Video Summary
The video above, presented without commentary, provides a look into the Dollar Index via the microscope of complex wavelet convolution. Wavelet convolution is one of the more precise methods of time series decomposition available today and well suited for extracting modulating, periodic components from noisy data sources.
We start at a wide ranging, long term bandwidth between the year 2003 to 2023. This includes wavelengths spanning 2400 days to 150 days and is excellent in identifying the prominent periodicity of the ‘nominal’ 54 month component at around 40 months and the ‘nominal’ 18 month component at around 19 months. It is clear, as we mentioned in our last report into the Dollar Index, that there is likely a 2:1 harmonic ratio between this smaller component and the component at 40 months. We examine the magnitude and phase time series of the total bandwidth, noting significant phase alignment (waves are synchronised) at large peaks and troughs over the timeframe.
T…