I am currently replicating a pretty large DSGE model. I was able to find the steady state numerically (analytical steady state seems to be impossible to find) given the parameters values authors provide in their paper.
My question is how to calibrate the model now? I have been looking through the forum but found no general example of how one can proceed with calibration in a large model, where there is no no analytical steady state. I understand the idea behind calibration, but have never done it myself. Any help would be appreciated.
I have also tried modularization and change_type techniques, but it encounters the impossibility to compute a steady state for some values and terminates.
It very much depends on what you are trying to achieve. There is a continuum of approaches to fix parameters, ranging from pure calibration via loose moment matching to full-blown estimation. But in all cases, you need to be able to solve the model and, in particular, to compute the steady state. For large models, that typically means to reduce the size of the problem until you have a small set of equations you can numerically solve.
My goal is to match steady-state endogenous ratios with ratios from data (such as C/Y, K/Y, etc.). This is the first part I am currently working on. I have successfully solved for the steady state using the initial parameters provided by the authors. Now, I am attempting to recalibrate the model. Do you have any approaches you would kindly recommend?
The second step would be to match theoretical moments, but currently, that is beyond the scope of my work.