Yes db, I didn't want to raise experience as a difference because Soph is obviously trying to learn, and I didn't want there to be any uncomfortable feelings as a result. But thank you for doing so. But there are at least some other valid differences that could have created the difference. Since, as you said, the observation wasn't interesting, I am surprised you commented at all. I guess the urge to post was overwhelming.
This backwards approach is seductive and can waste months of a novice's time. If you want to begin to put together a strategy, take a day like today, figure out where you'd enter and where your stop would be, then forward-test it over 30 days' worth of charts. Looking forward is not quite so easy as looking backward. Also be aware that the market teaches bad habits. Once it has taught you, for example, to maintain wide stops, say 10 pts, it will then hit them with regularity, teaching you to tighten them, at which point it will begin to trend, which you will miss because your stops are too tight. --Db
Thanks for the input, DB, it's appreciated. Since I followed the trading live using 15-minute charts, and posted a query during real-time trading (p. 27) trying to get into the minds of the traders while the scenario was unfolding for real, I really did attempt to avoid any "backwards approach" traps. At least I think so! But not unlike some of your earlier posts in this thread about hypothesizing a certain potential win/loss and risk/reward ratio for this sort of method, I do think there is a place for "coldly" looking back at the charts and trying to make rhyme or reason out of whether a method works to begin with, and how/why certain decisions get made at which points. There has to be a place for both backward and real-time/forward testing, no? Anyway, I definitely don't want to take this valuable thread off-course, so I'd enjoy discussing this further via PM if you'd like. Thanks again.
Asking questions about the process in real time was a smart thing to do, and, yes, one does have to go back for chart review in order to come up with some sort of hypothesis in the first place. The danger is in thinking if only I'd done this or if only I'd done that, then getting lost in a maze of what ifs. Therefore, rather than hypothesize that a method or methods such as those we've been discussing "missed" what now seems should have been a profitable day, it becomes necessary to come up with a set of rules that not only would have enabled entry but would have kept the trader in the trade for whatever length of time would represent maximum profit. It's entirely possible that one might be able to do so and yet maintain the simplicity that the thread is all about. Natalie, Mike, Cathy and I are all doing different things, yet we're all trying to keep it simple. The thread isn't about one method, but simplicity. Therefore, I can't imagine why anything about simplicity would be off topic or require private messages. --Db
Some time ago, I briefly used a method where I took yesterday's close on the FTSE Future and then calculated a range either side of it. I had observed (and collected a lot of statistics to support this) that the FTSE often gapped on the day and then later in the day would turn itself around. I found that if the FTSE came back into a range around previous close, that there was often a much better than 50% chance that it would Yield up some points for the taking without taking much risk. So I looked at ways of increasing the probability. It didn't work every day, because not every day applied, but the basic premise made money on the days when it did come into play. I worked out the range based upon avarage range of last 20/40 days, the open relative to high/low, and frequency of crossover of the prev. close line. It was also based upon a statistical measure of probability for each 1 point increment above and below. I would end up with a table of probabilities calculated daily something like this. +/- 10 42 +/- 12 45 +/- 14 48 +/- 16 53 +/- 18 61 +/- 20 58 +/- 22 57 And it was based upon the probability of taking 20 points from the FTSE on a cross up or down into the range. I also worked the table through for probabilities for much higher returns too and sometimes the volatility in the market would yeild them up again without taking undue risk. I guess something similar could work for the SP and NDX though I havn't looked into it. It's basically about taking minimal risk but catching the days when there is a pretty even price distibution either side of previous close... I probably havn't explained it very well, mainly because I havn't used it much in the last 3 or 4 years, but it's another idea to think about. Natalie
Girlpower, That last post needs to be moved to a new thread called "The Importance of Brilliance." What happened to "when Moving Average A crosses Moving Average B then enter, if not then take the kids to the park?" Just kiddin' you and complimenting your range Girlpower.
FYI I found a link to a similar breakout system that is a bit more sophisticated than what I am doing. But still simple. Check it out http://www.tradingeducators.net/3.htm Mike
I coded this in to Excel when TriPack first put it on his Gems thread. Of course, I haven't tracked it at all. The methodology is interesting in that it attempts to help capture a profit from the "nemesis" of range breakout trading, that being the false breakout. It is worth reading and understanding.
Certainly a very good idea. I did not realize that this thread became so large. I promised to contribute some of my thoughts to it in a more complete form, but still did not find enough time to do this. I feel guilty about it, but I am too busy working on both some of my mechanical systems that keep doing quite well and my discretionary system that would incorporate the 30 min breakout, but in a more anticipatory way, as I mentioned it previously. However, here is some exercise for you, Natalie, and perhaps Mike, if you want to check this: verify statistically how often after a long first 30 min bar (at least 7 ES pts) a reversal at 62% FR (fibonacci retracement) happens and takes you in the original 30 min bar direction. Today we had yet another example of this phenomenon, a pullback in ES took us to the 62% FR level and then we proceeded in the same direction scoring at least 6 points. This is not so infrequent from what I have noticed. However, it's good to have some additional reasons to enter at this level, namely you want to see that the pullback is weakening and today it was the case indeed. The stochastics were very divergent in 1 min chart, you could see a triple divergence and then a realiable candlestick pattern indicating a top. Well, I will check out later what I missed here, but in part would not like to do it because I am afraid that I may get too many ideas and will not be able to finish what I am working on. My discretionary system will incorporate the 30 min breakout in my anticipatory way and some other elements that I found useful and apealing. I hope to have enough time to elaborate on my system on this board in due time. Thanks, wally_
as this is is the simplest way to extend your 30 min breakout system so that it scores points also when in its original form it is not capable of doing it or at least not capable of doing it well. My crude observations suggest that pullbacks of this type happen at least 25% of the time for longer 30 min bars and so are certainly worthwhile exploiting. You may want to look into this to improve your system performance.