Dynare forums

Hi, I've always solved this deterministic model with zero lower bound on nominal interest rates because it is immediate to write this constraint and the impulse responses are exactly as I expect. However, by removing the ZLB (first equation), the impulse responses are really weird, even if the model solves correctly (I suppose). Why does it happen ? Thanks

Ok, I think I've got it: since the nom.int.rate is not constrained by the ZLB, the central bank can reduce it to fully accomodate the negative demand shock. That's why I see that gdp doesn't fall. This should be the explanation.

Hi, one more question: I'm removing the shock to d and leaving only that to eta. There is something I don't understand: I have one sequence with eta=0.2 and a second sequence with eta=0.202. If I set the shock in periods, say, 4:6, why do the second sequence changes from period 1 ?

I don't understand the question. Please provide an example file.

Ok. see the attached file gshock1. As you can see at line 74, I have commented out the shock to d. However, at line 109, I set a shock to eta from period 6 to 8. Now take a look at the 2 sequences generated, geta and gdpeta. You will see that both gdp and g vary from the first period in the second sequence, the one where eta=0.202. Why ?

You are simulating under perfect foresight. Agents will know that a shock happens at time 4 and will of course react to that today. Do you mean this?

Yes, you are right, I forgot agents have perfect foresight. With regard to the first question about dynamics without zero lower bound, am I right ? Thanks

That could be, but without knowing your model, it is impossible to tell for sure.