Structural VAR

[jvillama]

Hello,

I was hoping to get some guidance from the more knowledgable people in here.
For my dissertation I am looking to measure the dependance of Credit in Latin America to external factors.
Consequently I want to construct a SVAR to eventually measure impulse response functions.
I am going to run a SVAR for each of the countries i will cover and I want to have 4 Domestic factors and 2 external factors (Interest Rates in US and a commodity index) as my independent variables.
I however want to control for endogeneity, since it would be ludicrous to have the deposits in one of my LA country affect IRs in US or the commodity index.
What is the best way to go about this and can it be done in eviews?

Thank you so much for reading and responding. My sincere gratitude

JPV

[dakila]

If I understood your question correctly, in SVAR all variables are endogenous unless you specified some variables like foreign factors exogenously.

[trubador]

You can either define 4 domestic factors as endogenous variables and treat the other two as exogenous variables as Dakila suggests (along with the constant), or make a 6-variable VAR and then orthogonalize the error terms (with structural factorization or a simple Cholesky decomposition) to make sure that those two external factors have no feedback from the remaining domestic factors.