Failure to compute steady state using initval

Hi Prof,@jpfeifer

I am trying to replicate the basic model in Iacoviello(2005) and introduce shocks to housing demand, productivity and monetary policy. I am using initval to compute the steady state but I keep getting the NaN error. See error message below:

warning: Some of the parameters have no value (epsilon) when using steady. If these parameters are not initialized in a steadystate file steady_state_model-block, Dynare may not be able to solve the model…
Randomize initial guess…

Residuals of the static equations:

Equation number 1 : NaN : 1
Equation number 2 : NaN : 2
Equation number 3 : NaN : 3
Equation number 4 : NaN : 4
Equation number 5 : NaN : 5
Equation number 6 : NaN : w_t
Equation number 7 : NaN : y_t
Equation number 8 : NaN : 8
Equation number 9 : NaN : 9
Equation number 10 : NaN : r_t
Equation number 11 : NaN : pi_t
Equation number 12 : NaN : a_t
Equation number 13 : NaN : j_t
Equation number 14 : NaN : eta_t
Equation number 15 : NaN : 15
Equation number 16 : NaN : l1_t
Equation number 17 : NaN : 17
Equation number 18 : NaN : 18

error: The steady state has NaNs or Inf.
error: called from
print_info at line 32 column 5
steady at line 102 column 5
driver at line 326 column 1
dynare at line 293 column 1

Attached is my mod file, any assistance will be highly appreciated.

prac.mod (1.9 KB)

Your model is not linearized. Why do you initialize it with 0?

Hi Prof,
I am using a nonlinear model. I used initial values of zero because if one doesn’t explicitly specify a variable then dynare normally uses zero as an initial value when computing steady states. But kindly correct me if my reasoning is flawed.
Thank you.

As documented on many occasions that will not work with variables that must not be 0. Take consumption. How is 1/c supposed to work if c=0?

I understand.
Would it be better to work with a linear model in this case?

No, there the same issue would appear: linearization requires you to have the steady states of the nonlinear model as parameter values.

Thank you for your responses.

Could you kindly make suggestions on how to resolve the issue?

Set the initial values based on economic intuition. See Remark 15 (initval vs. steady_state_model vs. steadystate-file) in Pfeifer (2013): “A Guide to Specifying Observation Equations for the Estimation of DSGE Models ”.

Thank you very much.