I am trying to create an extended GARCH model with external regressors both in the mean and the var.
One of them gives the Pr(>|t|)="0" (not close to 0, but actually "0") for all parameters (3 ext regressors in both mean and var). The t values are indeed very high (100-400). Does this actually mean that the parameters are very significant? Can this be linked to some kind of problem? The Robust Standard Errors are not equal to 0 but also small.