Strategy backtests are inaccurate when running multiple days at a time

[irochester]

Hi,

I'm noticing a pretty big disparity between results when backtesting a strategy multiple days at a time vs. a single day at a time. Single day backtests are accurate, multi-day backtests are not. I am having all my variables reset EOD. This is making me wonder if the problem has to do with how much data is trying to be run within such a short amount of time? Is it possible to have a strategy disable/enable every "day" within a multi-day backtest? Getting this solved would save me an immense amount of time backtesting one day at a time.

[NinjaTrader_ChelseaB]

Hello irochester,

What specifically is inaccurate?

The values used in the conditions that are trigger the orders is incorrect? What are you comparing this to, to know it is incorrect?
The order fills are not within the bar, (or filling on the tick provided with intra-bar granularity)?

First, have you implemented 1-tick intra-bar granularity to improve backtest order fill accuracy?
Below is a link to a forum post that discusses and includes a link to a reference sample that demonstrates.

https://ninjatrader.com/support/forum/forum/ninjatrader-8/strategy-development/94098-isfirsttickofbar-vs-onbarclose-for-backtest-live?p=773377#post773377

If so, we would need output showing what has changed. I would recommend writing to file as this could be a lot of output.
Below is a link to a forum post that demonstrates writing information to file to compare real-time to backtest, but the process would be the same to investigate.

Please output the time, all values used in the conditions, and the order object from OnOrderUpdate() and provide the output text file with your reply.
Then please let us know what values are incorrect, the correct values you are expecting, and why.