If somebody has an experience of copying losing strategy with reverse, please post your opinion in this thread.
About 5 years ago, there was a strategy I found explained in a book (by Kathy Lien and Boris Schlossberg) which I thought probably looked
really bad, and it seemed to be hopeless on back-testing by eye, so I did actually try a lengthy back-test a little more formally, switching the long entries to short entries and vice-versa to test the opposite (using software called ForexTester-2 over a decade's price data, making plenty of allowance for slippage and commissions, and so on), and sure enough it really did make pretty steady profits with a PF around 1.3, and frequent trades, with absolutely no substantial drawdowns at all.
I'm embarrassed to say that I didn't quite have the nerve to try it live, just in case the markets had changed (and because the reverse strategy's PF of 1.3 was actually lower than everything else I was doing at the time).
But it was certainly a system that
was bad enough to make it profitable to reverse the strategy and trade it.
I always remember that when I see forum posts claiming that the opposite of a losing system will always be a losing system too, by the time you factor in all the dealing costs, because I don't believe that's true.
There was also an academic paper published somewhere quite a while ago, which purported to show that fading a moving average crossover was profitable, but I can't remember the details now (I saw it refered to and cited in a textbook but can't remember where, now. As I remember one of the MA's was displaced to the right by about half of its own periodicity, and the stop-loss and targets were ATR multiples. Something like that, anyway).
(There was also something published once purporting to show that the original "Turtle methodology" could profitably be faded, at the time, on a range of instruments. Which I can fairly readily believe.)
So the overall idea perhaps isn't as outlandish as it appears, anyway.