I want to include the confidence intervals in my impulse response functions that I obtain using a vector error correction model. Stata does not provide this information. How can I estimate the confidence intervals?
Hello, I am not familiar with your terminology so I may be completly off road here. However when there is no theoretical results to estimate the confidence interval we usually use monte-carlo simulations tweaking the entry parameters within their range to get an estimate of the variability of the output. Hope this helps...
Have you been able to add it? I found one solution in http://davegiles.blogspot.co.uk/2011/11/impulse-response-functions-from-vecms.html, do you work witk Stata? thanks