Different backtest resault

[stibflo]

Hello

When I backtest a simple strategy, I get different resault if I have added a second timeframe, even I don't use the second timefram to anything in the strategy.
I tought the resault would be the same as long as I don't do anyting else than add a second timeframe.

I have attached 4 screen picture. 2 with backtest resault, and 2 whit the code. The only difference in the code is that the first sampel has:
Add(PeriodType.Minute, 405);
and the second has:
// Add(PeriodType.Minute, 405);
else the code is the same.
Anyway, I get very different backtest resault.

Is this correct, is there a software problem, or have I forgot something that need to be in my code?

[NinjaTrader_Austin]

stibflo, I just tried testing this out and for some reason I could not get any results when I tested YM 12-10 with the settings you have in the screenshot and with the secondary series added. I then changed the session template to the Default 24/7 series and just like you reported, there are differences in the performance. I will forward this on and have someone reply to you on Monday.

[stibflo]

stibflo, I just tried testing this out and for some reason I could not get any results when I tested YM 12-10 with the settings you have in the screenshot and with the secondary series added. I then changed the session template to the Default 24/7 series and just like you reported, there are differences in the performance. I will forward this on and have someone reply to you on Monday.

NinjaTrader_Austin, I'm still waiting for the reply

[NinjaTrader_Austin]

I'm not sure how the reply got overlooked but the difference here is the backtester waits until there are enough bars (usually 20) for every series before starting the backtest instead of just the one series.

This means if you add a daily series, then that is potentially up to 20 extra days it will wait until it starts trading in the backtest.

[borland]

This means if you add a daily series, then that is potentially up to 20 extra days it will wait until it starts trading in the backtest.

This seems like a significant shortfall with NT if used to autotrade.

Why can't NT simply, gather up more historical data, if while evaluating formulas using technical indicators, it sees a look back for which there is insufficient data?

[stibflo]

Ok, I too saw that the difference got bigger when I added a big second bar series, so what you are saying, makes sence.
If I understand correctly, in my example, when I test with a 4 range series only, the backtest start making trades after for instance 20 bars og 4 range, but when I add a 405 minutes series, even I don't use it, it will force the backtest to start after 20 bars of 405 minutes instead. This will make many trades in the start of the backtestperiod unawailible for the backtester, but the remaining trades will be equal in both backtests.
I'm going to make a test to see if the remaining trades are equal, but it makes sence that this is the answer.

Thank you for your reply NinjaTrader_Austin.