Do you care about reading details ? I'm talking about paying employees time of my former business not about me. As for me I don't care about 24h work if I could because work for me is just like leisure: I only work in domain that pleases me. That's why I didn't even count the cost of this project in term of my own time: I count it as cost 0 - otherwise it would be much more than 1.5 Million francs. That's the interest of not being an employee: you do what you want. But you must financially be able to do so and support the risk. I don't consider my time as precious except during certain hours of course when I don't want to be disturbed.
And generally when I post here it's because I'm waiting for my software to finish a calculation . That's why I make often sporadic posts since I have to interrupt from time to time hee hee !
And it is my pleasure not to be obliged to sell anything because I can answer what I like to sarcastic or jealous guy aren't you : maybe you have to work hard as an employee for a boss, if it is the case well sometimes it's cool to be just employee I've been for two years nothing exciting but just cool : now never I will support to be an employee
Hey I like Harry he is one of my favorite posters Reading his posts are like figuring out a jigsaw puzzle ... they keep you from getting to complacent. Keep up the good work Harry your cheaper and more fun then puzzles anyday.
The probability of predictions for a limited period of time can not possibly be a criteria for the "true patterns" as the time frame change could drastically distort the "reliability" of such patterns. It is very naive to participate in the quest of searching for "true patterns' as it would be naive to try finding the "holy grail"....but what ever blows your skirt up! ..... Cheers,
??? Since when "limited period of time" has something to do with a probability distribution . Let's take a simple example: you predict a pic pattern at X within 3 points, a probability can be perfectly calculated for that assuming a normal law, rectangular law, triangle law or more sophisticated laws like Levy, frechet,.. whatever law you want (if you pretend that there is no prob law, it's even worst for you as it means that the predictibility due by chance is even more reduced). In fact there is even a fashionable new tool in financial risk management called "EVT" for Extreme Value Theory. And if you know EVT you perfectly know that it is extremely difficult to predict the queue of a distribution - since a pic is an extreme value haha !
example for yesterday trendline: the slope was slightly upward so that one could think it's just random slope that could be horizontal as well depending of the crowd buying at random. But where did the break occurs to less than 1 point from theorical value = 10469.7 (as I said I don't intend to prove anything I just show an example but I know now by experience that it repeats again and again and on multiple time frames. I don't care about that kind of observation as I said I don't trade traditional trendline and patterns it's just to show that the trendline doesn't occur at random) http://www.econometric-wave.com/elitetrader/slideshow/1/ A traditional TA would would consider a horizontal support at 10450 as I saw for that day this is 19.7 below my theorical point ( and so represents roughly 190$ of loss of precision per future contract).
I guess by reading your arguments that you one of those guys who are preaching against the traditional chartists who draw trendlines . To tell you the truth I was also against that since I didn't see any rationality in the pseudo psy reason they gave for such trendlines or patterns to exist in market. I have never believed in psychology of market as driving force of TA because there is no rationality behind that argument when one looks at the details - and many physico-financial researchers had made enough simulations to have never been able to demonstrate TA by modelising such crowd - and my model is based on economic reasoning not on psychology of crowd. Even after one year of producing my model I didn't see much relation since I don't look at them but after a while I finally see that there was a link: the pattern or trendline is indeed an accumulation zone until my theorical point is reached: this is the criticality point refered to in chaos theory. As a metaphore the market somehow waits until that point of criticality is reached because it is there that the avalanche has its optimal force. Moreover it was even in reaction against the psy pseudo-mystic explanation that I built this model and It proves right that these patterns if they drive psychology of crowd, it is not the crowd that is behind, the crowd only follows. You are talking of holy grail ? I just remark that you pretend to be able to have a consistent edge based on price analysis and according to official theory of efficiency of market this is not possible on long term basis (meaning consistently with a prob > 50%) and so "probably fraudulent claim" and I used the expression of CFTC if you remember the post I made on them. Then you can also classsify yourself into Holy Grail researchers . Secondly I am not searching true patterns per se, because I don't need to since I don't trade them, I just say that these patterns are not so illusionary as one could think as "rationalist" and as I said that's what was my behavior until I have some proof of the contrary after several years of looking at intraday charts at minute scale and seeing the link between the result of my model and some traditional TA patterns. So you will probably stay skeptical and I understand that since I stayed skeptical also for many years. And you don't know my model so you have no reason to change your opinion with that I agree.
btw you are misunderstanding what I am doing: I'm not doing research, it is a few years already that the research has been done. Then it has been traded discretionarily then the backoffice has been automated and now I want to automate my own brain if I can express it like that. It's the most complex phase as the circonvolution of my brain is sometimes difficult to grasp as you know ... even by my one self . Also I have to cope with a guy called a programmer and that slows much the process more than I would like. Also I come from industry and software engineering and when I do something I want to garantee the results that is for me or for clients and the only way to do that is to have a process that is completely streamlined to the finest details. What I will automate is not only entering and exiting position but the risk management part notably trailing stop but more efficient than the usual trailing stop used in candlestick or sar which have too many false signals. For that I have to automate the path of the stock market inside the day and not only at entry or exit. To get an idea a process is like this (I took it from Borland site) in fact what is important is the iterations (those who know RUP - Rational Unified Process - will understand). <IMG SRC=http://www.elitetrader.com/vb/attachment.php?s=&postid=471666>
That's why I will retire the current site presenting the model. As a substitute I will create a portal ... that will essentially serves for me to store my financial bookmarks: you know that I have tons of unorganised links but be reassured I won't put them all . Nevertheless for people that has already subscribed or being on the mailing list I won't upset by refusing them if they are interested - at the moment still there is nothing I can offer as long as automation is not finished and thorougly tested. But there will be no more acceptance for newcomers except if they can prove that they are professional traders or firms.