What are “gold pick” cycles?
You got me. I am a bit dyslexic.
That is why I use bandpass filters. They don’t discriminate between peaks and troughs. As you can see from my chart the filter output phases closely to the 8 year peaks whereas on some of the shorter waves the filter output phases closer to the troughs.
Yes, but traditional filters are always late to any change in phase, magnitude … I preferer my top analysis to stand along and I can always combine both when needed. It is cleaner this way at list in my view.
I agree there is some error at nowtime from the extrapolation through the lag period, but fortunately the error is proportional to the cut-off frequency used. By the time I dial down to my shortest waves, any error is insignificant. I don’t use the filter output as an action signal, but to identify the cyclical highs and lows in the near past. They are very adaptable and require no adjustment since they are based on relatively long cycle period averages.
Thanks guys for posting your analysis!
Are those models adjusted for price behavior in the past? If not really nice composite prediction. I assume its a composite anyway. Are they projected filter outputs or some other math basis?
Thanks for sharing your analysis.
Model is a composite; I use decomposition (guess you can call it filters).
The trick is to know what to look for - the market structure. Cycles constantly and so slightly change in size, amplitude and phase. Some models are more stable than others. But as long as model predicts next nominal 20-day cycle, it is good in my book. To confirm daily model, I build another model on the weekly bars.