Steady state ratios in log-linearized equation

[Julie]

Hi Professor,

I would like to know in a log-linearized model such as Smets wouters, if we use calibrate the steady-state ratios using the averaged ratios based on the actual data. Should we use ratios based on log(data), or ratios of actual data?

[jpfeifer]

The steady state ratios pretty much always refer to the actual data. The log only shows for the log-deviations from the steady state.