NaN values in deterministic simulation

[mricci]

Dear all,
I am working for my master thesis to a simple RBC model with government.

The purpose of the model, which is calibrated for Italy, is to produce a deterministic simulation of 20 periods (from 1989 to 2009) where exogenous variable such as Gov Consumption, Gov. Transfers, Tax rates are feeded in with their historical values and to study the trend of the main endogenous variable which is government debt. The scope is to see how well the model-produced government debt time series fits the (actual) time series of the Italian government debt.

My problem is that, when I launch the code in Dynare (4.3.3), I face the following situation: the program is able to calculate the SS, the rank condition is satisfied and the convergence is obtained. But apart from the SS, the values for all the endogenous like “y” (output), “b” (government debt) along the simulation are NaN (except for the first and last values which reflect the initial and terminal conditions). I don’t understand why this is happening neither I have found something analogous in the other posts.

I attach in the following my .mod file. I would be extremely grateful, if anyone could help me.

Mattia

[jpfeifer]

Dynare 4.3 unfortunately does not provide an explicit error message if your deterministic simulation did not converge. More recent versions say:

Simulation terminated after 9 iterations.
Total time of simulation : 3.563
Error using sim1 (line 144)
Simulation terminated with NaN or Inf in the residuals or endogenous variables. There is most likely something wrong with your
model.

Thus, Dynare was not successful in the solution of your model. As the Blanchard-Kahn conditions only tell you something about local saddle-path stability, you might want to first check the performance of your model for only small deviations from steady state and see if it works. Only then transition to the full-blown model.

[mricci]

Dear Pfeifer,
Thank you very much for you answer and suggestion

Best,

Mattia