but, is the stock market "natural phenomena"?

see, it's not as easy as coming up with a model, fitting it to time and then hoping it works as a predictive tool - you need a reason to apply it. and, this is actually the point where huge amounts of mathematical modelling fails outright.

stated as tersely as possible, fractals are self-repeating patterns, and there's consequently good reason to think you can fit trading data to a fractal-based model. if you're meticulous enough, you can no doubt get it very, very close by solving for the appropriate variables.

and, then, what you have is a model that explains what already happened, in the past. it might be nearly perfect - almost no error bars. but, then using that model to predict the future is truly nothing more than a leap of faith - unless you have some argument, otherwise.

when you're modelling physical data, that's where the actual physics comes in. you can claim that the model is predictive because the force of gravity exists between all massive bodies, or because electricity and magnetism are the same thing or whatever other physical idea you want to present.

but, when you're modelling stock trading, or large scale market behaviour in general, you're just presenting a circular argument: the model is predictive, because you want it to be, essentially. that is, you're trying to understand the phenomena by predicting it with a mathematical theory, rather than using the math to explain the theory. and, see, you might be right. we've learned physics this way, certainly. but, it's just a guess, and it doesn't actually get you anywhere in the end, or at least not in an epistemological sense - what you actually want to know is why you're right.

my understanding is that this kind of deconstruction really applies to any attempt to analyse patterns in the market. and, the reason is that there isn't actually a natural phenomenon underlying the decisions we make. humans are not electrons. and, while we nowadays think electrons are actually pretty random, humans don't even have a wave function.

my opinion is that it's a fool's errand. but, that opinion is only valuable until somebody shows me a truly predictive model - and then we need to figure out why, which is likely to have implications about human quanta that are hard to even imagine, right now.