Do rollovers or trading hour templ. affect backtest results?

[Ndakotan1313]

My friend and I were both testing a strategy on MCL, and have records of the backtest results from our optimizations. the next week we revisited our strategies to get updated data, and the results are different, including a new larger drawdown between March 2023 and October 2023. We both get the same huge drawdown that we weren't getting before. In the intervening period we both did the MCL 04-25 rollover and updated the trading hour templates. Can either of these updates introduce the 2023 drawdown?

[NinjaTrader_LuisH]

Hello Ndakotan1313,


Thank you for posting on the NinjaTrader forums.


Based on your description, there are two main changes that could have affected your backtest results:


1. MCL 04-25 Rollover

When rolling over futures contracts, NinjaTrader merges data from the expiring contract into the new contract based on the Merge Policy set in the platform.

• Possible Issue: If the Merge Policy changed (e.g., from "Do Not Merge" to "Merge Back Adjusted"), historical price levels may have been adjusted, altering backtest results.
• Check: You can confirm this by setting your Merge Policy to "Do Not Merge" and re-running the backtest.

2. Updated Trading Hours Templates

If the Trading Hours template for MCL changed (e.g., new session times or holidays), historical bars may be built differently, affecting entry/exit logic.

• Possible Issue: If a backtest previously ran on a different session template (e.g., "Default 24x7" vs. "CME Energy RTH"), orders could have executed at different times or been skipped entirely.
• Check: Compare the Trading Hours templates used in both backtests. You can do this in Tools > Trading Hours and ensure they match.

Thank you in advance!