iarudi (Arjun) proposed an EA based on Super Trend here, but I was curious about a simpler approach.
I hand-backtested two simple ideas on H4 charts, from July 1 to Oct 19:
* When ST reverses, enter the next time the market pulls back behind the DP.
* When ST reverses, enter immediately at the close of the bar that caused ST to reverse.
Both approaches close trades at the close of the bar that causes ST to reverse.
Entering at the DP sometimes gets you a better entry, but sometimes you miss the entry because the market runs away from you. In this test, the simple "enter on close, exit on close" approach worked the best.
I tested this on 13 pairs: AUDCAD AUDCHF AUDJPY AUDNZD AUDUSD CADJPY CHFJPY EURAUD EURCAD EURGBP EURJPY EURNZD EURUSD. (I got sick of testing after that.
)I really didn't expect much out of this. Super Trend-type indicators have been around forever, and hundreds of people have tried to apply simple rules to them. It never works.
But this did.
The market would frequently go sideways and whipsaw the thing to death. But there were many times when a strong trend would kick in, and in most cases that was enough to keep you in profit. Only 3 of the 13 pairs showed a loss for the July-Oct test period.
There was a really nasty sideways/dead period in late July / early August, and it kicked the results in the teeth. But other than that these two simple approaches worked pretty darn well:
Even with the drawdown in Jul/Aug, it averaged about 80 pips per month, per pair. Nothing to take the world by storm. But if you trade 20 or 25 pairs...
Am I fooling myself? From what I can tell, ST doesn't repaint or future-peek or do any other illegal or immoral acts. I *think* you could do this in realtime.
I'm not the EA guy so I didn't try to code this. (Though I might try it in Tradestation.) But for such a simple approach, it seems to work remarkably well, at least in the conditions in the last 3 months. You could almost certainly improve on these results with some good trade management, though the obvious things (BE stops, trailing stops, etc) don't help much because the whipsaws that kill you never get much open equity. But I'd bet it could be improved.
