Re: st: SVAR estimation with Stata

[Nick Cox <njcoxstata@gmail.com>]

This seems to overlap with a question asked and answered earlier.

Nick

On Tue, May 3, 2011 at 10:25 PM,  <ting.wang@yale.edu> wrote:
> Dear all,
>
> I am running Structural VAR model with Stata. Suppose the contemporaneous
> coefficient matrix is G0, that is,
>
> e(t)=G0*u(t),
>
> where e(t) is the structural error and u(t) is the error of the reduced form
> VAR.
>
> Here is the Stata codes:
>
> matrix A (exactly the G0 matrix)
> =(1,.,0,0,.,0,.\.,1,.,.,0,0,0\0,0,1,.,.,0,0\0,0,0,1,.,0,0\0,0,0,0,1,0,0\0,0,0,0,.,1,0\.,.,.,.,.,.,1)
> matrix B
> =(1,0,0,0,0,0,0\0,1,0,0,0,0,0\0,0,1,0,0,0,0\0,0,0,1,0,0,0\0,0,0,0,1,0,0\0,0,0,0,0,1,0\0,0,0,0,0,0,1)
>
> svar `varlist', exog(`dum') lags(1/6) aeq(A) beq(B)
> {`dum' refers to the 11 dummy variables treated as exogenous}
>
> Now create the Impulse Response Function:
>
> irf create model2, set(myirf, replace) step(8)
> irf table fevd, noci {noci is to suppress confidence bands}
> irf graph oirf, impulse(r) response(cpi) {orthogonalized IRF}
> irf graph sirf, impulse(r) response(cpi) {structural IRF}
>
>
> Questions:
> 1. Is it the correct way to involve "lags" to the estimation?
> 2. The convergence is not achieved, any possible reasons for that?
> 3. When I create the IRF, Stata was stuck and could not move on. Does anybody
> know how this might come across?
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/