Estimating output gap using IRIS

[whogeeyao]

Hi all,

I am trying to estimate the output gap using a bivariate state space model outlined by Clark (1989) with unemployment. I don't have the econometrics toolbox so I'm trying to see if I can do it in IRIS.

My GDP data is in logs, unemployment in percent. I have two unit roots, one for trend of GDP and one for trend unemployment.

I run into several issues with the estimate command.

1. the model is able to estimate if i take input real gdp and exp(urate)
2. I can assign my model parameters and solve the model, but the re-estimation often fails.

Does anyone know what's wrong with the way I set up my model/estimator in the attached zip?

Thanks!

db.zip