Brookelise, Richard, David. There was a real convergence for me when I read your discussions about BBB, system testing and MSWIN's limitations. I appreciate what you've been sharing with everyone. I had a simple idea when I saw Brookelise's queries about BBB. It was a lightbulb for me. Perhaps others will find it relevant.
Assuming I'm using the BBB scan in my nightly activity, I will focus my attention on a stock when BBB triggers me to do so. I created an MSWIN indicator, based on the discussion:
BBB:
(C > 1.05*Ref(C,-5)) AND (V > Mov(V,66,S)) AND (C > 0.85 * HHV(C,260))
Displaying this on a chart (shows as a bar in histogram form), lets me see where BBB fires. It is interesting to see where BBB brings me into a stock as it climbs the LHS of the mountain. What I see wasn't quite what I first expected, but it makes sense when I see it, and I gained some insight just from this step!
I can make a basic "BBB trading system" (as a basis for comparison with other systems which I can refine) once I define a simple exit strategy.
For example:
Using a simple exit rule (it's not very sophisticated).
Define Indicator "Trail-up" as Max(LLV(C,10),.9*HHV(C,10))
This is a "trailing-stop" at Higher of 10-day low and 90% of 10day high. (This would keep me in until I give back more than 10% or move lower than 10 days ago. This is sort of "buy and hold", but get out if price goes down. By temperament, I am most attracted to systems that are short to medium term, but this system gives me a reference for comparison).
Enter Long:
Fml("BBB")
Exit Long:
Fml("Trail-up")<Ref(Fml("Trail-up"),-1)
Using MSWIN's system testing, I can use BBB to enter Long, and test various exit strategies. I can also use other entry strategies which I would use after BBB signal, but before BBB exit.
While system testing with MSWIN, I can test multiple systems, but only with one stock at a time. I've been envying the power of WOW (until I read a recent message that outlined WOW's testing shortcomings. The grass is always greener.)
I question this "pick one stock and try" approach, but MSWIN drives me to it. I find myself checking entry and exit points to evaluate each system, rather than comparing $ - because there are so many reasons why different systems can't be compared apples to apples.
(ie. amount of data is small for 250ema's, can't program subtleties that I'd see when I checked key indicators, etc.)
I intend to explore with the approach of comparing to a BBB-system - and use this approach for a while, and see if it helps me understand things better. (I think it will drive me to focus heaviest on exit strategies, rather than entry).
My sincere thanks to all of you, and others. I'm learning a lot from everyone here. I'd appreciate any comments you may have on this approach.
Paul |
