Page 1 of 1

Dealing with exogenous, observed variables

PostPosted: Thu Jun 23, 2016 2:30 pm
by dhcooper01
Hello,

I am estimating a model in Dynare using MLE where I have a few variables -- such as oil prices-- that I want to treat as strictly exogenous and observed. I know that to avoid stochastic singularity there have to be as many shocks in the model as observed variables. However, is there a way to get oil prices into the model without having to specify a shock process for it such as measurement error. I have not seen any examples of this online so I guessing that perhaps it is not feasible. In AIM you just do something like "EQ > poil = 0*one" when something is strictly exogenous (and given).

thanks,
Daniel

Re: Dealing with exogenous, observed variables

PostPosted: Mon Jun 27, 2016 10:11 am
by jpfeifer
My guess is that the univariate Kalman filter in Dynare can handle the arising stochastic singularity.