The RATS Software Forum

The code below is used to compute 30 different GARCH models for each one of a sample of 55 companies. Is there any way I can create a matrix to display all coeffs (both mean and variance ones) and the respective stderrs, t-stats, p-values and %converged?

Open Data C:\Users\buddas\Documents\Data2.xls
Data(Format=xls,org=col)
Stats(noprint) rt;com nobs = %nobs; com step = 501

do j=1,nobs,step
Display 'Garch results'
SMPL j j+500

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,iters=500) / RT
# RMT DUMMY_2DAYS CONSTANT

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,iters=500) / RT
# RMT DUMMY_0_10 CONSTANT

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,xreg,iters=500) / RT
# RMT DUMMY_0_10 CONSTANT
# DUMMY_AFTER

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,xreg,iters=500) / RT
# RMT DUMMY_2DAYS CONSTANT
# DUMMY_AFTER

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,iters=500) / RT
# RMT DUMMY_0_2 CONSTANT

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,xreg,iters=500) / RT
# RMT DUMMY_0_2 CONSTANT
# DUMMY_AFTER

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,iters=500) / RT
# RMT DUMMY_0_5 CONSTANT

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,xreg,iters=500) / RT
# RMT DUMMY_0_5 CONSTANT
# DUMMY_AFTER

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,iters=500) / RT
# RMT DUMMY_10_0 CONSTANT

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,xreg,iters=500) / RT
# RMT DUMMY_10_0 CONSTANT
# DUMMY_AFTER

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,iters=500) / RT
# RMT DUMMY_10_10 CONSTANT

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,xreg,iters=500) / RT
# RMT DUMMY_10_10 CONSTANT
# DUMMY_AFTER

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,iters=500) / RT
# RMT DUMMY_1_0 CONSTANT

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,xreg,iters=500) / RT
# RMT DUMMY_1_0 CONSTANT
# DUMMY_AFTER

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,iters=500) / RT
# RMT DUMMY_1_1 CONSTANT

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,xreg,iters=500) / RT
# RMT DUMMY_1_1 CONSTANT
# DUMMY_AFTER

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,iters=500) / RT
# RMT DUMMY_2_0 CONSTANT

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,xreg,iters=500) / RT
# RMT DUMMY_2_0 CONSTANT
# DUMMY_AFTER

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,iters=500) / RT
# RMT DUMMY_2_2 CONSTANT

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,xreg,iters=500) / RT
# RMT DUMMY_2_2 CONSTANT
# DUMMY_AFTER

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,iters=500) / RT
# RMT DUMMY_5_0 CONSTANT

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,xreg,iters=500) / RT
# RMT DUMMY_5_0 CONSTANT
# DUMMY_AFTER

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,iters=500) / RT
# RMT DUMMY_5_5 CONSTANT

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,xreg,iters=500) / RT
# RMT DUMMY_5_5 CONSTANT
# DUMMY_AFTER

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,iters=500) / RT
# RMT DUMMY_ANNMT CONSTANT

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,xreg,iters=500) / RT
# RMT DUMMY_ANNMT CONSTANT
# DUMMY_AFTER

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,iters=500) / RT
# RMT DUMMY_PREVIOUS DUMMY_ANNMT CONSTANT

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,xreg,iters=500) / RT
# RMT DUMMY_PREVIOUS DUMMY_ANNMT CONSTANT
# DUMMY_AFTER

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,iters=500) / RT
# RMT DUMMY_PREVIOUS DUMMY_2DAYS CONSTANT

garch(p=1,q=1,pmethod=simplex,piters=20,method=bhhh,reg,xreg,iters=500) / RT
# RMT DUMMY_PREVIOUS DUMMY_2DAYS CONSTANT
# DUMMY_AFTER


end

Thanks in advance

Have you looked at REPORT with the REGRESS option? That's designed to organize output from various regressions (or other estimation instructions). Once the report is done you can either export it from the view on the screen, or export it directly using the UNIT option on REPORT(ACTION=SHOW,...)