Hi all,
I am estimating the model of De Graeve (2008) with some minor modifications. When I run the estimation command I specify 5 blocks. Blocks 3, 4 and 5 behaves normally but blocks 1 and 2 have 0% acceptance ratio.I though it could be that MCMC runs to the boundary of the stability region or to the prior bounds. Nevertheless, as far as I know it does not make sense that it happens just to some blocks (which are always the same 1 and 2) and from the very first draw.
I attach mode and data file.
The mh_replic=2000 because i was trying to solve the problem but I have done it also for 250000 replications finding the same problem.
FF_DSGE.mod (6.2 KB)
GD_data.mat (4.8 KB)
Could you please also provide the mode-file you are loading in the estimation-command?
1 Like
Thank you for the quick answer. I realize that some of the variances of the shocks were at the boundary of the prior bounds. Therefore, I get rid of the boundary for the shocks but I still have the same problem.
FF_DSGE.mod (7.6 KB)
FF_DSGE_mode.mat (7.9 KB)
GD_data.mat (4.8 KB)
FF_DSGE_results.mat (95.4 KB)
Which Dynare version are you using? When I use
estimation(datafile=GD_data,mode_compute=0,presample=0,mode_file=FF_DSGE_mode,first_obs=1,prefilter=0,mh_replic=2000,mh_nblocks=6,mh_jscale=0.1,mh_drop=0.2,bayesian_irf,irf=20,moments_varendo,filtered_vars,smoother,conditional_variance_decomposition=[1 2 4 10 40]) Y C I L PIE W R Prem ; //
in Dynare 4.5.4, everything works as expected:
Estimation::mcmc: Number of mh files: 1 per block.
Estimation::mcmc: Total number of generated files: 6.
Estimation::mcmc: Total number of iterations: 2000.
Estimation::mcmc: Current acceptance ratio per chain:
Chain 1: 36.65%
Chain 2: 38.15%
Chain 3: 37.05%
Chain 4: 35.5%
Chain 5: 36.55%
Chain 6: 35.05%
1 Like
I was using Dynare 4.5.4. Now I have tried in another computer and it works as in yours. So I will use this one.
Thank you very much