不同GARCH模型对股市风险价值评价的比较研究

The Calculation of VaR in Stock Market Based on Garch Type Models Comparison

通过给出了四种GARCH模型(GARCH-t、EGARCH-t、GJR-GARCHt、Beta-Skew-t-EGARCH),进而实例研究了对数收益率序列的一系列特征,计算了基于不同GARCH模型下的风险价值,并对计算出的风险价值进行了Kupiec检验.通过比较发现GJR-GARCH-t模型计算VaR更为准确,而Bata-Skew-t-EGARCH模型对股市反应较为敏感,在计算VaR上也有一定的优越性.

VaR (value at risk)is a widely-used standard to measure risks in the financial market.By working out four GARCH type models(GARCH-t;EGARCH-t;GJR-GARCH-t;Beta-Skew-t-EGARCH),an empirical study is made of a se-ries of features concerning the logarithmic Return Rate of the four models,and a Kupiec test is then made to verify the VaRs results so as to find out about which model fits best in the reality.