What is it about?
Impulse response functions are a typical mean for representing the dynamic dependency of economic variables of interest. They are obtained from econometric estimates, e.g. of vector autoregressive models. Consequently, the time path of impulse response functions is subject to uncertainty. Confidence bands provide a mean for describing this uncertainty. The methods presented in the paper aim at obtaining confidence bands which include the true impulse response function in, e.g. 95% of all cases. A part from this correct nominal coverage, they should be as small as possible in order to be informative.
Featured Image
Why is it important?
Standard methods built in many econometric software tools do not consider the coverage for the path of an impulse response function, but only one particular point in time. This might result in erroneous interpretations of the significance of impulse responses.
Perspectives
While from an a priori point of view, it appears that the methods based on the idea of Wald tests should perform better, we provide both theoretical and simulation based evidence that this does not apply for the specific setting of impulse response functions.
Peter Winker
Justus Liebig Universitat Giessen
Read the Original
This page is a summary of: Confidence Bands for Impulse Responses: Bonferroni vs. Wald, Oxford Bulletin of Economics and Statistics, September 2015, Wiley,
DOI: 10.1111/obes.12114.
You can read the full text:
Contributors
The following have contributed to this page
