Hi everyone,

during backtesting my strategies on DAX 15mn from 2006 to the present day, I keep getting weird results such as huge losses on single trades that should simply not happen with stops. Zooming in I’ve witnessed that the data is simply weird in some places: sometimes bars are just flat lines with the same value for open, close, high and low, sometimes there are missing bars, creating huge gaps between two bars (with 3 intermediary bars missing, for example). This throws the backtest completely awry with not even stops behaving like they should (for example it will open a position on the close of the 9:00 bar, which is to say at 9:15, and close it on the open of the following bar which should be 9:15 but is really 10:00, with the 9:15, 9:30 and 9:45 bars missing, with a 80 point gap, and not trigger the stop loss at, say, 5 points).

Has anyone encountered similar glitches? Is their a way to remedy it or should I just accept that the data is incomplete and that I should make do with what we have, using backtest to assess the relative performance of different strategies and not their long-term absolute performance?

Thanks for your help!