[...]
A kind of a self-fulfilling prophecy, yes. However, there was a 1w EMA downcross at the end of october 2013. How did that work out?
(a) Maybe I missed it, but I don't recall TERA specifying exactly which weekly EMAs crossed over, so it's hard to tell if you found a bad signal or not.
(b) That said, even if you did, keep in mind, we're not playing 'mathematical proof' here, where one counter example is enough. We're playing 'is my stochastic system better than a monkey throwing darts'. And if a signal is, say, right 3 out of 4 times, and assuming you have some form of risk control in place (basically something that allows you to let profitable positions run, while cutting short unprofitable ones), then there's a chance you have a system that beats the dart throwing monkeys*.
* Details very much matter of course, so the above is nothing but the most high level description of why history based trading works even if individual signals go wrong.
