Steady States calculation (growth model)s

Stochastic simulations
1

I have a quick question. As far as I understand, Dynare interprets all of the underlying variables as logs.
For example, if one of my equations in the model is y_{t} = ak_{t}^{\alpha}n_{t}^{1-\alpha}, in Dynare I have to write the code as exp(y) = exp(a) * exp(k(-1))^(alpha) * exp(n)^(1-alpha).

Here, one of my equations takes the form of
y_{t} = exp(z_{t})k_{t}^{\alpha}n_{t}^{1-\alpha}.

Should I write the code as exp(y) = exp(exp(z)) * exp(k(-1))^(alpha) * exp(n)^(1-alpha)? I’ve tried it but it said there is an error, so I was wondering whether the error comes from this part. Thank you!

2

  1. No, Dynare by default does linearization, not log-linearizations. That’s why you need to do something if you want to have logs.

  2. If you have exp(z), then z is already in logs. There is no point in adding another exp.

  3. Generally, I recommend appending auxiliary equations with logged variables instead of doing a full exp-substitution. See Question about understanding irfs in dynare - #4 by jpfeifer