Dear professor,

I am doing an estimation but before that I perform the identification and sensitivity analysis with observable variables (filtered variables). The identification test shows that both matrices J and H are full rank and all the parameters are identified. However, the sensitivity graph does not show any bar on the graph.

on the other hand, when i go ahead to trigger the estimation, it got singularity problem with the same observable.

What might be the cause of the problem?

Did you try the unstable version? It is more informative when the prior mean is 0 so that the normalization of the bar graph cannot be computed.

Identification and stochastic singularity are two completely separate issue. One is not informative of the other.

Dear professor,

After fixing the issue of identification and performing the estimation of the set of parameters identifiable by the observable, I got a weird Bayesian IRF. In fact, the IRF changes sign for some variables . Specifically, the IRF of output to shock on government consumption spending turns negative and thus the fiscal multiplier.

Also, I got a very huge estimated std value for a shock mainly shock on labor income tax in the model.

I wonder how these unexpected outcome came about.

Finally, the mode check graphs of parameter got few “red dots” for some parameters.

I would appreciate any hint to fix all these issues about the model

Sorry, but that is impossible to tell without knowing the model in detail. I would try to identify which parameter values drive this behavior.