ABSTRACT
Traditional time series analyses rely on methods that involve either the time or the frequency domain. But wavelet transforms permit an analysis that combines both time and frequency information, the latter in terms of levels of time resolution. Usually wavelet transforms are made with equally spaced observations whose number is an integer power of two. Here, we show how to go beyond these constraints by using a form of semi-nonparametric regression analysis to construct wavelet patios for selected international commodity price series. These patios show the magnitude of the variations in the series at different time scales for different subperiods of the full sample.
