Page 1 of 1

explosive Garch

Posted: Wed Aug 03, 2016 12:46 pm
by alm
hello
I estimated a Garch(1,1) model using a dummy variable as a calendar effect in conditional mean equation, but I faced a problem. The sum of the coefficients in the conditional variance equation exceeds one, so the model is explosive.
How should this be analyzed?
what should I do?