其具有高峰厚尾特征.最后,通过上证指数的VaR计算,得到在金融风险度量中基于稳定分布的PARCH模型比基于正态分布的PARCH模型更加有效。

In this paper, Stable distribution and PARCH models based on normal distribution and stable distribution are introduced. It is found from histograms and stablized PP plots of some stock - index return data that their distributions have a high - kurtosis and fat-tail characteristic. The calculation of the VaR for Shanghai stock indices shows that PARCICH medel with stable distribution is more efficient than PARCH model with normal distribution in risk valuation of finance.