By "feel free" did you mean "please" ...? Is English your second language, or it's just lack of manners ...?
I'm not an expert but something like wavelet scaling or wavelet multi-scale decomposition might be what you're looking for. http://www.google.com/search?q=fractal+wavelet+scaling+decomposition+stock+market
The markets contain a random component. Markets can be modeled as trend plus noise, where the noise is random. Whether or not you can use the noise to make money is an entirely different question. In many cases you're better off using the deterministic component to trade. Under certain conditions you're better off using the noise/randomness to your advantage. Also consider regime switching and seasonality, so the actual model would be trend + current regime + noise. Determining whether any of these components is best modeled as fractal or not is left as an exercise for the reader.
Under certain conditions you're better off using the noise/randomness to your advantage if it is noise as far as you can tell then by definition you can not use it and sir Gordon Deco to say the market is fractal does not mean saying it is random not going to explain this shit read up
AD, Not meant to you, but because Sir aside - I wish MTT would post more RN http://www.elitetrader.com/vb/showthread.php?s=&threadid=22501
One guy's narrow range inside bar is another guy's Bollinger narrowing and could be another guys doji... One guy's trend reversal is another guy's continuation... One guy's stop placement is another guy's bread and butter... One guy's noise is another's volatility...
yes, quite true. The hurst exponent but be exactly 1/2 in order for the fractal description to imply randomness. Many times it is not. In any event it is easy to see that the pure fractal model breaks down for markets. Just open up a tick chart and you can see that scale invariance is not held.