| | | I added a new feature to my context sensitive RSI and tested it on weekly natgas. I'm thinking some NG traders would kill for this algo <g>.
The new feature is the thin red/blue line you see at the bottom. Without looking at the chart and only looking at the RSI, I marked every occurance where the RSI was either overbought, oversold, or at the middle *and* the new indicator was red.
The idea is that the marker should point to new outsized moves over the next several bars (up or down). I'd sell in overbought, buy in oversold, and buy a strangle in the middle and dump the side that didn't work (or go with the RSI). Out of the 11 signals it gave over a 4 year period, 10 made money and one broke even (if you held it, or lost a little if dumped it).
BTW, this was not developed with NG in mind. I was looking at SOX when I was working on it...and these two are as different as it gets.
PS The very last one is not quite by the rules. The signal went red and about half a bar later the RSI went oversold. So I included it b/c I figure if a human was looking at it he would act that way. But even excluding it, 90% win rate is good.
Same mindless strategy for 2010 - 2013.
10 signals - 2 losses, one break even, 7 wins. Presumably adding some intelligence would improve the batting average, but you'd never know with some traders <G>
BTW, this period - and especially 2013 - is widely acknowledged to have been the most difficult and unusual trading environment.
 |
|
