Pardon my skepticism, but it's intrigued a bunch of us, and I'm finding quite a few errors in the postings, which raises questions. And now this latest revelation of not using these posts for trades, breeds further skepticism. Someone on this thread will find an optimal value for L5 (perhaps one that is better than EminTrader's), but a lot of wasted hype, and errors in postings could be averted, if we could collectively perform a valid backtest. It is clear that EminTrader himself could benefit from a real backtest, with valid data, and a real backtesting environment. Apologies to anyone I offended with my last post, but I like calling them as I see them.
EminTrader, this is a valid concern on your part; I certainly understand that. Your 18 year chart and AAPL chart doesn't tell us anything at all....sorry. Anyone who trades a system on AOF (act of faith) is looking for trouble. Each and every one of us need to perform backtesting to our own level of satisfaction. Surely, you concur with that. Would you trade a system of mine blindly ? No way, I'm sure. Babe's point of you calling your entries before the market opens is correct. That is a noble effort on your part. However, I just finished spreadsheeting HCBK, and there are more than a dozen error in your L5 posting, your entries, and your exits, in a short 2-month period. I would not have discovered these, unless I took the time to both backtest and spreadsheet the #'s. It has not been proofread (since you are not trading them), and I know for a fact at least 2 members on ET are trading it already. I guess that is the problem I am having here. I hope the discussion continues here; if nothing else, it is a fun academic exercise.
The OP should made it clear in the first post that all the following result are all based on paper trading. For me, from the first time I saw other posters pointed out discrepancy between their real trading result of this system with OP's , I began to doubt whether OP was on simulation trading. And with some experience , you will know paper trading is different with real account trading. As someone who really were trading this system with their real money already told you in this thread: their results are less impressive than the OP's.
The problem is quite simple. Using the backtesting software that he uses only tests the OHLC. That wont work because as I pointed out before in a previous posts you can be long a stock at 10 bucks, target of 10.50 and the stock triggers are 10, goes to 10.25, retraces to 9.50 and then ends up at 10.75. OHLC, you will probably show a winner on his software when in reality, you have at best, a break even trade. Thats what I think is happening and why there are discrepencies. One thing I will say in EminTraders defense, I would NOT post my L5 code for all to see. I think thats quite obvious. Why does he post on elite? Maybe to sell his system, maybe to find investors for his system, gain credibility, etc. All I do know is that since he has been posting he has been making money, no doubt about it.
You are right dafishman... Using only OHLC is not enough to give correct results. It is my fault. I don't have take care of this aspect here (i was not clear). Sometimes i have time to do all controls and sometimes not. BUT, about 90%-95% of time, using only OHLC is enough ; but not all time. That's why backesting my trading system is hard. I need 2 timeframes in the same time to be correct at 100%. I made many manual tests before trading in real. It was a hard work. As said Lautentb perhaps i must to go on new technologies as Neuroshell to overcome my problem of backtest. A lot of work since i win money by using this system everey day. After, i use proealtime datas and it seems that they have sometimes bad ticks...
Results : HCBK : 12.53. - 2.4469% MRVL : 10.27. - 0.895% RIMM : 67.79. + 0.1727% JNPR : 21.72. SCHW : 17.42. - 0.2543% 2009-04-28 : HCBK : L5 = 2,0992% MRVL: L5 = 2,6144% RIMM : L5 = 2,493% JNPR : L5 = 2,2844% SCHW : L5 = 2,2117% Have good trades... Overall %result : + 129,85% FSLR: L5 = 2,8512% SKF : L5 = 4,4556%
I observe, using data from bigcharts.com (graph attached) that SCHW opens at 17.68. Your L5 value is 2.2117. I calculate the buy target price as 17.68 * 1.022117 = 18.07. I notice bigcharts.com reports a high price for the day of 17.99. I do not calculate a buy signal for SCHW today. Please explain how you show a trade for SCHW today.