MV-GARCH Diagnostics
Posted: Fri Jun 03, 2011 8:48 am
Dear Tom,
I am trying to get some diagnostics for the models I estimated. Multivariate Q statistics works but I am getting an error while computing standardized residuals. It says
## MAT15. Subscripts Too Large or Non-Positive
Here is my code. Thanks in advance.
I am trying to get some diagnostics for the models I estimated. Multivariate Q statistics works but I am getting an error while computing standardized residuals. It says
## MAT15. Subscripts Too Large or Non-Positive
Here is my code. Thanks in advance.
Code: Select all
EQUATION C R_C
# CONSTANT R1_C R2_C R3_C R4_C R5_C Mon
EQUATION SM R_SM
# CONSTANT R1_SM R2_SM R3_SM R4_SM R5_SM Mon
EQUATION LH R_LH
# CONSTANT R1_LH R2_LH R3_LH R4_LH R5_LH Mon
GROUP MEANM C SM LH
GARCH(P=1,Q=1,MODEL=MEANM,MV=bekk,XREGRESSORS,METHOD=BFGS,PMETHOD=SIMPLEX,PITERS=15, HMATRICES=H, MVHSERIES=MVH, RVECTORS=U,ITERATIONS=1000, subiterations=50, CVCRIT=0.001)
# APP cattle COF CP FO GS HP LDPO OCO wasde_mix wasde_pure $
MON JAN FEB MAR APR MAY JUN JUL AUG SEP OCT NOV
* standardized residuals
set z1 = U(t)(1)/sqrt(H(t)(1,1))
set z2 = U(t)(2)/sqrt(H(t)(2,2))
set z3 = U(t)(3)/sqrt(H(t)(3,3))
@bdindtests(number=40) z1
@bdindtests(number=40) z2
@bdindtests(number=40) z3
* Multivariate Q statistic
dec vect[series] zu(%nvar)
do time=%regstart(), %regend()
compute %pt(zu, time,%solve(%decomp(H(time)),U(time)))
end do time
@mvqstat(lags=40)
# zu