In a prior post your cycles were 3.3 and 7.2(?) yrs, and now the longer one is at 8.3yrs. This lengthening is surprising. While Hurst found a multitude of discrete frequencies, he focused on sets wherein the components are harmonically related, ie the next longer member is either 2x or 3x the prior one, but strength and influence of each member changes over time.
Your 3.3yr breaks down into shorter wavelengths that are observable and relatively stable. 2x the 3.3yr gets 6.6yrs, approx what was observed since 2000. Therefore, I wonder if the "forcing function" applied by the Fed and other central banks is the reason why your period is not harmonic, or if there is another reason built in thru length of dataset, etc.
The question of longer cycles gets into Kitchen, Kandriattof, and others. Please see my previous comments re K-wave implications.
Your conclusion matches mine, that the trend is now down, and the next/existing trend is downward on the 3.3yr towards the next due date of May2019. There should be ample opportunity to get short on one of the interim rallies and hold on to the conclusion.
Thanks for your response and also anything you can add re your selection of 8.3 yrs as the longer term member of your set.
OT: you may want to imvestigate MBXIX, a mutual fund run by one of the most reputable and historically successful managed futures managers. High Sharp ratio, moderate drawdown, and LT consistency. Correlation to SPX is .20, so it is good diversification.
Oddlot
News 
Market Data 
Discover 
