647 – Business Mobility, Income Inequality, and Dynamic Models
A Dynamic Correlation Analysis of CPI Subcomponents Using Continuous Wavelet Transforms
David Doorn
West Chester University of Pennsylvania
The continuous wavelet transform allows for the estimation of the spectral characteristics of a time series as a function of time. We make use of this to employ cross wavelet transforms and wavelet coherence analysis in investigating commonalities in the time-frequency behavior of the eight subcomponents of the CPI. Cross wavelet transforms allow for assessment of common power structures across series, while coherence analysis is in effect a localized correlation measure that can reveal the strength of co-movements of the series over time and frequency and any changes in that strength. The goal is to provide a foundation for the development of a weighting scheme across the subcomponent series for an alternative measure of core inflation. Rather than dropping some of the subcomponents completely to arrive at a core measure, the goal is to extract the common frequency components of all of the subcomponent series to develop an index based on actual co-movements in all prices over particular time scales. Initial results indicate strong correlations between some subcomponents at certain time scales and frequencies and less for others, along with significant changes in these over time.