tangweicas wrote:
Dear Prof.Walsh  
Many economists use the method of vector autoregression to study monetary policy.But I have  a question about VAR:
Does the method of VAR need the stability of relationship among economic variables?

another question related the first one:
China is in the economy system transition period and the relationships among economic variables are not stable.Is VAR appropriate for studying Chinese monetary policy?

Thank you!
                                          
Sincerely yours,
Tang Wei
                                         2009-05-13   

Dear Tangwei,
In my view, VARs do require stability in the underlying dynamic
relationships that characterize the economy. This makes them problematic
for studying economies such as China that are undergoing rapid
structural changes.
Sincerely, Carl Walsh