Dynare forums

Hi there,

I was wondering whether it is important to do Cholesky decomposition of the variance-covariance matrix when we do historical decomposition.

Thanks for your taking your time to answer.

Cheers,
Qazi

Dear Quazi,
yes, if you have correlated structural shocks, the covariance matrix needs to be decomposed to assign the effect on the endogenous variables to a particular shock. Dynare will use a Cholesk decomposition for this purpose.