# Dynare can solve only the log-linearized version of the model

Hi everybody,
I wrote a model in levels and Dynare can’t solve the model saying that Blanchard Kahn conditions are not satisfied: indeterminacy. However, when I wrote the same model in linearized version it worked. The two models are identical in terms of parameters, variables, etc. I solved the steady state of my model analytically and I used the same steady state values in my linear model wherever it needed (for example, in market clearing condition when you need shares). Can somebody give me a hint what could be the problem?

That typically means there is still a timing error (or some other mistake) in your nonlinear model.

Hi Professor,

Thank You for Your reply, I checked my model and found that the problem was arising because of one equation, namely:

[name = ‘Restricted household marginal utility’]
1/c_3 = mu_3*p_c_3;

When, I write it in linear version my model was solved, but I can’t understand why this is a problem. Can You give some intuition?

I have three types of household in my model and this is the marginal utility of rule of thumb household, who has no access to bond markets.

What is the full error message?

There are 31 eigenvalue(s) larger than 1 in modulus
for 32 forward-looking variable(s)

The rank condition ISN’T verified!

Error using print_info (line 45)
Blanchard Kahn conditions are not satisfied: indeterminacy
Error in stoch_simul (line 100)
print_info(info, options_.noprint, options_);
Error in TANK_DSGE_extended (line 1234)
info = stoch_simul(var_list_);
Error in dynare (line 235)
evalin(‘base’,fname) ;

When you introduce the linearized version of that equation, how do you link the linearized variables with the nonlinear ones in the other equations?

-c_3 + c_3ss = mu_3 - mu_3ss + p_c_3 - p_c_3ss;
where c_3ss mu_3ss and p_c_3ss are steady state values

But that is only correct if everything is already in logs.

So it is wrong equation if the rest of the model is in levels. But I still can’t understand why Dynare can solve the model only by changing this particular equation. What can be the possible explanation for this?

That equation will alter the steady state.

1 Like

Thanks, I understand what is my problem

1 Like