Re: st: RE: SVAR: Return code 498

[Charles Koss <hqtiger@gmail.com>]

Thank you Dr. de Souza, the code worked, I appreciated it!!! In
summary, this is what can be concluded:

Stata can produce the standard errors for irfs after svar only  in two
cases: parametric bootstrap and asymptotic standard errors. When the
option of bootstrapped residuals is used, the described error message
appears.

Add these lines to the previous code and shall see the error message
irf create order1,                        step(8)
irf create order2,  bsp reps(50) step(8)
irf create order3,  bs   reps(50) step(8)

Therefore, it seems that there is a bug with the bs option. Opinions
or comments?

Charles

-- 
Charles Koss
http://charlesonnet.blogspot.com

On Mon, May 9, 2011 at 9:27 AM, DE SOUZA Eric
<eric.de_souza@coleurope.eu> wrote:
> The instructions below work. The model is over-identified.
>
>
> Eric de Souza
> College of Europe
> Brugge (Bruges), Belgium
> http://www.coleurope.eu
>
>
> set more off
> log using irf3.log
> use http://www.stata-press.com/data/r11/lutkepohl2
> matrix A = (1,0,0\.,1,0\0,.,1)
> matrix B = (.,0,0\0,.,0\0,0,.)
> svar dln_inv dln_inc dln_consump if qtr<=tq(1978q4), aeq(A) beq(B) var dfk
> matrix Aest = e(A)
> matrix Best = e(B)
> matrix chol_est = inv(Aest)*Best
> matrix list chol_est
> matrix sig_var = e(Sigma)
> matrix chol_var = cholesky(sig_var)
> matrix list chol_var
> irf create aqmch11irf_2, set(aqmch11)  step(10)
> irf graph sirf, impulse(dln_inc) response(dln_consump)
> graph export aqmch11irf_1.emf
> des
> log close
> set more on
>
> -----Original Message-----
> From: owner-statalist@hsphsun2.harvard.edu [mailto:owner-statalist@hsphsun2.harvard.edu] On Behalf Of Charles Koss
> Sent: 09 May 2011 16:08
> To: Stata List
> Subject: st: SVAR: Return code 498
>
> Dear list members, I am using the command SVAR, but when I calculate the IRFs I obtained an error message as follows:
> predict may not be used after fitting an overidentified SVAR model to get these statistics for the underlying VAR, fit it with var, then run predict r(498);
>
> This is the code I am using:
>
> clear
> set more off
> use http://www.stata-press.com/data/r11/urates
> matrix A = (1,0,0\.,1,0\.,0,1)
> matrix B = (.,0,0\0,.,0\0,0,.)
> svar missouri indiana kentucky, lags(1/4) aeq(A) beq(B) set seed 05062011 //compare irfs with rats irf create order1, set(jmulti_gretl_stata)  step(16) bs reps(50)
>
> I wonder, why the error message appears? When the system is just identified the irfs are calculated. It seems that the irfs can not be calculated when the system is overidentified. As such, it seems to me that the estimation of the irfs do not take advantage of short run restrictions when the system is just identified. Any commenst on this issue? Your help will be appreciated.
>
> Thank you,
>
> Charles
>
> --
> Charles Koss
> http://charlesonnet.blogspot.com
> *
> *   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/
>
> *
> *   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/
>

*
*   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/