same model but different IRFs amplitude

[kam]

Hello,

I need some help about those two questions please:

Is it problematic that my residuals are different from 0 ?

I just turn my model, but my IRFs do not have the same amplitude (are really far) than the paper in wich I work...

any suggestions...

Thanks.

Residuals of the static equations:

Equation number 1 : -4.0213e-007
Equation number 2 : -3.4829e-006
Equation number 3 : -1.3097e-006
Equation number 4 : -2.2163e-006
Equation number 5 : 0
Equation number 6 : 2.5e-007
Equation number 7 : 0
Equation number 8 : 3.3088e-006
Equation number 9 : -1.7534e-007
Equation number 10 : -3.0147e-006
Equation number 11 : -2.5e-007
Equation number 12 : -2.5165e-006
Equation number 13 : 0
Equation number 14 : 0
Equation number 15 : -3.7419e-006
Equation number 16 : -3.7051e-006
Equation number 17 : -8.8775e-006
Equation number 18 : -7.2369e-006
Equation number 19 : -2.5e-006

[jpfeifer]

The small residuals might be harmless unless they signify a larger problem. If your IRFs are different from a paper you are trying to replicate, there is most probably something wrong with your code.

[kam]

Thank you for your reply, I think I fixed the residuals equations problem, almost all of them equal zero now...but the IRF's problem still, I verified my code again and again, all seem right.

Equation number 1 : -0.042837
Equation number 2 : -0.00042837
Equation number 3 : 0
Equation number 4 : 0
Equation number 5 : 0
Equation number 6 : 0
Equation number 7 : 0
Equation number 8 : 0
Equation number 9 : 0
Equation number 10 : 0
Equation number 11 : 0
Equation number 12 : 0
Equation number 13 : 0
Equation number 14 : 0
Equation number 15 : 0
Equation number 16 : 0
Equation number 17 : 0
Equation number 18 : 0
Equation number 19 : 0

[kam]

here is my graphs and the graphs of the papers.

[jpfeifer]

Almost is most probably not sufficient. Qualitatively, the IRFs look similar. Are you sure that you have the correct shock size (and multiply with 100 if the IRFs are in percent)?

[kam]

yes, I'm sure. it's one percent monetary policy shock (std dev=0.01) and when I change it to a value greater than 0.04, I have no IRFs (NaN).
I do not know if that's reveal something.
Thx.
kam.

[jpfeifer]

This means that you should put order=1 into stoch_simul. Also note that a one-percent federal funds rate shock in a quarterly model implies a shock of 0.0025 or 0.25%.

[kam]

thank you.
Now, my IRFs are the closest to the paper, but I still have a problem with the investment that is ten times lower. In addition, I have to evaluate the effect of trend inflation on stochastic means (deterministic steady state vs stochastic steady state), so I have to use the second order.

kam

[jpfeifer]

Then use the pruning option and check the model parts (and calibration) relating to investment

[kam]

HI again,

what did you mean by the pruning option?

I looked in the manual, but I did not find...

kam

[jpfeifer]

You need to use at least 4.3.3. There you can find it in the manual.