Live vs Historical tick chart discrepancies

[imjohn]

I use Ninja Trader Continuum and trade using tick charts.

I screenshot my trading at the end of each session before I log out for the day.

Today I was comparing screenshots (of my live trades) to historical charts and noticed small discrepancies.

I then did some research and found this:

https://ninjatrader.com/support/helpGuides/nt8/en-us/?discrepancies_real-time_vs_bac.htm

Differences in chart data

•	If you run a strategy in real-time on DAY1 and then DAY2, you are now backtesting your strategy on DAY1 data instead of processing like it did in real-time so there could be differences. You should understand how chart bars are built.

•	If using tick based charts, all it takes is a single tick difference between real-time and historical data to generate completely different looking charts. This in turn would impact the calculations of your strategy should the data sets be different.

I do use tick charts and make decisions based on close of bar. Basically, what I realized is, if I were trading what I now see "historically", the outcome would definitely be different than what I achieve trading the live feed coming in (Im not saying better or worse, just different)..

My question

1) Are these discrepancies more prevalent using NT continuum?
2) Will I have the same issues if I switch from NT continuum to say, kinetick or rithmic feed?
3) Which is the most accurate feed (for comparing real time to historical)? Kinetick?


I guess what I want is to be able to pull up historical tick charts, and have them match the charts which I based my real-time trading decisions on.... Is that possible with Continuum?... Kinetick?? etc.

[NinjaTrader_EricB]

Hello,

Thank you for your post. For 1, no this would not be more prevalent with NinjaTrader Continuum. For 2, the same would be seen with all tick data. For 3, the same loading behavior would occur with all providers so if the chart is opened or refreshed after the session has already started, these charts will be different than the historical data that is downloaded which will include the data from the beginning of the session without a new bar forming where the chart was opened regardless of provider.

In other words, if you open a chart before the start of the session and never reload this, this would match the historical data. See the below section of our help guide for more on this matter, "Understanding why a chart can look different after reloading historical data from the server."

https://ninjatrader.com/support/help...aFromTheServer

Let us know if we may assist further.

[imjohn]

I see. Thanks for the response.



"In other words, if you open a chart before the start of the session and never reload this, this would match the historical data."

--
How about if I open a chart after the start of the session (let's say at 09015 ET), and reload historical data right when I open. Will the incoming real-time data for the remainder of the session match what is printed for historical?

[NinjaTrader_EricB]

Thank you for your reply.

No, bars would load from the historical data servers on reload and then tick bars would begin building as received. From our help guide: "The only way to ensure that data always looks the same is if every connectivity provider sent ticks with time stamps AND that all vendors synchronized on time stamps. Unfortunately, this is just not a reality nor plausible scenario."

Please let us know if we may be of further assistance.

[afcmaff]

imjohn,

I had the same issue with Kinetick -- I use 'on bar close' logic with tick charts and even 1-tick Close[0] difference changes the equity curve on back-tests vs live. There's no good way of ensuring your live data will precisely match the server historical data as NinjaTrader_EricB stated, but you can reload historical data in between sessions (the 1-hour break period) to see how often there are discrepancies and try to program some error-buffer logic, or just accept some slippage and/or intraday trade discrepancies that hopefully offset a little so as not to be too far off from the back test when running live -- this should hopefully boost confidence in a tick-bar-close-based algo, even slightly, if you know the results are not disastrously different.

Happy Trading!