Don't worry sometimes I don't understand myself my brain is faster to find ideas that I can follow him
Now seriously read this article from Mandelbrott to get the basic on fractal and market geometry: This article originally appeared in the February 1999 issue of Scientific American. For a copy of it, please contact Scientific American at 212.754.0550. We highly recommend a subscription to this important magazine. A Multifractal Walk Down Wall Street http://www.elliottwave.com/education/SciAmerican/Mandelbrot_Article2.htm This is the classical and immediate way to describe fractal but what is immediate is not surely what is the most pertinent: law hide behind apparence you cannot grasp it so easily and for me the L System could be more subtil than the trivial I mean visual approach.
Mandelbrott said at the end of the article: "These techniques do not come closer to forecasting a price drop or rise on a specific day on the basis of past records." Sure it can't forecast anything since it is only descriptive techniques as descriptive statistics. There is no equivalence of some kind of statistical induction in fractal domain. That's why one need first a physical model (with cause and effect) before using fractals if one must do prediction. If not so fractals can only be used for stochastic evaluation of risk which is better than classical theory but far from enough for speculating as an individual trader ( traders for institution don't have the same problem: they trade with OPM: others people money )
and reach the Hall of Fame ? First I should have improved my english (wait just 100 years), second there is an expression in french "pour vivre heureux vivons caché" (to live happily live hidden) although I seem an extraverted person I am not but I like to express my ideas but not if I have to become famous as a person, third as I said I don't want to attract the mass traders to my stuff only the ones I filtered that has passed the antimarketing filter - which would the first sign of their intelligence , 4th I have to write all this in a doctorate tone that's what I do when writing reports for clients but it's so boring happily I am switching to trading from partial to full time.
Just a useful remark: you can see an hour indication associated with the 7556 target which is 32:30 : 16 proj min base: 7711.35 proj: 7556.77 17:00 (32:30) it means in linear theorical time on my model it should have happen in 32/16=2 days (why 16 because I have noticed that US+Europe is like a single market which trade 16 Hours). So the low we made Friday was a fake and we turned back quickly to the "normal" target of the day (a normal day is 10 unit of times on hourly scale and 1/2 unit of time on daily scale except when it is consolidation which will need an expanded time chart at the moment this version of chart has not enough length for showing consolidation target) on hourly scale which was 7631/7613 on future/spot (you can see BRK which means break zone, the highest break zone is a very good buying point because it cuts like a knife and stop can then be very tight ): <IMG SRC=http://www.elitetrader.com/vb/attachment.php?s=&postid=215750> This is what I called a forcing phenomena, in fact I have already presented it. On scale 1, forcing implies at least the retest of the forced zone within 24hours and often less depending on the following periods projection.
Statistically (see concept of "Resistant statistics" http://www.met.rdg.ac.uk/cag/courses/Stats/course/node13.html ) "the extreme minimum and maximum values .. is a good example of a statistic that is not at all resistant to the presence of an outlier " that is to say if a model can predict consistently the bottom and the high of the day then it is out of probability that it is just by pure hasard. And my model just do that: it doesn't predict the close it predicts highs and lows often with 2/3 point precision (except in days like Friday the precision is less good since it is then rather 5/6 points on hourly scale and 10/12 points on daily scale) less than the intrinsic standard deviation of a 1 minute bar on Dow Jones which is merely impossible by pure stochastic law !
If you wonder if knowing the high or bottom of a day within 2/3 points or even at worst 12 points does make money or not, well you must be a great philosopher that like to ask infinite question which should also give you a headache I have posted trades in past years at that time I coudn't yet predict top and bottom with so much accuracy. I intend to do it a few days/weeks demonstration again in the future: this is part of my Big Yellow Project in reference to Big Blue. The automation is not yet achieved because the demonstration aims to show that all the trades can be mechanic with a very high win/loss ratio up to 10 sometimes more and a high probability of 80% - whereas in past years I did it with my own discretionary feeling. You see I know you as traders: when I trade I am like you a pavlov dog I only understand buy, sell, stop. I don't think when I am focused on action. But action is very brief the rest of time I can think. So I will talk this simplest language during the demonstration phase so that some of you will not get a headache Now don't be frustrated not to understand what I post since you never cope with my model. And I won't make effort to explain to people who won't make effort either. I like to post just as others smoke it's for my own distraction you're are not obliged to find that distracting yourself . But I assure you you should do the same : posting can give you ideas which has nothing to do sometimes with the subject of ypur post but it is like a writer writing a novel it can inspire. Sometimes I am even inspired by stupid arguments of others because it pinpoints me where things are wrong that's why I like people's opinions even contrary to mine What I don't like is people having no opinions at all and just attack like Brutus.