Hi,
I am new to the forum and still struggling to figure out how to post a new topic. But, I thought to continue this discussion on Recursive VAR since I could not find any suitable reply to this question. I am using EVIEWS9 and trying to replicate a Blanchard and Perotti (2002) paper published at QJE. The authors seem to have done Recursive VAR model.
Their model specifications are as follows:
The basic VAR specification is
(1) Yt=A(L,q)Yt-1+ Ut,
where Yt [Tt,Gt,Xt]' is a three-dimensional vector in the logarithms of quarterly taxes, spending, and GDP, all in real, per capita, terms.
Ut [tt ,gt,xt]' is the corresponding vector of reduced-form residuals, which in general will have nonzero cross correlations.
I have been able to solve for this VAR. Now the authors use the following identification methodology:
Without loss of generality, we can write
(2)tt, = a1 xt + a2 egt + et
(3) gt= b1 xt + b2 ett + egt
(4) xt = c1tt, + c2gt + et,
where et, ef , and et are the mutually uncorrelated structural shocks that we want to recover.
Can anyone please help us to specify this model in Eviews. I am struggling to incorporate this uncorrelated structural shocks in Eviews. I have able to get the residues from the VAR (eq 1) model. Any help on this will be really appreciated.
I am new to the forum and still struggling to figure out how to post a new topic. But, I thought to continue this discussion on Recursive VAR since I could not find any suitable reply to this question. I am using EVIEWS9 and trying to replicate a Blanchard and Perotti (2002) paper published at QJE. The authors seem to have done Recursive VAR model.
Their model specifications are as follows:
The basic VAR specification is
(1) Yt=A(L,q)Yt-1+ Ut,
where Yt [Tt,Gt,Xt]' is a three-dimensional vector in the logarithms of quarterly taxes, spending, and GDP, all in real, per capita, terms.
Ut [tt ,gt,xt]' is the corresponding vector of reduced-form residuals, which in general will have nonzero cross correlations.
I have been able to solve for this VAR. Now the authors use the following identification methodology:
Without loss of generality, we can write
(2)tt, = a1 xt + a2 egt + et
(3) gt= b1 xt + b2 ett + egt
(4) xt = c1tt, + c2gt + et,
where et, ef , and et are the mutually uncorrelated structural shocks that we want to recover.
Can anyone please help us to specify this model in Eviews. I am struggling to incorporate this uncorrelated structural shocks in Eviews. I have able to get the residues from the VAR (eq 1) model. Any help on this will be really appreciated.