Inspired by Adam Grimes' latest webinar, I've started to dip my toes into the world of quant-ism. My question is, are the markets fractal? And if so how do you account for noise in smaller timeframes?
For instance, if I get my hands on a bunch of EOD yahoo data and conduct some statistical analysis for something like a simple MA crossover over 10 years and it gives me a 50.1% probability with certain money management, and let's assume that markets are fractal in nature, will this probability hold in an hourly timeframe? How do you account for the noise (randomness) in the lower timeframes? Is every test in a different timeframe independent? Perhaps my question is misguided by an assumed trader-ism of lower timeframe charts being much noisier than higher timeframe.
I would appreciate if someone with experience creating financial models could chime and give their 2 cents.