profit discrepancy in strategy analyzer and market replay

[Playdc]

I use the same strategy to backtest both in strategy analyzer and market replay. The data is 100 tick for one day. The total profit in strategy analyzer is $62k, while market replay shows $36k. Which one should I trust?

[NinjaTrader_ChelseaB]

Hello Playdc,

May I confirm you have properly implemented 1-tick intra-bar granularity (and TickReplay if not using Calculate.OnBarClose) for accurate results?

Below is a link to a support article on comparing strategy results.

https://support.ninjatrader.com/s/article/Developer-Guide-Comparing-Strategy-Results

In general Playback with Market Replay data will be more accurate than a backtest, however with 1-tick intra-bar granularity and TickReplay the differences would be minimal and either may be closer to live performance.

[Playdc]

Hi Chelsea,

I still have two questions:

1. where could I setup 1-tick intra-bar granularity?

2. Since if the condition meet, I would buy or sell at the open of next bar, I am a little confused why 1-tick intra-bar granularity could help to emulate the real order filled price, is the open price of a bar not the real open price?

[NinjaTrader_ChelseaB]

Hello Playdc,

The Improving backtest order fill accuracy with intra-bar granularity guide linked in the Comparing Strategy Results guide linked from post # 2 provides direction for implementing 1-tick intra-bar granularity.

If the Calculate setting is OnBarClose, you could also choose to set Order Fill Resolution to High 1 tick.


It would depend on the order types being used.

Market orders would fill at the open of the next bar in historical and backtest if submitted after bar close.

Stop and limit orders that fill intra-bar in real-time would fill at the bar close in historical which would a different time and likely a different price.