GARCH模型的相依性

Quadrant Dependent Property for GARCH Model

验证广义自回归条件异方差(GARCH)过程中的绝对值序列和平方序列都是两两PQD序列,讨论了GARCH(1.1)模型的两两正相依性,研究了关于GARCH序列部分和的不等式。这些性质和GARCH过程的条件异方差性是一致的,同时也说明用GARCH模型来描述金融时间序列的波动聚类现象是合理的。

It proves that the absolute value and square sequences of GARCH process are pairwise PQD sequences. The pairwise positive quadrant dependent property of GARCH model was discussed. Some inequalities of sums of GARCH modelwhich are consistent with conditional heteroskedasticity property of GARCH process is obtained. It is reasonable that using the GARCH model to describe the volatility cluster phenomenon of finance time sequences.