The RATS Software Forum

I am using Replication file for Jordi Gali(1992), "How Well Does the IS-LM Model Fit Postwar U.S. Data". The stored impulse response data doesn't match with the impulse response produced as the graph. Can we use history series created for historical decomposition of other variables. I mean to say is HISTORY(5,7) represent the variation in inflation due to IS shock.

abhishek wrote:I am using Replication file for Jordi Gali(1992), "How Well Does the IS-LM Model Fit Postwar U.S. Data". The stored impulse response data doesn't match with the impulse response produced as the graph.

I'm not sure what you mean. The last set of impulse responses are for one set of drawn coefficients, while the center in the IRF graphs is the median of the whole set of simulations. They shouldn't be the same.

abhishek wrote: Can we use history series created for historical decomposition of other variables. I mean to say is HISTORY(5,7) represent the variation in inflation due to IS shock.

That's correct.

How can we save the median impulse response?

Use @MCPROCESSIRF.

Thanks Tom

I'm interested in modifying the Jordi Gali(1992) by adding one more variable exchange rate thus making variable list to 5. I want to create a corresponding shock namely external shock. Further I want to bifurcate supply shock in internal and external shock. Can you suggest something on this or any reference?