unit roots in the small open economy DSGE

[litmuszhu]

Dear guys:
I try to reproduce the results of an small open economy DSGE. Though there are some error in the mod file, there are always 4 unit roots. What differentiates this model is that it contain some foreign price shocks. I see most open economy DSGE contain real shocks, like technological shocks or real demand shocks. I wander if the unit roots problems are caused by the structure of exogenous shocks. So I cut some price shocks. It really reduces unit roots. But I can not cut them all. So I have confidence that the original model on the paper has unit roots.My questions are: 1, Does the unit roots problem matter? 2.Is it possible to use a DSGE model with unit roots to do policy analysis?
Thank you!

[jpfeifer]

The unit roots could come from equations like

Code: Select all

PiH=e-e(-1)+PiHF;
QC=QC(-1)+PiT-PiN;


They basically look like unit root processes and I wonder if that is desired.

If the unit roots are desired, they pose no problem unless you want to estimate or simulate the model. For IRFs at order=1, this is no problem. See also:
http://www.dynare.org/phpBB3/viewtopic.php?f=1&t=5982