How to estimate VARMA (1,1)

How can I estimate VARMA(1,1).I cannot use vgxvarx, since it is only for VAR. I have successfully set the model as VARMA(1,1), but I do not know how to fit the matrix<93*3 double> to the model and get the estimates.

my VARMA setting is: ErrVAR1full = vgxset(ErrVAR1diag,'nAR',1,'nMA',1,'ARsolve',[],'MAsolve',repmat({Errdt},1,1));

and after I type vgxdisp(ErrVAR1full), I get:

Model : 3-D VARMA(1,1) with Additive Constant Conditional mean is uninitialized AR-stable and is uninitialized MA-invertible Series : 1 Series : 2 Series : 3 a Constant: Uninitialized AR(1) Autoregression Matrix: Uninitialized MA(1) Moving Average Matrix: Uninitialized Q Innovations Covariance: Uninitialized

Answer 1

[Shashank Prasanna]

Currently MA terms can't be estimated in a VARMAX model. You can however use the following functions VGXSIM, VGXPRED and VGXINFER to simulate and forecast VARMAX parameters which include MA matrices if you know them before hand. Sorry about that. Try looking online to see if there is a File Exchange implementation or elsewhere.