S&P Long Term Cycle Phasing

David’s patterns, in my humble opinion, adds more confusion to already complex subject. David is using 8 sequence to identify market position within larger cycle and at times it is very subjective. This is an approximation of even harmonic. If you know how to phase cycles, sequence is irrelevant. 8 points in the sequence identify 4 cycles. Even bigger issue is when you apply David’s sequence to market instrument or time frame where odd harmonic is dominant. For example 9 18 54 - even sequence fails. Using fld as a trigger or conformation or getting price targets are all mathematically sound and helpful.

I do agree that the FLD pattern is muddy much of the time. I found this one to be fairly clear where the pattern cues seemed to be right on time when overlaid. To see how it might be different within different portions of the 80w odd harmonic with the 54 month I made several different segments. Its still often not ideal looking and so, difficult to interpret at times. This one struck me as being more clear than most so thought I’d share. Although it doesn’t look like that last leg will occur as predicted based upon yesterday’s action.

  • BillC

Yes, it feels great when price follows market model exactly to the tee. But even an average model is better then no model at all. Building models on different time frames helps. We are looking for weekly model to support daily model and so on.
The kind of volatility we had in last few days will kick any daily model in the gut but weekly model did not even blink and has swallowed this volatility increase like nothing has happened. Daily model will need half of the smallest cycle to adjust.
I am working on some way to reduce influence of sudden volatility increase on the model with promising results. The goal is to transform price chart, make it adjusted by historical volatility before sending to model engine.

Been awhile since I posted. The forum appears pretty dead. Could be a by product of the crazy market we are in and the cycle being fairly tough to read.

Here is my current phasing. Its a tough and distorted seeming market where the bottoms aren’t necessarily break FLD’s and some of the spike lows don’t necessarily form great VTL’s either, but I picked what seems to be the most probably bottoms. There is a lot of low amplitude on many of the cycles right now as well, so some of the lows do not appear as prominent as might be expected.

Second picture shows what the computer thinks from the Butterworth filer perspective.


  • BillC

Every model will fail sooner or later, because model looks at finite number of cycles. It is a tradeoff speed vs precision. So long term phasing always helps. The point of contention is crash low in March of 2020 it must be attributes to 9 years cycle, but with bullish left translation of price cycle low in February 2016. Every on still remembers unprecedented, well almost unprecedented events of March 2020 it becomes apparent.

Even when looking at the long term model, price is still stronger indicating that model is missing even larger cycle that is influencing market.

A different perspective.

Hi William, you have your latest 3,5year low ~July 2019, where your data shows a low of 2400, lower than the March 2020 crash low. Price never corrected lower than 2700 in SPX during 2019. What is the source of your data and how do you accept it to come around to this cycle picture?

Disclaimer: I do not subscribe to the always low synchronized, simple harmonic ratio approach to cyclical analysis. I still think ST is a great piece of software.

I noticed other analysts assigning the March 2020 low as a 9 year low. That resulted in one of the prior 4 year cycles being almost 7 years long if one followed the simple harmonic ratio approach. An unlikely scenario in my opinion. I felt that the magnitude of the drop into the March low was exacerbated by the Covid news, i.e. a fundamental interaction.

I’m not privy to ST’s underlying phasing algorithms but it is fairly obvious that amplitude plays a role in assigning the magnitude of the troughs. I decided to try a little experiment for fun. I modulated the amplitude of the 40 week wave over that short period before running the data through ST. It resulted in a relatively symmetrical analysis over a 25 year period which is in accordance with the historical wave averages based on Hurst’s bandpass filter approach.

I’ll be honest in that I didn’t even try to use ST’s “pinning” function. I might get the same result. The result shown is in accordance with the bandpass filter approach I employ.

1 Like

March 2020 low was not a 9 year cycle low. May be I was not clear, 9 year cycle low happened to be between March 2016 and March 2020 lows. I think March 2020 amplitude can be attributed to 54 month cycle low, exaggerated by the interaction with weak 18 year cycle, topping and 9 year cycle just starting to accelerate to upside . The main difference in analyses can be summarized , if William is correct market collapse should start sometime in the first-second quarter of this year and continue down for several years.
I think market still has some room to grow at list before the 9 year price top sometime in the second quarter of 2022.

Hi b3cker,

I was not making reference to your analysis with respect to March 2020 as a 9 year low but to David’s most recent S&P update.

Personally I think the 18 year wave is just an artificial construct and not a price wave intrinsic in the data. Even the 9 year wave lacks cohesiveness over the history of the market. However the 4 year wave is extraordinarily consistent over the history of the market.

I’m certainly not forecasting any “market collapse” based on that ST analysis. It was merely an exercise demonstrating the effect of amplitude modulation.

I agree with your assertion that this run is far from over from a probabilistic viewpoint based on history.

My mistake. Yes, David always has couple of models up his sleeve :grinning:. One can only envy . One model calls for immediate collapse.