Couple of questions to clarify; GLW good example

[prox]

Using the 6/4 MA channels as an entry trigger , which specifies (for a long) a close above the upper MA channel and then a takeout of the high of that bar. My question is: What are the exact rules that should be applied on continuing to use a bar for taking out the high ? Say, if I had a large first bar close and then we have 2 or 3 inside bars that close above the upper 6/4. At what point do I start trying to take out the high of one of these following bars and give up the first bar ?


Also, another question on the 1.4 pullback on the oscillator. If just a couple of bars are slightly taken out by the 1.4 , when do you decide to give up on the pullback ? I've noticed many examples where the 1.4 doesn't hold, but the trade still turned out to be a winner (granted, with the gift of hindsight).

[MR]

Over the years several 6/4 DMA and Regression Trend Channel breakout methods have been developed, particularly with the advent of the XTL to help improve the breakout interpretation. The real answer is, it depends more on the real risk/reward ratio involved in the setup, the price movement/scale already involved, and the quality of support/resistance behind the setup that determines which breakout bar you take, and possibly other variables not identified here.

If you are using a breakout technique to either confirm or trigger a trade, you need to have at least one 'close' above the breakout bar. After that it is open to healthy debate if you need more data bars to better substantiate a signal.

I don't normally do this because I am a very private person, but let me give you one real example of a chart I posted to myself yesterday in my 'Marc Rinehart Group.' I don't have time to trade right now, but I still want to keep my trading skills sharp, so I am posting trade ideas to myself so I can track them. (disclaimer: The M.R. group in File Share is setup as 'closed' only because I want to sometimes be very specific with my trading ideas without worrying about strangers misinterpreting public comments or taking trades based on my personality profile and preferences, which is not always wise for others but works for me and a couple friends.)

Below is a copy of the GLW daily chart posted yesterday. This isn't a 6/4 DMA breakout, but it is the same breakout logic.

I was using an eSignal Scanner and found this breakout. In this case I didn't even wait for the breakout to 'close' (GLW was posted at Mon 01:28:49 PM) because I determined it was quality, and a good risk/reward ratio if initiated when posted.

The point of all this is under a different setup I might have to wait for a couple more data bars before triggering a trade. It is a conditional issue based on the quality of the setup and the profit potential basis the risk involved.

As for your 1.4 question I would interprete it the same way.

People I work with tell me if it breaks 1.4 forget it. But I don't agree. I will allow for some very slight deviations if the 1.4 rule if the Type 1 setup has other real quality behind it and the risk/reward is better than 2 to 1.

Again, this chart shown below was posted to my private site yesterday. It is not intended as advice to anyone, but is being shown only as an example of how I personally am willing to trade a breakout data bar. The only difference is in real life I would have done 1,000 shares w/GLW, not 100.

[MR]

Oh, buy the way, here is an update of that GLW daily chart shown below. It is one day later and it shows the change.

I didn't even look at the 1.4 but it also fits what you said because it did technically get violated by a small fraction.

Again, these are not ideas necessarily for new people, but they define how a more experienced trader might be
willing to make adjustments within the benchmark rules.

[donv]

I agree with Marc, it depends on many factors. Personally I prefer to assume more information risk and less price risk. Regarding RTC and 6/4 dma the longer you wait for confirmation of entry (more price info)the larger the stop and additional price risk you take. The key is finding the right balance given the specifics of each trading opportunity and not get too focused on a strict rule for each trade.

The guidelines suggested by AGET are a good starting point and will keep you out of trouble but it is always better to refine these depending on your personal style and risk profile.

As traders we can't have it both ways, waiting for more price information to confirm entry and the best price and tightest stop.


donv

[prox]

Thanks for the replies. The need for an exact set of rules would be most helpful in backtesting and to avoid having to think in the line of fire of live trading.

You are right about the added risk with the confirmation through the MA channels. When using a MOB as a price target, I find it very difficult on an intraday chart to even achieve a 1 to 1 R/R on the MA channel entry. The RTC entry is of course, more aggressive but often inconsistent and subjective. Now, that you mention risk/reward .. buying blindly right at an ellipse projection might not sound like such a bad idea.

[MR]

prox,

I apologize. When I answered your question earlier I wasn't thinking about backtesting needs for someone attempting to define mechanical system parameters.

Having read and thought over you last post, here are some ideas to see if they can help. I am very creative, can think of lots of ideas, I just don't know how to program and test them out.

Here are some breakout backtest ideas....

Start a test with 1 breakout bar. Run tests as high as 6 to 13 data bars after a breakout-- but, be warned, the higher you go the more you need to add into an equation some sort of a buy/sell a percentage of the next pullback parameter because it grows more risky the further away it trades from the initial breakout. You need to also keep factoring into a breakout backtest different risk/reward considerations? Maybe attempt a secondary, backup trade trigger, initial stop loss methodology. See if this helps filter better results, particularly if you desire a strategy incorporating multiple buy/sell entries?

Make sure you also build into the test some sort of a risk/reward parameter. I am not a programmer, but if you can program a more sophisticated test, one idea is to build into an equation some sort of an auto 'cross-referenced' or 'cross-linked' integer test which accesses shorter time frames for key information? Attempt to define money management considerations based on the shorter time frame information. For example, if the trade setup is based on an hourly (60 min) chart, you may attempt money management parameters defined in a 30, 15, 10 minute, or other shorter time frame variable?

Hard to explain, but my instincts tell me if you experiment with these ideas you will capture better risk/reward ratios and results. You should find quicker trade triggers and tighter trailing stop tools when using things like a MA, DMA, or RTC w/high Pearson, on a shorter time frame.

(Oh-- by the way, please excuse my digression-- but do you know what 'DAM' is? It is either a dyslexic trading a DMA, OR it is a DMA setting being used for a stop loss that fails and results in a loss.... sorry, I'm both dyslexic, have a poor sense of humor, and try too hard to be funny.)

Back to the subject-- if you have the ability to program a sophisticated backtest, try incorporating an AGET XTL breakout into a test? For example, test using 1.5 and -0.5 XTL breakouts as triggers? Backtest the XTL with different settings? For example, use a lower setting if you want to test for quicker trade parameters? Use higher XTL settings if your breakout objective is to attempt staying with a trend longer?

I don't know if this helps. These are mostly just ideas because I sincerely want to help you generate good backtesting results.

If you generate promising results, remember your new buddy and send me a copy!!! I promise to try and keep it a secret!

Take care, best wishes, good luck!

[MR]

Follow-up on the earlier GLW daily example. Scroll down to see the exact post on 01-06-2004 11:23 PM--