Why are you getting so nervous? How do you know whom I know? Anyway, most floor traders I have met never went to college. Regardless, What is your point anyway? With what issues do you dissagree? I find it extremely hard to understand what your point is. Maybe we agree, but your attitude makes understanding it very hard.
Maybe - but you don't read the posts. There was a discussion about statistics and my point was to use it careful - that is all.
What I am looking for is how to incorporate this into the actual trading algorithm. This is in turn would allow you to optimize HOW OFTEN you optimize your parameters. I know this sounds horrible and it seems like a double dose of curve fitting but I believe if you can find out HOW OFTEN you should optimize your parameters it would help identify changes in market cycles and thus improve your results. I still don't think anyone's hit on this specific topic.
I use three main cycle principles: no two price cycles are identical, price cycles are nested within price cycles and longer cycles dominate. Each of these principles applies to cycles for price and volatility. Based on these principles it is not possible to optimize parameters to tell you how often to optimize your parameters. For example minor cycles are nested in major price cycles. If I could optimize the minor price cycle so as to predict their periods, then this in turn would allow me to predict a typical period for the major cycle. Then I could predict price tops and bottoms. The above does not work that way with cycles because of their third principle. Longer cycles dominate minor cycles constantly affecting their period. This is why principle number one states no two price cycles are identical. Then only after a major cycle has completed will you know all of the particulars in the relationship between the major and minor cycle and the periods of cycles. So what I am saying you cannot accurately take the current projected periods of the major and minor price cycles and optimize them to project when all the minor cycles will complete to coincide with changes in the major cycles. Who ever succeeds in doing so will own this planet.
Cycle types as you describe (major and minor) are based on the time frame selected to trade in. You can also chose to trade based on multiple time frames which could take into account both. What I am looking for (optimizing optimization periods) may or may not be effective. This is why I am trying to test this theory and find out for myself. The cycles I am referring to are optimization cycles which may or may not correlate to market cycles.' You sound like you have done extensive research on market cycles so I would be interested in hearing your input. Right now I am in the process of exploring algorithmic trading methods. I started automating my discretionary strategies in mid January. I have coded and tested all the methods I have observed over the past three years of full time trading. Now I am looking for more obscure mathematical approaches to improve on these results and possible create new strategies all together.
About 14 years ago I was working on project similar to yours. I wanted to find out all of the metadata about the cycles (and the cycloids) of market indexes and select stocks (not part of the index) that I could find. Gathering the cycle metadata individually from major cycles of indexes and stocks was not a problem. But trying to drawn conclusions and find correlations from these cycles was next to impossible. Even though it is true some correlation existed between major market cycles and major stock cycles. It was not enough to help in designing my strategies. As it turned out the wild card in this process was volatility cycles. Volatility cycles are the defining metadata that twists a stocks price cycles so they have no correlation to market cycles. For example a stock volatility cycle can drive a stocks price so no correlation can be seen between the stocks and the markets cycles. So how you define and use volatility cycles is crucial IMO to building strategies that trade the markets.