Dynare forums

Hi,

It seems to me that also in Dynare 4.1 there is no pruning of the solution when computing IRFs.

If this is the case, it is a bit unfortunate. I understood from discussions on this forum that future developments of Dynare would
take account of this issue. (and the contribution by Stephan Fahr for second order solutions was a step in the right direction)

Is there any particular reason why Dynare developers have decided to neglect the problem of spurious explosive solutions when pruning is not implemented?

Best

Gianni
Please see the messages on this forum discussing pruning as well as Kim, Kim, Schaumburg and Sims (JEDC,2008) and Lombardo and Sutherland (JEDC, 2007) and the references to Judd therein.

Hi

Pruning is indeed planned for the future, but probably for Dynare 4.2.

Best

Hi,

Pruning is just part of the story. To compute IRFs in non-linear setups you need to use a generalized impulse response like in the papers by Pesaran and Potter. Be careful when reporting impulse responses!