VAR Heteroskedasticity-consistent standard errors

Questions and discussions on Vector Autoregressions
comac
Posts: 25
Joined: Mon Jun 29, 2009 7:31 am

VAR Heteroskedasticity-consistent standard errors

Unread post by comac »

Dear all,
is there some code for computing heteroskedasticity-consistent standard errors for a VAR system?

Thanks in advance,
C.
Last edited by comac on Fri Sep 10, 2010 1:32 am, edited 1 time in total.
TomDoan
Posts: 7814
Joined: Wed Nov 01, 2006 4:36 pm

Re: VAR Hetreskodasticity-consistent standard errors

Unread post by TomDoan »

Use SUR with the ROBUSTERRORS option as is shown below:

Code: Select all

system(model=varmodel)
variables dinv dinc dcons
lags 1 2
det constant
end(system)
estimate(sigma,resids=resids) 
sur(model=varmodel,robust)
ying2728
Posts: 5
Joined: Thu Nov 01, 2012 6:17 pm

Re: VAR Heteroskedasticity-consistent standard errors

Unread post by ying2728 »

In the following SUR:

system(model=varmodel)
variables dinv dinc dcons
lags 1 2
det constant
end(system)
estimate(sigma,resids=resids)
sur(model=varmodel,robust)

If each equation uses different different indep var: e.g.,
dinv = constant dinv(-1)
dinc = constant dinc(-1)
dcons = constant dcons(-1)

How can I do it? Thanks.
TomDoan
Posts: 7814
Joined: Wed Nov 01, 2006 4:36 pm

Re: VAR Heteroskedasticity-consistent standard errors

Unread post by TomDoan »

Isn't that what SUR.RPF does? In general, SUR is used when the equations aren't the same.
Post Reply