Kinetick vs NT's historical market replay data

[nightriderx]

Hi,
I an having quite a difference in the results of testing strategies on market replay vs real time data feed from kinetick. Can you tell me what can cause such anomalies?
My kinetick data is not being filtered and there have been no disconnects in my test sample.
I know slippage is a factor but how do you recommend dealing with this?
thx

[NinjaTrader_Matthew]

Hello,

Thank you for your post.

The subject of your thread refers to Kinetick vs NT's historical replay data - keep in mind that this not the same exact data as they come from separate providers. The market replay data is recorded with a live Zen-Fire connection, while your Kinetick feed is using their servers. In this case, it would be expected to see different results.

If you would like to use the same data you are receiving from Kinetick, you can record this in real-time by going to Tools--> Options--> Data tab and ensure you have checked "Record for market replay"

In addition, the Market replay uses a algorithm to file you in a simulated fashion. This algorithm attempts to simulate a real exchange but it is not the live market so the live market will always act differently in terms of filling orders.

Furthermore, it is expected to see different results in real-time versus market replay. You may be familiar with the expected discrepancies you will see in real-time versus historical backtesting, but many of these variables will also apply to market replay. Please review our Help Guide article on this topic and let me know if you have any questions:

http://www.ninjatrader.com/support/helpGuides/nt7/index.html?discrepancies_real_time_vs_bac.htm

[nightriderx]

ok can you please explain how the "filter bad ticks ... off market" works in the options/data section?

the default is 0.1% - what is that 0.1% of?

is that for real + simulated trading?

[NinjaTrader_Matthew]

nightriderx,

The filter bad ticks... option will exclude any ticks that are greater than the percentage offset configured. The percentage offset is based off the value of the last tick you received.

This is applied to any real-time data you are receiving from a market vendor.

Please let me know if you have additional questions.

[vlc]

I had a long position in Z futures, the trade was in good profit and so my trailing stop-loss was in place well above my entry price. The market was going up, but at about 7:50am the price as reported by Kinetick dropped dramatically, incorrectly triggering my stop-loss... a Sell To Close order to sell at market price was sent to Collective2 and that order was executed at about 8:00am when the market opened. That closed my trade, even though the real market price never hit my stop ( C2 executed the trade at a price well above my stop-loss position ) and in fact the market continued up in what would have been my favour. The trade ended with a good profit, but should not have been closed. Since this behaviour is a daily occurrence ( at the same time every morning), is there no way to avoid it? Do I really need to remove my stops for any Z overnight trades? I've been told that this is due to low volume where there is no bid or ask price and it gets transmitted as 0, but (although this is not quite unique to Z, and occasionally happens with YI and a few others ) there are loads of other instruments which have regular periods of low volume and it doesn't happen with them.
I tried setting the Filter bad ticks option to 0.2% but that does not help. I suppose if two bad ticks are received consecutively they're not seen as bad ticks. This problem with Z futures is daily at about 7:50am - London time ( ten minutes before the market opens locally ); very easy to repeat and therefore should be possible to resolve.

[nightriderx]

Can you add a volume minimum, so that it has to hit a certain volume on a tic/second or whatever time frame in addition to your price parameters for the stop to go off?