As far as I know, the posted data involves taking every trade. Whether or not Quah takes each trade is irrelevant. The posted system data is a result of taking each trade. I was considering only one of the two contracts. You are right in that respect then re: 6-8 contracts, as far as potential occurrence. Although, combining the data to smooth the equity curve was an after-the-fact decision. Each method was developed to stand on its own. Perhaps, then one should consider each trade as the system. I mean after all, the time frame by which to post profits is not the system. Okay then. Btw, how about that guy who posted that there would be a squeeze this afternoon. Not a bad call.
In case anyone's interested, I've been looking for a way to trade those points at which there is no clear signal as to direction. This need not be a case of one line not being above or below another. They may be roping. But I do know that when I have to guess, the probabilities slip back to 50:50. Therefore, if I don't see a clear indication of direction at the bar at which I'm supposed to enter, I wait until such an indication is given. Sometimes this is the next bar. Sometimes it's several bars later. In any case, this is just one avenue that can be explored. The times in parentheses are those at which entry was actually made. 0933 (0936) S +3 0935 (0936) S +3 0938 (0939) S +3 0943 (0946) SAR 0 0951 (0953) S +3 1004 (1010) S +3 1025 L +3 1059 (1113) S +3 1154 L +3 1323 SAR 0 1547 Not taken 10W 2L --Db
For ES: +3.50 points 16 trades, 10 winners, 6 losers For NQ: +7.50 points 15 trades, 9 winners, 6 losers http://home.carolina.rr.com/quah/quah1web.htm
If I have to guess as to direction, the indication is not clear. Which is another way of saying that it doesn't really matter. What is clear to me may not be clear to you. If you're unsure as to whether the indication is up or down, then you pass until it's clear to you. --Db
Well, what I mean is, what do you consider clear. Attached is the shot from that period...the line is on 10:08. Why did you not choose that time to enter? For that matter, the 10:06 was also well below. zorak
I've always like 1m charts, and when I ran across this thread and learned that Quah used 2m charts, I began running both to see if I could adjust his system to a 1m chart. I won't go into all that because it's not important. But the upshot was that I was running 1m and 2m charts simultaneously with various sto settings. And I noted that the 1m and 2m charts frequently gave conflicting signals. In kindergarten TA, we learned that the macro trend is the more powerful, and that the micro trend must bend to it. We also learned that macro trends begin in micro timeframes. There's an inconsistency here. So rather than try to reconcile those inconsistencies, I looked for those occasions when both the 1m and 2m charts were pointing in the same direction. In other words, if the 1m chart was giving a short signal and the 2m chart was giving a long signal, I waited. Which is why some of these entry times are so much later than one might expect. Though I've been keeping notes on this for more than a week, that's really nothing. And I've only begun to keep formal notes. But given the number of posts regarding unclear and conflicting signals, I thought this might provide a clue to an answer for somebody. I want first to avoid the SAR. And most of all, I want to avoid the dreaded DW. --Db
Very interesting work Db. Across my 3 screens I routinely have the 2 - 5 -15 - 60 - D - W time frames so I see the micro trend bend to the macro trend all the time, and I see the macro trend begin in the micro trend all the time... every time. When a friend once suggested that I use the next shorter time frame to enter my trades to get a head start, I pointed out that a shorter time frame can break and will break down several times before the next longer time frame finally executes. Additionally, the two time frame/ same direction method is very powerful. While each time frame operates independently from any other time frame, when trading in the direction of the next longer time frame, the move tends to be more powerful. Your work demonstrates this very well. Did you limit your observations to the fib times in the original method? And since you have put a significant amount of work into this project, I'll understand if the answer to the next question is no. Are you willing to post the stochastic settings that you used to generate the trades you posted today? Thanks either way.