Discrepancy between Kinetick Range data and real market data

[nikolaalx]

Hi,
I have a strategy that is based on crossover of Exponential moving average.

It struck my attention that the optimization results, significantly differ from the real market data.

Please take a look at the screenshot of the backtest for today's data when compared with the actual data that has been recorded from the real market movement.

Kinetick state that they do not filter their data in any way, however as you can see the difference is so significant that 80% of the crossovers do not take place.

[NinjaTrader_PatrickH]

Hello nikolaalx,

Thank you for your post.

Are you connecting to any other data feed providers?
Which screenshot is the backtest and which is the real-time?
Under Tools > Options > Data > Real-time data > is 'Save chart data as historical' enabled?

[nikolaalx]

I am not connecting to any other data feed providers. I backtest via Kinetick...and in real trading I connect only to IB.

Saving real time as historical is enabled, yes.

In the attached/uploaded file name(at the bottom), you will see which one is for the backtest and which on is in real mode.

I have connected to Kinetick and have reloaded all historical data and the discrepancy is now gone.

I however do not consider this a solution, as on one end I have had an isolated IB data which was not matching isolated Kinetick data for the same period.

What guarantees do I have that over the next e.g. 2-3 days, data from IB will not be different from the historical one from Kinetick (based on which I have done my optimization).

[NinjaTrader_PatrickH]

Hello nikolaalx,

Thank you for your response.

Interactive Brokers data is aggregated data, there fore it is possible to see differences when compared to non-aggregated data.