Hello Darrell,
I am testing a strategy using a few of your indicators that “looks” to be promising. I just started testing, so far two months worth of data on the EUR/USD. Both months had the same win percentage. In your opinion, how far back should a person collect data to determine if a strategy works and how many instruments should they test? I personally only plan on using this on FOREX, so how many pairs would you test?
Thank you
I would have to know about your system to tell you what to test exactly. Otherwise I would say you would want to do 6 six months back testing and do it on several FX instruments that are non corresponding pairs. Meaning do not do ALL your testing on only JPY pairs or on ly on USD pairs etc, mix it up.
Test as many as you want.
At a certain point though you will get correlated markets having the same results
The more data the better
20 days minimum ideally at least 60 trading days and if possible up to a year to account for different seasonal cycles.