Quote:
|
Originally Posted by tim_nn
Yes. In a sense, BunnyGirl has (among others).
See her long and interesting thread called "WMA Cross". Similar principle of MA crossover plus filter after the cross (in her case 30 minute charts with 5 & 20 WMA cross; but a lot more to it as well and a _highly_ discretionary system).
The underlying theory of such systems, in a nutshell, is that by being willing deliberately to miss a significant proportion of the trend, you catch the smaller residual proportion of it with _greatly_ increased reliability.
|
You are, of course, right. I have read through that thread, all in one afternoon. To be honest, it did my head in. Although I would never dispute anyone as gifted as bunnygirl, I also think in order to get the hang of the discretionary side, one have to start from a simpler beginning and gradually work in the more complex filters and decisions.
What I was wondering was if anyone has tried to implement - maybe only in testing - this system, though perhaps with other parameters, so to get the same realiability, though of course for smaller profits - and thus exposures - on a smaller timeframe.
If noone else has, maybe I will find the time to do so myself. It seems worthwhile. Also, since this is a very objective system, maybe it is possible to do a backtesting write-up for eSignal (what I use)... I shall have a ponder.
Once again, thanks to all for being so willing to share your knowledge. It really is great reading!