Calib_smoother with non-stationary model

Dear Sir,

I am writing a semi-structural model to decompose variables into trend and cyclical components. I used A nonstationary model for output gap with a const in drift - #2 by jpfeifer to execute the model. I am wondering if I can I use calib_smoother to decompose variables for non-stationary models. Using this results in large autocorrelation for oo_.SmoothedVariables.varname_cycle.

Without more context, it is impossible to tell. But generally, you can use the smoother in non-stationary models.

Thank you for your response Prof Pfeifer. The plot of cyclical component of smoothed GDP,
forum.mod (6.0 KB)


forum_data.xlsx (15.6 KB)
data and code is attached for your reference.