Can structural vector autoregressions (VARs) discriminate between macro models? In simulation exercises, VARs will only infrequently reject the true data generating process. However, equally important is power: the rejection rate of false hypothesis. For a set of DSGE models, we report power results for both the standard test of the sign on impact and a… (More)

@inproceedings{Gust2009ThePO,
title={The Power of Long-Run Structural VARs},
author={Christopher Gust and Robert Vigfusson},
year={2009}
}