Cholesky decomposition in VARs and SVARs
Posted: Thu May 21, 2015 1:21 am
For our thesis, we are estimating a SVAR model. However, we wonder what is the difference between a VAR and a SVAR IRF and variance decomposition? We've understood that the Cholesky ordering is a "solution" to the lack of theory wrt variable ordering, and can be used both in VARs and SVARs by specifying it as a restriction matrix. But, however, when interpreting IRFs and VDs with both methods, are we correct if we specify VAR results as shocks whereas SVAR results as structural shocks?