In this paper, we measure, with Canadian data, the scope of the revisions to real-time estimates of the output gap generated with several univariate and multivariate techniques. We also make an empirical evaluation of the usefulness of the output gap estimates for predicting inflation. Our findings suggest that, for all techniques, the standard deviation of the revisions is of the same order of magnitude as the output gap itself. We also find that, with the exception of the Beveridge-Nelson technique, all revisions are persistent, which means that there is a long lag before the scope of the output gap revisions is fully known. Finally, we find that the output gap estimates do not significantly improve the inflation forecasts. We infer from these results that estimates of the output gap are not very reliable.

Also published as:

The reliability of Canadian output-gap estimates
The North American Journal of Economics and Finance (1062-9408)
December 2005. Vol. 16, Iss. 3, pp. 373-93