I’ll definitely post if I do go back through and check the RRR results. I just do this manually, so there might be an error or two. I do check my work before posting, however. Very tedious!
I simply closed my initial long as if the Rejection Rule was triggered by a new trade. I treated it as if I was flat until a subsequent candle traded above the rejection candle (the one marked with the red X), and a new long position was entered using the last bar technique for the stop.
I actually did two backtests on GBP/USD; the first one was the application of the Rejection Rule only on the entry (closing the trade only by Last Bar, Bishop, or 10-day low), and a second time removing profitable positions (like my example above) if the Rejection Rule was triggered as if on a new trade (and of course, using the 5 day condition). The results were significantly different:
Rejection Rule only on entry (2008-2013): +2297
Rejection Rule continuously applied (2008-2013): +3839
I did the same on EUR/USD however, and the difference was a wash. It seems to improve the performance of choppier pairs, and not affect the performance of smoother pairs. I’ll have to do more backtesting.