在用eviews对金融数据做garch(1,1)检验时, 得到ARCH(1)和GARCH(1)的P值均大于20%,此时用gacch(1,1)系数估算波动率是否有意义,如果没有意义,如何改进其波动率的计算,另外,我想询问一下如何判断金融数据有garch效应,如果garch(1,1)不满足,那么garch二价应用时,是否有garch(1,1)那样良好的波动率期限结构
请注明:姓名-公司-职位
以便审核进群资格,未注明则拒绝
typically GARCH(1,1) has been enough. no one would like to use GARCH(p,q) where p or q is greater than2.
You can use LM test to investigate whether there is ARCH effect. standard time series text will normally give this test method.
谢谢!!!