IRF graph seem not normal

Hello Prof @jpfeifer

I get these graph from running Bayesian estimation. I think something goes wrong. The behavior of these IRF does not seem normal. Am i right ? What do you suggest if it is the case?

Oscillating IRFs are caused by complex eigenvalues, usually due to timing errors or implausible parameter values in feedback rules like debt feedback. Does a calibrated version of the model work as expected?

The calibrated model give also the same behavior. As you say, the model has debt equation and there maybe some parameters value that are not consistent. It is quite tricky to solve the timing error issue. Maybe you can take a look at it

You should always make sure the calibrated version of the model works as expected. Often, there is a fundamental timing issue that can be “solved” by choosing implausible parameter values.

Thanks a lot. I finally solved it