Dave Evans suggested a short-term and long-term exits. The short-term one used these indicators: MACD (8,17,9), Parabolic SAR and StochRSI(14). The long-term one involved looking at Dahl and DNS.
So I suppose the short-term one could be expressed like this:
fml("MACD (8/17/9)")<0 and
fml("StochRSI")<70 and
SAR(.02,.2)>C and
slope(SAR(.02,.2),3)<0
The long-term one could be:
fml("DNS")<3 and
fml("Dahl's primary trend")<0 and
slope(fml("Dahl's Primary trend"),3)<0
Dave might formulate these differently, of course. I'm just guessing, in a way, based on what he said at TAINJ.
The short-term exit gives a lot of signals. You'll get whipsawed during periods of low volatility if you're not careful about what stock you're getting into and when. Once you find a real breakout, though, the system seems to pinpoint highs and lows.
The longer-term system gives fewer stgnals, but in some cases, still quite a few. Maybe others have some better suggestions.
I've been thinking of just using some sort of moving average for exits -- for esample when the 9-day MA crosses under the 21-day MA (an exit Henry Brookins mentioned). I haven't tested this yet, though.
I've also been thinking of turning the entry indicators into a binary system. But Dave's system really depends on that first indicator, StochRSI(14), it seems. That's the linch pin, it seems. You'd have to meet the first condition, that StochRSI had crosssed back up over 30.
Brooke |
