Why the IRFs are so abnormal? Is it either normal or not?

Stochastic simulations
1

Hi everyone
I wanted to ask what the issue is with the IRFs, specifically whether it is a calibration problem or if the issue is more substantial.
Thanks much to help me.
modelthesisFRA.mod (19.7 KB)

2

Those IRFs indeed look weird. They look like oscillations due to a complex eigenvalue. It’s either a timing problem or a wrong parameter setting (often in feedback rules).

3

Thanks, I’ve just tried changing the values of both tau1 and tau2 (tau1=0.1 and tau2=0.2), and the oscillation problem has disappeared. However, the IRFs do not seem to return to the steady state values.
modelthesisFRA.mod (19.7 KB)
Could you take a look at the new IRFs to check their correctness?

4

That looks a lot better. The IRFs do return to the steady state, but they are very persistent. See also

5

Ok, thank you very much. Now I have another problem, which is the presence of indeterminacy. Basically, if I leave the current open settings (in particular 120 and 126), the model runs, but the IRFs (Impulse Response Functions) don’t seem perfect, as it appears that following a shock to the energy input price, inflation decreases (since the price of big firms good increases, but that of small-medium enterprises decreases, so I believe the problem lies with the latter). The issue is that if I use the same production function (line 124 and 126) for both firms (which is perhaps what I need to do), I get indeterminacy, but I don’t understand what it’s due to. Can you help me? It’s quite urgent.
Thanks in advance.

6

In my opinion the main point is to capture an increase of Psme_1/Psme_2/Pbf_1/Pbf_2 (maybe the marginal cost MC is important as well?).

7

I am getting a message about instabililty, not indeterminacy. Quite often the problem is a timing one, e.g. one where the timing of capital is wrong.

8

Yes sorry, the problem is about instability.
Where could the problem be that prevents me from opening both YSME and YBF production functions? I’m looking around but I don’t really know.
Thanks

9

As I said, it might be a timing problem. You may want to try the setup in a simpler model.