Theoretical Moments_ NaN


I am solving a multi industry model with a monetary and a macroprudential rule. When I run the model I do not get any particular error but the Theoretical Moments show up as NaN.
Model diagnostics detects a colinear relationship between the productivity shock and the macroprudential rule, and suggests there maybe a redundant equation. However, I have four optimizing agents and I am excluding one of the four budget constraints, therefore I was not expecting to have issues concerning the Walras’ Law.

Thanks in advance for your help

Macroprudential_model_100521.mod (9.4 KB)
SS_100521.m (4.1 KB)

You did not provide the mat-file you are trying to load.


by running the m-file the mat-file is created. Sorry for the confusion.


I am getting that your a1 and A1ss are complex numbers.

I am not sure why is that. If I run the .m file I get A1ss = 0.9132.

MM_120521.mod (9.4 KB)
SS_120521.m (4.4 KB)

I think now it should be possible to run it (by running the .m file SS_120521.m).

In the model I have four optimizing agents among which two are savers and two are borrowers. The problem seems to be a collinear relationship between the two Euler Equations of the savers. I understand that since they have the same discount factor two Euler Equations may seem redundant, but I thought that since consumption of the four agents is different, there must be four Euler Equations anyway, correct?

Thanks in advance.

I think this could potentially cause a problem. For example, if I have two optimizing agents with the same discount in their Euler (presumably the Euler originates from optimizing over a budget constraint) and an equation which pins down the aggregate objective variable (consumption, for example), how do I pin down the steady-state value of their respective objective variables?

I believe this is the reason why Iacoviello chooses different discount factors for all of the agents in his Financial Business Cycles paper.

My guess is similar and has been discussed in

(and many other posts)

Thanks to both. A similar model with only one agent saving works correctly indeed.

That is expected. But with two agents, they can provide each other with credit after shocks. With the permanent income hypothesis, the effects will be permanent.