a1 Cass Business School
a2 University College Dublin
a3 Sheffield Hallam University
This paper proposes the use of wavelet methods to estimate U.S. core inflation. It explains wavelet methods and suggests that they are ideally suited to this task. Comparisons are made with traditional Consumer Price Index–based and regression-based measures for their performance in following trend inflation and predicting future inflation. Results suggest that wavelet-based measures perform better, and sometimes much better, than the traditional approaches. These results suggest that wavelet methods are a promising avenue for future research on core inflation.
(Online publication June 10 2010)
The authors thank Tim Cogley, Ramo Gençay, Julie K. Smith, Theo Panagiotidis, and two anonymous referees for helpful comments on an earlier draft. They also thank Julie Smith for permission to use her data. Cotter acknowledges the support of Science Foundation Ireland under Grant Number 08/SRC/FM1389.