VARMA GARCH Model

[Bella]

Dear Tom

When I run the code about a VARMA(2,2)-GARCH-BEKK, I get different results if I change the order of variables. How is it caused?
It's like
system(model=varmah)
variables mb zjz
lags 1 2
det constant eps(1){1} eps(2){1} eps(1){2} eps(2){2} sqrth(1) sqrth(2)
end(system)

But if I change the code into
system(model=varmah)
variables zjz mb
lags 1 2
det constant eps(1){1} eps(2){1} eps(1){2} eps(2){2} sqrth(1) sqrth(2)
end(system)

I get different estimated coefficients.
It is the code and data enclosed.

[TomDoan]

Note--you don't get different estimated coefficients, you get no estimated coefficients. Neither model converges.

Is there a reason that you're trying to estimate a VARMA(2,2)-Asymmetric-BEKK-GARCH-M model with this data set? You only have 168 data points, which is nowhere near enough to estimate such a complicated model, and neither of your time series has very convincing evidence of GARCH effects.

[Bella]

Thank you Tom. I think you are right.