one way to test L5 without intraday data is to enter at the entry price calculated from L5, but always exits at the daily close price, assuming the cut loss and cut win cancel each other for long term. the backtest in this way should produce positive gains if there is indeed an winning bias on the entry price based on L5. I doubt if the entry price does not have winning bias, this L5 approach would be a winning strategy.
Please, those who are here to just critic please go away and continue for researching the holy grail...You are just losing man An error of me. Oh my god....I decide to put a certain capital per stock and not all my capital only on one stock !!! But you can if you want (you are free). Cf. previous response Prorealtime is a real backtesting platform. Even Tradestation has sometimes bad ticks no ??? With prorealtime i can have 3 years of 1min data. Again....RIDICULOUS Make a good stop-loss is one of the key of succes. I can still works on this aspect. But you need to put a sto-loss to be a winner
The more simple for avoid unnecessary speculation and for those who believe in this trading system is that i just put my L5% values...I will do that next morning
Does your trading model include a slippage assumption? In my work with Emintrader system I assume zero slippage but in real trading slippage can sometimes be significant.
No slippage, no commissions, no human errors, no software errors... Slippage important ? It depends on how many shares you buy per order and average daily volume of your stock. Buy 300 shares for a stock with an average daily volume is more than 2millions, if you have an important slippage than change of broker !
I assumed NO slippage, but did factor in $ 0.005/share commission (my broker is IB). I use real Open Prices, as provided by Yahoo! Finance, and 1-min data provided by TD Ameritrade API. I am currently tweaking my facsimile of L5, to be as close to the author's values, and will then publish my full results again.
We are not here to bash, but if you are going to post a journal like this, you must be confident enough to accept some critiquing. There are a lot of reasons why you are posting "superior" results, and you seem very critical of anyone trying to point out reasons to help explain it. ProRealTime has a horrible reputation, and your use of OHLC for your backtesting (as you have stated) is poor testing methodology. Also, one has to question why you are posting here, and on other similar boards. For example, http://www.3stocksonfire.com/tradin...1340f2ba8a704077e048a1b574b5259&topic=11665.0 It is the nature of many discriminating readers of ET, to always ask "why ?"
I am first very critical with me ! I say i made some errors and try to be clear (see my previous posts). I use this trading strategy every day, in real. I am not so fascinated by backesting because i don't have need it. Again you don't understand when you said : "I am currently tweaking my facsimile of L5, to be as close to the author's values" -> it is just ridiculous...Think by yourself instead of trying to make a pale copy. Have a look on laurentb's work. I give here some ideas to develop a different approach of trading "Also, one has to question why you are posting here, and on other similar boards." -> and why not ???? What is really the problem ? I am just loyal
I guess that about says it all. I've tested countless systems that performed well over a short duration, but did not over the long haul. I guess everyone has to learn at their own pace. Best of Luck !
Hi All, Emintrader was generous enough to share his ideas and post results daily. I understand that everyone wants the complete system but we shouldn't be mean. And if you want to reverse engineer L5%, at least don't disclose it on the thread. I've been using something similar and have posted the details here. http://www.elitetrader.com/vb/showthread.php?s=&threadid=162357 Feel free to rip my head off