Investment specific shock process in NK_baseline.mod

Dear Forum,

I am trying to understand the NK_baseline.mod that comes with the dynare example. As mentioned, the code follows Fernández-Villaverde and Rubio-Ramírez 2006. In their notes they have a random walk with drift process for log of investment specific technology shock i.e. \text{log} \; \mu_t = \text{log}\;\mu_{t-1} + \Lambda_{\mu} + \sigma_{\mu}\epsilon^{\mu}_t but in the NK_baseline.mod it is written as log(mu_I)= Lambdamu+epsmu_I which is nothing but a white noise process. I have tried to specify it as a random walk with drift i.e. log(mu_I)= log(mu_I(-1)) + Lambdamu+epsmu_I but the model doesn’t run in this case and I get following error- The steadystate file did not compute the steady state.

So can anyone explain what is happening here?


PS: PFA code for reference
NK_baseline_steadystate.m (4.7 KB)
NK_baseline.mod (12.1 KB)

I think this is an implementation of the section “Stationary Equilibrium” and “Change of Variables”.

A random walk with drift does not have a steady state. For that reason, it’s unsurprising you get an error message.

Thank you for the confirmation.