Page 1 of 1

variance decomposition

PostPosted: Mon Jul 11, 2016 1:41 pm
by Cathy
Hello

I don't know exactly how to interpret a result for shock decomposition for output attached in the document.

The decomposition of shocks shows the contribution of government expenditure shock (e_g) and tax shock (e_tau) to the dinamics of output. Results in the graph shows a negative contribution of tax shock in 2008 to the dinamic of output (when the deviation of output from its steady is positive) and a positive contribution in 2009 (when output gap is negative). I'm really confused about the economic interpretation. Could you help me to clarify the explanation about the contribution of tax shock?

Thank you.

Re: variance decomposition

PostPosted: Mon Jul 11, 2016 7:18 pm
by jpfeifer
Sorry, but it is your job to come up with an understanding of what is happening in your model. After all, that is the purpose why you built this model. If you are not able to explain what happens there, this is a very troubling sign four your model. Crowd-sourcing the intuition for your model is not a good idea. If you ask enough people, someone will surely come up with a convincing story. But you will not know whether this is a true story or whether there is still a problem with your model.

Re: variance decomposition

PostPosted: Mon Jul 18, 2016 1:30 pm
by Cathy
The results are a replication of the model used in this study (http://www.sciencedirect.com/science/ar ... 9312001071) so I didn`t considered that there might be a problem with the model.
I will try to understand better the model and the variance decomposition of the shocks.

If there are any other examples with variance decomposition explained (except from those in An Introduction to Graphs in Dynare) it would help me very much.

Thank you :).

Re: variance decomposition

PostPosted: Mon Jul 18, 2016 6:21 pm
by jpfeifer