Output, Money and Interest Rate in the United States: New Evidence Based on Wavelet Analysis

The aim of this article is to investigate the relationship between real output, real money supply and interest rates in the United States using wavelet analysis, based on quarterly data from 1960Q1 to 2023Q2. The initial data analysis reveals that the time series for real output, real money supply measures and interest rates share similar characteristics: they are non-stationary, exhibit non-normal distributions and display positive skewness. Given these characteristics, wavelet analysis is chosen as the appropriate method for examining the raw data or at levels, because it contains more information. The key findings indicate that the coherency between real narrow money supply and real gross domestic product has significantly decreased over time, particularly at medium and long-term frequencies. Specifically, the real narrow money supply measure is found to be neutral across all frequency bands during the “Great Moderation” period, spanning from the mid-1980s to the late 1990s, but it is not neutral outside this period. Similarly, the real broad money supply measure is neutral only at the medium frequency band during the “Great Moderation” but remains non-neutral at the long-term frequency band, even during this period. Furthermore, quantitively, the coherency between interest rates and real output decreased during 1960–1985 but increased during 2008–2023 at short and medium-term frequencies. Moreover, the findings confirm that interest rates are neutral in the long-term frequency band. These results hold true regardless of whether the effects of real monetary aggregates on real output is controlled for or not, using partial wavelet analysis, further validating the robustness of the results.