Quote from benwm:
1 1 min data is fine - ideally 5-15 sec bars but I'm just trying to get an idea of where I would have got fills on a possible strategy
My experience with backtesting in these timeframes indicates that the result is noise. Especially nowadays when these timeframes are ruled by HFT robots. I have yet to see a trading system developed on 1 min data that won't go bust after a week's time. Most do immediately. By that I do not imply such systems do not exist but backtesting is unlikely to produce one.
