Nonstationary variables in dynare

[Jaime_Fierro]

I am estimating potential output. The model runs fine if I write potential output and GDP in first differences; however, it seems dynare does not accept nonstationary variables such as potential output and GDP in the model because the model does not solve. Is there a way I can incorporate nonstationary variables in the model code so I do not need to accumulate outside the model?

[jpfeifer]

What is the problem you are facing? The non-existence of a steady state due to a random walk with drift?

[Jaime_Fierro]

Yes, potential output has a unit root y(t) = y(t-1) + g + e(t); dynare says “Impossible to find the steady state”; so I have to run the model as dy(t) = g + e(t) and then obtain potential output outside the model by accumulating dy(t). My question is: Can I run the model in dynare including the nonstationary variable y(t) in the model code?

[jpfeifer]

See

Nocheck command?

Hi, The nocheck option is generally used in models where you have random walks with drift, ie equations of the form: a_t = \alpha+ a_{t-1}+\varepsilon_t There is no steady state for a, except if \alpha=0 in which case there is an infinity of deterministic steady states. In this situation (where the steady state does not exist) you can provide an arbitrary value for a in the block defining the steadystate or in a steady state file routine. This is safe as long as the elasticities and the rati…