Smets and Wouters 2003


#1

Is there a successful replication of Smets and Wouters 2003? I’ve tried to search, but it seems to be quit difficult for several reasons.

First, the paper has several imprecisions/errors in the linearised formulas. I’ve read in some slides by Uhlig that the supposed dynare code for the model also contained several mistakes…

Also, what do we understand by a proper replication? Just running the same dynare code in another computer with the same data, or reprogramming from scratch? If it’s the last one, then it’s even harder to find a proper replication.


#2

Yes, you can find the respective model’s dynare code (mod file) in our database. It is necessary to download the replication package and look for the subfolder EA_SW03. Additionally, you can use the MMB in order to compare results from different models (including SW03) of your interest or several monetary rules’ implications within the same model.

Please, have in mind that our replications do not provide estimations of models, but rather the use of estimates from papers directly (parameters’ values). A proper replication, therefore, assumes writing the code (defining model’s variables, parameters and their values, as well as characterizing equations) which can reproduce precisely the same results (impulse response functions) as in the paper.

Hope this helps.