SVAR just/exactly identified short-run
Hello, I would like some assistance with the interpretation of an exactly identified SVAR model. I am working on a bivariate model with the variables fiscal deficit and public liabilities, in which the restriction is imposed that public liabilities do not affect the fiscal deficit.
The results I obtain are the following. I am unsure about the interpretation of the coefficient a_21 which is negative. Because in the Stata SVAR manual (tsvarsvar is the name), in example 1 of a just-identified model, it states:The estimates of /A:2 1, /A:3 1, and /A:3 2 are all negative. Because the off-diagonal elements of the A matrix contain the negative of the actual contemporaneous effects, the estimated effects are positive, as expected.
Therefore, the contemporaneous effect would be negative. For me, this is crucial because the sign determines whether to establish a fiscal dominance regime (if the sign is positive) or simply to deny that there is statistical evidence to verify it.
Matrix:
. matrix A = (1,0\.,1)
. matrix B = (.,0\0,.)
Results:
svar d.DEFISCAL_PIB d.PASIVO_PIB, exog(BCOMERCIAL_PIB CONSUMO_PIB) lags(1) aeq(A) beq(B)
Sample: 2002q4 - 2022q4 Number of obs = 81
Exactly identified model Log likelihood = -472.1765
------------------------------------------------------------------------------
| Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
/a_1_1 | 1 (constrained)
/a_2_1 | -.7916936 .1204449 -6.57 0.000 -1.027761 -.555626
/a_1_2 | 0 (constrained)
/a_2_2 | 1 (constrained)
-------------+----------------------------------------------------------------
/b_1_1 | 4.2862 .3367556 12.73 0.000 3.626171 4.946228
/b_2_1 | 0 (constrained)
/b_1_2 | 0 (constrained)
/b_2_2 | 4.646256 .3650443 12.73 0.000 3.930782 5.36173
------------------------------------------------------------------------------
Just to clarify: a_2_1 is -0.7916936 and variables are expressed as GDP %.
Thanks a lot!
The results I obtain are the following. I am unsure about the interpretation of the coefficient a_21 which is negative. Because in the Stata SVAR manual (tsvarsvar is the name), in example 1 of a just-identified model, it states:The estimates of /A:2 1, /A:3 1, and /A:3 2 are all negative. Because the off-diagonal elements of the A matrix contain the negative of the actual contemporaneous effects, the estimated effects are positive, as expected.
Therefore, the contemporaneous effect would be negative. For me, this is crucial because the sign determines whether to establish a fiscal dominance regime (if the sign is positive) or simply to deny that there is statistical evidence to verify it.
Matrix:
. matrix A = (1,0\.,1)
. matrix B = (.,0\0,.)
Results:
svar d.DEFISCAL_PIB d.PASIVO_PIB, exog(BCOMERCIAL_PIB CONSUMO_PIB) lags(1) aeq(A) beq(B)
Sample: 2002q4 - 2022q4 Number of obs = 81
Exactly identified model Log likelihood = -472.1765
------------------------------------------------------------------------------
| Coef. Std. Err. z P>|z| [95% Conf. Interval]
-------------+----------------------------------------------------------------
/a_1_1 | 1 (constrained)
/a_2_1 | -.7916936 .1204449 -6.57 0.000 -1.027761 -.555626
/a_1_2 | 0 (constrained)
/a_2_2 | 1 (constrained)
-------------+----------------------------------------------------------------
/b_1_1 | 4.2862 .3367556 12.73 0.000 3.626171 4.946228
/b_2_1 | 0 (constrained)
/b_1_2 | 0 (constrained)
/b_2_2 | 4.646256 .3650443 12.73 0.000 3.930782 5.36173
------------------------------------------------------------------------------
Just to clarify: a_2_1 is -0.7916936 and variables are expressed as GDP %.
Thanks a lot!
