Spectral Analysis Technique
I've been spending some time the last couple of months doing some spectral analysis. There are 2 periods of interest.These are not the same as Hurst cycles.
A 51 year period topped in 2003, and is entering its steepest deline. This cycle is due to bottom in about 15 years. Since this top occurs between two market tops, there are large-amplitude smaller-periods completely out of phase with the 51 year period top. Hurst might call this a cycle straddle, and I've read poster's comments that straddles should be treated with suspicion; however, high corelation with the market should give reassurance.
A 2.57 year period should be topping any day now. There is a possibility this year long bear market rally makes one more high.
Individually these periods of interest corelate better than 94% with the DJIA. I have not calculated the corelation of the combined periods. Or of other periods.
Based on more available shorter periods, I'm finding periods merge, split, and fracture. I'm trying to figure how to read the spectrum to detect these changes.