No Steady State - Help!

Dear Johannes Pfeifer & All,

I am trying to replicate the results of Melina & Villa, “Fiscal Policy & Lending Relationships”. My model has 38 equations. The paper had listed all equilibrium conditions, the steady state equations and almost all parameter values (attached). So in a way the task at hand was only to write the equations and replicate results.

The only values assumed by my professor were:

y = 1.01 - output
rn = 1 - nominal interest rate
q = 1 - tobin’s q

To my best knowledge, the model has no error’s. Iv’e checked that many times. No matter what I do, I always get this error:

Error using print_info (line 74)
Impossible to find the steady state. Either the model doesn’t have a steady state,
there are an infinity of steady states, or the guess values are too far from the

Error in steady (line 92)
print_info(info,options_.noprint, options_);

Error in prac2 (line 464)

Error in dynare (line 180)

I know what the error means. Iv’e been reading dynare forums all day. I tried to compute all steady state values by hand and substituted them in the initval block. Upon running the code, I again got the same message.

I would be extremely thankful if someone could give me a way out? I have attached all files for your reference.
model2.mod (10.1 KB)
model1.mod (10.1 KB)
parameters.pdf (79.4 KB)
equilibrium_conditions.pdf (172 KB)

You should not normalize output to 1.01 unless you have a scaling factor for TFP. The nominal interest cannot be fixed a priori. It is a function of the discount factor that determines the real interest rate and steady state inflation, which must be assumed. Tobins’s q must also be assumed.

When you use


Dynare will find the steady state, but given how weird it looks there must still be problems.

Does having a scaling factor for TFP is the same as adding TFP in the production function and in all other places necessary? If I got that correctly?

Secondly, to which model file do you refer to which finds the steady state after the command steady(solve_algo=4,maxit=10000) ? without making any changes to my .mod file, if i used the steady with the option you suggested, It still gives the same error…

Yes, you got that correctly. The only other way to normalize y is to fix the labor disutility parameter to get exactly the amount of hours worked to get the desired output.

I used your above mod-file with the most recent Dynare unstable version and found the steady state.