merge non back adjusted issue?

[Murphy67]

Hello,

I want to backtest historical periods, but it takes too much time to define each contracts periods and dates on ''do not merge'' merge policy.
Thats why i changed merge policy to '' merge non back adjusted''

I wanted to test between 12.June.2020 and 10.Sep.2020 period.
First, i loaded ES 06-22 contract, pulled 1000 days 1 minute data and reloaded again.

Then, i run ES 06-22 contract, and select that days on strategy analyzer.
I run another strategy analyzer, this time select 09-20 contract and choosed same days.
But results are different. due to datas are different.

Merge non back adjusted should be raw data, isn't it?
Pls look at the attached file to see differences.

[Murphy67]

If i can get accurate data ( not offset, raw data) for past periods on continuous contract, i will be able to test all past contracts periods at once.
I am planning to enter the past contract dates to text file and strategy will read the dates from there.

[NinjaTrader_RyanS]

Hello,

This is due to the date range being run on rollover days. When loading the 09-20 contract it is only viewing the 09-20 contract, while when loading the 06-22 you are loading the 09-20 until September 10th, when it is then using the 12-20 data. If you run the report from June 12th to September 9th you will see the same with both contracts.

[Murphy67]

Thank you Ryan.
Just for confirmation;

The rolloverdates on Instrument List shows the start date of that specific contract. 12-20 contract shows 10.09.2020 rollover date.

Continous contract 06-22 takes the data from 09-20 contract until end of 09.09.2020 date, because 10.09.2020 is start date of 12-20 contract

[NinjaTrader_RyanS]

Correct, on the roll date the platform will use the later/newer contract