Tick difference between Kinetick and NT Continuum

[gjing]

I use both NT Continuum live connection and Kinetick for my trading now. The reason I still keep Kinetick is that my script performance seems better when using Kinetick than NT Continuum when my trading NQ. I found their tick count is different (9 ticks for NQ), see screenshot. For RTY, only 1 tick diff between them, so the result is pretty much the same.

Is there anything setting for NT Continuum connection so I can get the same (or closer) tick count as Kinetick for NQ? I'm using NT8 now.

[NinjaTrader_DrewO]

gjing,

Thanks for posting.

The Tick Counter would simply count the number of ticks remaining in the bar before a new bar is formed, much like the Bar Timer indicator for time-based charts. Are both charts using the same instrument, expiry, and interval?

[gjing]

Yes, all the settings are exactly same for NQ 03-18.

[NinjaTrader_DrewO]

Hello,

Thank you for posting.

So that I may investigate I would ask that you contact me by email at PlatformSupport [AT] NinjaTrader [DOT] com. In your message please include a link to this forum thread and a copy of today's trace files. You will find these diagnostic files in the Documents > NinjaTrader 8 > Trace folder. I will need to review a copy of the files which are named 'trace.20180109.XXXXX.txt'. There will likely be multiple files for each day.

[gjing]

It's really easy to reproduce on your end. Try to launch NT8 on two different machines, open NQ 03-18 987 tick charts on both of them, one with the connection of Kinetick and the other NT Continuum, you'll see the difference.

[gjing]

Looks like the similar issue between Continuum and Rithmic as well:

https://ninjatrader.com/support/forum/showthread.php?t=87122

[NinjaTrader_RileyS]

Hello gjing, thank you for the post.

I tested this on my end and have similar experiences. Different data providers provide different data, and there isn't a setting for the CQG/NinjaTrader Continuum/Continuum connection to match the rate of incoming ticks.

• All data from Kinetick is true tick by tick data. That is not the case with many providers; in fact, many providers only report some of the transactions that occur. This is referred to as filtered data, but the Kinetick data is completely unfiltered, so you’re getting each and every tick and trade reported directly to you from the exchange, with no filtering involved.
• Comparison: Filtered numerical sequence followed by a filtered numerical sequence. With the filtered numerical sequence you can see that we are missing quite a few data points within that series.

More about this can be found in much more detail in our Kinetick webinar:

• https://ninjatrader.com/Webinar/Kinetick-101

Please let us know if any further questions come up.

[otislauwaert]

So is Ninjatrader Continuum filtered or unfiltered?

[NinjaTrader_PatrickG]

NinjaTrader Continuum is a CQG-based connection, and CQG-based connections are unfiltered.

Regarding the original report in this thread, I am suspicious of the millisecond-level difference in exact start time of real-time data that may be contributing to the observed difference. The help goes into great detail about how a very small difference can change what ticks are associated with what bars:

https://ninjatrader.com/support/helpGuides/nt8/en-us/index.html?how_bars_are_built.htm#UnderstandingWhyAChartCanLookDifferentAfterReloadingHistoricalDataFromTheServer

If we can identify a specific period of time in which the tick data is objectively different, we could look deeper into this, however, I suspect PC clock sync and slightly different initial load times can explain this discrepancy.