Matlab and Stata show different results (VAR). Am I doing something wrong?

Hello. I am running a simple var-model in Matlab first using the following command:

vgxset('n',2,'nAR',2,'Constant',true)

And then this one:

[EstSpec,EstStdErrors,logL,W] = vgxvarx(Spec,Y);

My results are the following:

    Model  : 2-D VAR(2) with Additive Constant
             Conditional mean is AR-stable and is MA-invertible
             Standard errors without DoF adjustment (maximum likelihood)
         Parameter          Value     Std. Error    t-Statistic
    -------------- -------------- -------------- --------------
              a(1)       0.432328       0.100162        4.31626 
              a(2)       0.583732       0.103499        5.63997 
        AR(1)(1,1)       0.348238      0.0812341        4.28685 
             (1,2)      -0.115787      0.0776992        -1.4902 
             (2,1)     -0.0379696      0.0839402      -0.452341 
             (2,2)       0.520637      0.0802876        6.48465 
        AR(2)(1,1)       0.144522      0.0803325        1.79905 
             (1,2)     -0.0140245      0.0781883      -0.179368 
             (2,1)     -0.0332596      0.0830086      -0.400676 
             (2,2)      -0.204997       0.080793       -2.53731 
            Q(1,1)       0.416339                               
            Q(2,1)      0.0524138                               
            Q(2,2)        0.44454

But in Stata the coefficients are different (the dataset is obviously the same).I am running a simplistic 2-lag var-model w/o any modifications. Could somebody tell me what's the deal here?