The RATS Software Forum

This replicates the three structural VAR models used in Cecchetti and Rich(2001), "Structural Estimates of the U.S. Sacrifice Ratio," JBES, vol 19, no 4, 416-427. Each of these are SVAR's of various sizes with short- and long-run restrictions. They are used to compute the "sacrifice ratio" of GDP lost to disinflationary policy. By using bootstrapping, they show that the reasonable-looking point estimates for the sacrifice ratio don't survive when you take into account the uncertainty in the dynamics of the model.

This is a good example of the use of bootstrapping for an SVAR, since the bootstrapping maintains the contemporaneous relationships among residuals. As a result, you can just do a standard bootstrap and re-estimate the SVAR with each draw.

Ceccheti-Rich JBES 2001

Detailed Description

Respected sir,
I am really sorry for not telling you that the result I reported in the earlier post is on my own data. It is not based on the data of Cecchetti and Rich (2001) paper.
But, after running the data of Cecchetti and Rich (2001) I found that the sacrifice ratio is 1.28369 (original is 1.2768). Further, though I got an IRF which shows 90% band but it does not match with the original work of Cecchetti and Rich (2001). I have taken same sample period (1959:1 1997:4) and lag length 4. For your illustration, I have attached the IFR figure below.


Sincerely,
Dinabandhu

First of all, bootstrapping is a random calculation, therefore you will never get anything that's exactly the same. Also, the description of how those graphs are created is sketchy at best. Second, if you've ever tried to replicate published work, you would know that 1.28369 vs 1.2768 is about as close as you will ever get based upon the description of an analysis in a paper. Sometimes it's hard to get closer than that even if you have (theoretically) the original data and the original program.