We may know the next bar Close necessary for a BB breakout. Details at http://groups.yahoo.com/group/amibroker/message/48394
mr tso is undoubtedly an astutue technical analyst but i could never figure out how his posts could be practically applied. what would be the strategy in this BB breakout out method?
I think he means that if you have strict technical requirements for a bb breakout (see bollinger on bollinger bands for specifications), you can compute the close of this bar necessary for the breakout to meet the spec. IOW, it's a way to do MTF analysis intra-bar (heading down a timeframe, instead of the usual up). IOW, if the necessary close is 1020.75 on the ES, and the ES is already at 1022.25 and moving fast with 20 secs left (lets say you're using minute bars), you don't have to wait for the close of the bar (given that waiting for the close could cost you an additional .5) Basically you're solving an equation-- neat idea, but I bet it can lead to a lot of jumping the gun if you're not careful. --laz
Suppose you trade daily AMZN. Suppose also that you go long when the close penetrates the upper Bollinger band. The formula warning of Sept22 [point X] was : "The next AMZN Close should be above 49.043 [+3.314%] for a BBandTop cross" On Sept23 AMZN opened at 47.41 and closed at 50.44 and a cross was detected. The trader who uses EOD would know the cross *after* the session and would go long by the next Sept24 open, without any other alternative. [AMZN may open higher today...and you should hope to open higher since your system goes long by the upper breakout !!] The trader who uses my formula has the ability to find a value between 49.034 and 50.44 *during* the Sept23 session and hope for a better profit. Another advantage is that he may place his order *before* the Sept23 session, since the alert level is already known by the end of Sept22. Another valuable info : When you see for a stock that you need a +18% for an upper cross and the past history of this stock has highest Roc +7.6% [for the last X months] you may suppose, at low risk, that the upper cross scenario is almost impossible. Something else will happen and it is better to exclude the impossible. As for the RevEng mechanism, see also http://www.traders.com/Documentatio...2003/Abstracts_new/Siligardos/siligardos.html
Another signal that loses money when backtested. Exit was a simple Exit on Low or high of the bar and take profits when the price crosses within the band again.
There was no suggestion for the system [although I use it] The word "suppose" was enough to make it clear. Well, *if we suppose you trade AMZN and *if we suppose you go long when..., then the text explains the benefits of the method. Every BB trader uses the bands from his own point of view. My method gives the next bar threshold for an upper or lower cross. I hope it is more clear now.
As for your comment, I don´t know the backtesting you refer, but here is the true numbers. The first system buys at the upper breakout and sells, say, 8 days later. Buy/Sell are at Open, delay+1, commission 0.5% [all stops disabled] The result over the whole N100 database from Jan2003 till now is a poor average profit+1.17% My method changes this result to the [interesting] +14.63%. Examples without the method BRCM +57%, JNPR +48%, AMZN +14%, MSFT -10%. The whole N100 market was at +1.17% average Example with the method BRCM +89%, JNPR +78%, AMZN +24%, MSFT -3%. The whole market was at +14.63% average The more important is that my method had 100% success [to add some profits or reduce some losses]: All the N100 stocks have had higher final equities for the above described trading system. Without exception. The filter request e1>e2 gives 101/101. The increment of the final equities was from 4% up to 57%. [Full and detailed codes are available for any comparison] Enjoy the equities improvement, the best password for a humble method.
It just depends on how you work around the Bollinger Band breakout. Bollinger Breakout Entry... then Bollinger Breakin Exit has negative test. Simple as that. But TA traders usually have other filters like what you wrote with x days exit, etc. etc. so it's different. Also, depends on the data set you use.
Just for the reference, AMZN opened at 50.35. The EOD BB breakout trader would buy at this price. My Buyprice was 49.03.