Unit root issue

I’m trying to solve the attached model but I have a unit root issue.
The model is a very standard 2-country open economy endowment model (shocks are specified as news shocks).

The model works fine in the version of complete markets and financial autarky (see attached files, respectively “newsCM” and “newsFA”).
The problem is with the bond economy (since it introduces a forward-looking variable the BK conditions are not satisfied anymore).

Does anyone know how to deal with this please?
Many thanks in advance!

newscm.txt (3.59 KB)
newsfa.txt (3.54 KB)
newsbe.txt (3.77 KB)

Hi guys!
Sorry to bother again… Any idea on this issue please?

Have you read Schmitt-Grohe and Uribe (2003) “Closing small open economy models”? The answer is there.

No, the answer won’t be there as SGU deals with small open economies and the initial post explicitly states that it is a two-country model. Are you sure you obeyed Dynare’s timing convention for predetermined states, i.e. the bonds? See the manual.

Actually, the code is well specified. In fact it works in the other versions (complete markets and financial autarky).
Not sure why it doesn’t in a bond economy…

Hi, the unit root problem is present in any bond economy (incomplete markets), irrespective of whether it is an SOE or a 2C model. did you try using a cost function for bond accumulation or an endogenous discount factor?