Hello,
I am trying to estimate the Smets and Wouters (2007) model using the TaRB-MH algorithm, in Dynare Unstable (2015-09-30).
The estimation command is called as:
estimation(mode_compute=1,use_tarb,order=1,datafile=usmodel_data,mh_replic=5000, mh_nblocks=2, mh_jscale=0.3);
The issue is that the MH algorithm is runs VERY slowly. Did I do something wrong? Or is the TaRB-MH algorithm so computationally expensive?
Thank you