Money growth rule cause rank condition problem

In a NK-DSGE system, when I change the Taylor interest rule into a Taylor type money growth rule, it cause the rank condition problem.

How can I use a Taylor type money growth rule correctly?

PMAF0608.mod (6.9 KB)
Here’s my mode file

I get the same error when using the “normal” Taylor rule so for me the issue probably does not come from the money growth rule. Why is your Phi_Pi negative though? And having Phi_star and Y_star endogenous is also new to me. But maybe this is a feature of your model.
Can you make it run with a Taylor rule?

I use some value like -1 or 10 for test, still not work. The normal Taylor rule works well.

Most likely it’s related to

Thanks, how can I improve that? Is there any precedent for reference. Is the problem about the fiscal rule in my model? Thanks a lot

sorry to bother again, more specifically, how can I find the source of explosiveness in the model? Any paper I may need to read?
On this problem, it solved after I add government debt into the government purchase rule.

Often, the only way out is trial and error. Try to see whether you can find a parameter combination that works. If it happens after you add a fiscal rule, often the cause is the parameterization of the fiscal rule.

Thanks! I still have some questions, Is each “>1 eigenvalue” corresponding to a variable in model? And how may I find which variable is explosive(?) when B&K condition not satisfied?

Typically, the variables in the model are linked. The eigenvalues will normally affect all the variables. If debt is explosive, consumption will usually also explode.

Thanks a lot! And here’s another question… I am now replicating the result of Bhattarai et al(2016), it compared the log-data density of different model, and chose the maximum model as the best model. How can it work? What do log-data density means exactly in DSGE bayesian estimation?
2016 Policy Regimes, Policy Shifts, and U.S. Business Cycles.pdf (938.9 KB)

Have a look at the references in Model Comparison Bayesian Estimation (again)