The RATS Software Forum

Dear all,

Am I right in conducting multivariate VAR(1)-GARCH(1,1)-M model (4 endogeneous variables) as following:

system(model=garchm)
variables x1 x2 x3 x4
lags 1
det constant
end(system)

dec symm[series] hhs(4,4)
clear(zeros) hhs
*
equation eq1 x1
# constant hhs(1,1) hhs(1,2) hhs(1,3) hhs(1,4)
equation eq2 x2
# constant hhs(2,1) hhs(2,2) hhs(2,3) hhs(2,4)
equation eq3 x3
# constant hhs(3,1) hhs(3,2) hhs(3,3) hhs(3,4)
equation eq4 x4
# constant hhs(4,1) hhs(4,2) hhs(4,3) hhs(4,4)

group garchm eq1 eq2 eq3 eq4
garch(model=garchm,p=1,q=1,mv=diag,pmethod=simplex,piters=10,mvhseries=hhs,iterations=500)

Any reply will be appreciated.

Owen Wang

That's set up correctly, though I don't think you want MV=DIAG.