Symmetric BEKK with explanatory variables

[Islam Hassouneh]

Dear Tom,

I am estimating a symmetric multivariate BEKK model with two explanatory variables in the variance model. My question is, what is the exact specification used to include the exogenous variables? In other words, how does the variance matrix, H(t) = C'C + A'u(t)u'(t)A + B'H(t-1)B, change after including these two explanatory variables?

Thanks,

[TomDoan]

(C + Ex(t))'(C+Ex(t)) + A'u(t)u'(t)A + B'H(t-1)B

That maintains positive definiteness and is invariant to different equivalent representations of dummy exogeneous variables.