Adding steady state priors


#1

I have a simple 3 equation NK modelnk.mod (1.2 KB)

The model is already log-linearized with respect to the steady state and (I guess) it runs correctly.

I have data on three variables (GDP growth, Inflation and the Federal Funds Rate) and I would like to estimate this model using Bayesian tools.

Apart from some key parameters, I would like to put priors on the steady state values of GDP growth and inflation and estimate them. However, given that my model is already log-linearized, I do not know how I should modify my existing code in order to estimate it, specially estimating the steady state values of this endogenous variables. I guess I should create new variables but I am not sure how to.

Thank you very much!


#2

Have a look at the Smets-Wouters AER 2007 code.
In the measurement equation,
the output growth rate in the data = average (steady-state or trend) growth rate + change in the model output-gap.

You can set the prior of the steady-state growth rate at the unconditional mean.Similar strategies can be used for the level (not growth rate) of the interest rate and inflation.

Cheers
Reuben


#3

Thanks Reuben!

Great Reference!