The RATS Software Forum

Hi,

I wonder wheter it makes sense using %psdfactor to change the order in a BQ decomposition, that is, something like:

bqfactor = Phi(1)*%psdfactor(inv(Phi(1))*Sigma*tr(inv(Phi(1))),||1,3,5,2,4||)




Thanks in Advance fro the help,

Rodolfo

You can, though I would recommend using @ShortAndLong. It's hard to come up with an (economically) reasonable set of long-run only restrictions to identify a five variable system.

TomDoan wrote:You can, though I would recommend using @ShortAndLong. It's hard to come up with an (economically) reasonable set of long-run only restrictions to identify a five variable system.

Great, thank you !