Blanchard Kahn violations when using OLR

[dbeltran@ucsc.edu]

Hi Michel,

Thanks for your patience in trying your best to reply to all these emails, and for providing such a wonderful software for economists.

I'm trying to use the OLR function. Dynare calculates the steady state and performs the stochastic simulations for my model just fine. When I try to compute the optimal policy using olr, it crashes saying the Blanchard-Kahn conditions are not satisfied (indeterminacy). OSR seems to work fine, but not OLR. I am sending you the dynare .mod file for Monacelli's SOE model extended for habit formation which I am using in my dissertation. The instrument I chose for olr was the interest rate "int." I then removed the monetary policy rule equation when running olr.

//int=rho_i*int(-1)+(1-rho_i)*(psi_pi*pie+psi_y_tilda*y_gap)+ei;

I tried other policy instruments and removing other equations, but this did not solve the problem. Could this problem be due to the initialization of the optimization algorithm?

If you have any clues for me that would be great. The .mod file is attached.

Thank you so much.

Best,
Daniel Beltran
Univ. of CA, Santa Cruz

[MichelJuillard]

Dear Daniel,

thanks for the message. I will have a look. There is no theory that I'm aware of that will tell you that OLR solution will always satisfy BK conditions. I have examples where a simple rule will be uniquely stable, but not the OLR solution

Kind regards


Michel

[dbeltran@ucsc.edu]

Hi Michel,

Thanks for looking into this. I'll also try using different parameter values and hopefully it will work for at least one parameter combination.

Best,
Daniel