Big difference in strategy performance using UniRenko bars

[NicoNiso]

Hi Support

I really could use some help from your guys.
I have created a strategy based on the unirenko bars and the Fisher Transform indicator. I have led the strategy run the whole night, and when I then compare the strategy performance on the chart that has been running the whole night, and with a new duplicated chart with the same setting, I get a very different picture of the performance. How can there be such a big difference?
I have attached to PDF documents where you can see the big difference.
The PDF with name: NinjaTrader Grid 2020-05-18 08-01 – ES
Is the strategy that has been running the hole night.
The PDF with name: NinjaTrader Grid 2020-05-18 08-00 - ES unirenko 2 – Duplicate
Is the duplicated chat with an export of the performance.
I really look forward to hear from you.
Thanks in advance!
Kind regards
NicoNiso

[NinjaTrader_ChelseaB]

Hello NicoNiso,

Please review this forum post that discusses differences between real-time, historical, and playback.
https://ninjatrader.com/support/foru...nce#post100192

As a heads up, TickReplay does not work with bar types that use IsRemoveLastBarSupported such as Renko, UniRenko, LineBreak, etc.

"Note: The system bar types "Line Break" and "Renko" cannot be used with Tick Replay and as a result, the Tick Replay option will be disabled when configured with those bar types. There may be other 3rd party bar types which may also disable Tick Replay by design. If you are a developer, please see the property IsRemoveLastBarSupported for more information."
https://ninjatrader.com/support/help...ick_replay.htm

If you want understand what is causing differences, it will be necessary to do some work and print (or write to file) the time of the bar, the real-time data, the historical data, all values used in conditions, TraceOrders, and the order object in OnOrderUpdate().

I am happy to assist with analyzing the output saved from the output window to a text file if you would like to go through the process of using prints to understand the behavior.

[NicoNiso]

Hi Chelsea
Thanks a lot for your quick reply, and it would be great if you would assist me with analyzing the file
I will try to include the print function to my strategy, to better understand the big difference.
All the necessary information you have listed, will that be printed if i just include the print function through the strategy builder, og do i have to do a bigger "surgery" on the strategy?

Kind regards
NicoNiso

[NinjaTrader_ChelseaB]

Hello NicoNiso,

The time and information of the bar and all values used for the conditions can be printed from the Strategy Builder as well as enabling TraceOrders.

I'm directly linking the forum post on using prints to understand behavior (that is linked from the first post on comparing results) below.
This includes a video on using prints from the Strategy Builder.

https://ninjatrader.com/support/forum/forum/ninjatrader-7/strategy-development-aa/99209-date-conditions-not-working?p=791121#post791121

[koganam]

Hi Support

I really could use some help from your guys.
I have created a strategy based on the unirenko bars and the Fisher Transform indicator. I have led the strategy run the whole night, and when I then compare the strategy performance on the chart that has been running the whole night, and with a new duplicated chart with the same setting, I get a very different picture of the performance. How can there be such a big difference?
I have attached to PDF documents where you can see the big difference.
The PDF with name: NinjaTrader Grid 2020-05-18 08-01 – ES
Is the strategy that has been running the hole night.
The PDF with name: NinjaTrader Grid 2020-05-18 08-00 - ES unirenko 2 – Duplicate
Is the duplicated chat with an export of the performance.
I really look forward to hear from you.
Thanks in advance!
Kind regards
NicoNiso

The "Renko Deception" strikes again.

ref: https://ninjatrader.com/support/foru...639#post644639

[tgn55]

You cannot translate back-test with UniRenko bars to live trade results. It just does not work. Why? In back-testing, an order is presumed filled at the open of the next bar. Which works fine for normal time-based bars. Use UniRenko... and you will get kicked in the, uhh... well, you can guess.

With Renko bars, the opening price is at best a calculated value, it's NOT real, and not reflected in the REAL live price action. You will find you frequently get "slippage" of the order of twice the size of the UniRenko bar, so if you have a 10-tick bar body... hmmm... pretty ugly. And when the bars reverse direction...look out for major badness.

And that's why strategy analyser results can be so VERY much better than actual trades.

Hope it helps...
T.

[NicoNiso]

Thanks a lot guys - Really appreciate it! Now i can move on