I am working with a New Keynesian DSGE with third order approximation. The problem is that the IRFs generated by the model are fluctuating wildly leading to problems in interpretation. For example, labor whose steady state value is 0.31 is exhibiting values ranging from -300 to +300 and so on. Similarly output whose steady state value is 1 is varying between -1000 to +1000. My model is expressed in levels (not in exp form). I tried by reducing the size of the shock but it did not help.
Can anyone suggest what am I missing here. I am attaching the mod file.
Stevedem_2020_wthWageRigandHP.mod (9.4 KB)
You should check your model. Even at
order=1, some of the IRFs are already extremely big.
Yes I am aware Professor. Is there a trick to start looking for the loopholes?
No really. You can play around with the parameters and see what this changes. That should give an indication which part of the model may be the cause.
You are right! The model behaves abnormally at some parameter values of the price stickiness parameter, interest rate smoothing parameter and interest rate response of inflation in the Taylor rule. Does that mean my model is mis-specified? What do I conclude from this?
It is at least an indication that there may be a mistake related to those features.