Consider an ARMA(2,0) process. Is the process invertible?
$(1-\phi_1L - \phi_2L^2)X_t = u_t$
I understand that for the process to be invertible, I must assess the root of on the MA side of the equation, however in this case it doesn't exist. Would the process still be invertible or not?
Would the answer be true for an ARMA(p,0) model?
If the lag polynomial for the MA part is equal to 1, that is to say $B(L)=1$ then the ARMA process will be invertible.