Sounds to me like a multidimensional optimization and curve-fitting across many assets and timeframes. No wonder why some of these investment houses went bust.
The goal is to fit a real underlying process and not fit random noise. And to detect when the underlying process changes. "Curve fitting" and "optimization" are terms that have a bad public press but are used every day to find solutions to real problems in a large number of fields. Most people working in this space aren't that stupid. They know the issues. Do the techniques need improvement? Of course. Are the techniques applied incorrectly by some people? Of course. Does that make the techniques invalid? Does that mean you stop trying to improve them?
This is middle office mumbo jumbo & buy side voodoo. Not trading strategy development. Risk managers use these models to police the front office. Risk to Trader: "You need to unwind that, because the model says your risk is too concentrated". The middle office is another career path for failed traders.
Actually that is sort of what I meant -- not so much for trading (day or swing) but more for balancing risk in a long-term portfolio of assets. My neighbor does not describe himself to be a trader but is a principal in a small financial firm that specializes in asset allocation. My guess is that longer term, there is bound to be more signal as you can model things like interest rates, inflation, GDP, real estate prices, corporate profits, etc etc...
I did find the authors' claim that certain system types had higher failure rates weak, since he didn't provide any empirical substantiation.
Many of the losers here may think you advertise the programs because they think like losers and will stay losers all their life. My main problem is that I cannot afford to purchase a lot of software at this point. But I believe some people have success using them. Is the APS the same with Price Action Lab?
The equity curve looked like commissions were missing. Looks like the entire post has disappeared now.