Question trying to use OccBin: A Toolkit for Solving Dynamic

[dmv]

Hi.

I am trying to replicate one of the models in this paper https://ideas.repec.org/a/eee/moneco/v70y2015icp22-38.html in particular the New Keynesian Model Subject to the Zero Lower Bound (I attach the .mod file with the reference model (where the ZLB is not binding)). Nevertheless, in the paper the discount factor's law of motion is log(ut)=(1-RHO)*BETA+RH0*log(ut(-1))+ SIGMA*eps_ut while in the code:

1) it is just log(ut) = RHO_U*log(ut(-1))+eps_u
2)there is another law of motion and another error term "at" (???)

Why is there a difference? What does it mean? Or is it that the code posted corresponds to a different model?

I am trying to adapt this code to use a higher order perturbation method, and would appreciate your help.

[jpfeifer]

That is equivalent. In

Code: Select all

log(ut)=(1-RHO)*BETA+RH0*log(ut(-1))+ SIGMA*eps_ut


the ut has mean exp(BETA). This can be rewritten with ut having mean 1 and simply multiplying BETA so that
BETA*ut with

Code: Select all

log(ut) = RHO_U*log(ut(-1))+eps_u ;


follows exactly the described process