SKST-APARCH model and VaR estimation of securities portfolio
Zhan Yuan-rui · Journal of systems engineering · 2005
The unconditional distribution of financial return series not only shows the skewness but also the excess of kurtosis.In the analysis of volatility and VaR of applying ARCH models,the traditional method combined the symmetrical fat-tails of Student and GED errors distribution and the asymmetrical leverage effect can not solve the asymmetrical problems in the series because the skewness and the excess of kurtosis are not independent in practical.The SKST(skewed student distribution) is introduced into APARCH model for empirical study on volatility and VaR of the SSE Composite and SSE A-share indices in this paper.The result shows for whether volatility or VaR in sample the SKST performs more exact compared with the symmetric distribution of Gauss,Student and GED.