NT7 Strategy Analyzer Process Confirmation

[successfulmike]

Hello NT Support or any other experienced trader/developer,

I'm just curious about one thing.

When I download all the data I'd like to backtest and optimize for my strategy (i'm now in Market Replay mode), is it necessary for me to replay ALL that data from start to end physically before running the optimize function on my strategy?

I'm only asking this because sometimes it appears as though the results change if I don't go through all those days' data. Eg. I initially had 2 weeks loaded, run my optimize and got my results. I then load another week and then jump to end of the last day in that new week, run optimizer, and get my results. When I then replay everything to ensure the results are correct, the results are different, and if I run the optimizer again after this, the results are different there as well now.

Please help with my understanding on this. Thank you

[NinjaTrader_PatrickH]

Hello successfulmike,

Thank you for your post.

When running the Strategy Analyzer before running Market Replay, you are running over the historical data itself.

When running the Strategy Analyzer after running Market Replay, you are running over the cache of data created for historical by Market Replay.

Historical data that you download from your data feed provider is not the same as Market Replay data when viewed historically after running Market Replay in most cases. This is due to Market Replay being a recording of the real-time data and not a historical database that you download data from the provider.

When we replay Market Replay data we are creating a historical cache of data from the point we begin replay. We can then see differences int he results of a test that would be run over that historical data and in this case historical cache of data.

Please let me know if you have any questions.

[successfulmike]

Thanks Patrick. So just for my clarity, running Market Replay is more accurate and realistic than running strategy over the downloaded historical data, correct?

I'm just trying to determine which results I should focus on more - Strategy Analyzer over historical data or Strategy Analyzer over downloaded and completely replayed Market Replay data.

Thanks,
Mike

[NinjaTrader_PatrickH]

Hello Mike,

Thank you for your response.

Running a test over Market Replay data will allow for real-time bar movement for calculations of your conditions, order executions, and for the bars to build from (helpful to Renko for example).

Technically while Market Replay data is real-time data, running a backtest or optimization means you are running over closed bars which were already built. So executions would be on the close of the bar, as would the conditions when they calculate.

You can debate that Historical or Market Replay is more "accurate" than the other but it is entirely up to your discretion to determine the best means to test your strategies.