Question about backtesting using Heiken Ashi bar

[dahutu]

Hello, I backtested a strategy based on common minute bar and then tried on Heiken Ashi bar just for a check without changing anything on my strategy. The results have considerable differences. My question is what is the backtest mechanism for a strategy based on minute bar putting on Heiken Ashi bar. The calculation of the logic based on OHLC would be Heiken Ashi bar's OHLC or minute bar's. what bar prices are the execution prices of signals based on? Thanks.

[dahutu]

Hello, I backtested a strategy based on common minute bar and then tried on Heiken Ashi bar just for a check without changing anything on my strategy. The results have considerable differences. My question is what is the backtest mechanism for a strategy based on minute bar putting on Heiken Ashi bar. The calculation of the logic based on OHLC would be Heiken Ashi bar's OHLC or minute bar's. what bar prices are the execution prices of signals based on? Thanks.

It also seems backtest results are different in strategy analyzer and chart for the same period using Heiken Ashi bar.

[NinjaTrader_Jesse]

Hello dahutu,

The Heiken ashi bar and minute bar would have differences in values overall due to how the heiken ashi is made to form the data. In the backtest the bars would be different between your two tests so we should expect different results.

Although the heiken ashi has a base series value that matches your primary it would only be used to generate the heiken ashi bars, the backtest will then use the generated bars to place executions. As the data is different it is very likely you will see differences overall in how the strategy logic works and where executions are placed. You can use the execution information or chart to visually see the differences to better understand why.


I look forward to being of further assistance.

[mic414]

This is another case where NT8 is a step back compared to NT7. NT7 handled the price correctly, i.e. used the base series, not the Heiken-Ashi prices. Heiken-Ashi in NT7 was a real indicator, whereas in NT8 it is a data series type.

Backtesting in NT8 is plagued with basic problems, things that worked perfectly in NT7 (e.g. position sizing is missing, backtesting long data series in NT8 crashes NT). I don't understand why these couldn't be solved in NT8, when it was already working in NT7. :-(

As a result, I'm still stuck with NT7, and won't upgrade until someone finally decides to address these issues. The main reason I loved and invested in NT7 was the powerful and easy to use backtesting functionality. This is absolutely not true in NT8 anymore.

[NinjaTrader_Jesse]

Hello mic414,

In NT7 the heiken ashi was generally an indicator, you can still do that in NT8 as well there are a few versions on the user app share. A BarsType was added in NT8 for those who want to use that as the base series directly for placing trades.

If you are having troubles with a part of the platform or want to see a new feature we suggest to make a post about that specific item and provide an example of the situation if relevant. We can use that information to file a bug report or create a feature request.


I look forward to being of further assistance.