Thanks. That confirms the excellent experience that I have with CQG data.
Why is TT data so bad?
- Thread starter canoe
- Start date
Thanks. That confirms the excellent experience that I have with CQG data.
GQG filters their real time data - if you read it refers to both pre and post processing real time data. Besides if you have multiple real time data feeds you can witness this event. Why do you think CQG gained an edge over other vendors all these years? They have clean data - they have clean data because of the pre processing they and post processing algo.
The sales staff at GQG probably does not know anymore about their product than new car sale people know about cars they sell.
If it were not for disclosure agreements with Eric Hunsaker we could get him here to tell you about the whole operation. I can guarantee you TT is faster than CQG given same routing, this is evidence of filtering. That along with the fact their data displays no spikes ever. Splain that.
The sales staff at GQG probably does not know anymore about their product than new car sale people know about cars they sell.
If it were not for disclosure agreements with Eric Hunsaker we could get him here to tell you about the whole operation. I can guarantee you TT is faster than CQG given same routing, this is evidence of filtering. That along with the fact their data displays no spikes ever. Splain that.
NinjaTrader_Ray
ET Sponsor
MarkBrown, I respectfully agree to disagree.
I am not sure what information you base your opinion on. My statement that CQG is unfiltered is based on:
1. Personal experience in continually monitoring and analyzing many data feeds that NinjaTrader integrates with
2. Confirmation and explanation from CQG Executive that their real time data is unfiltered
I am not sure what information you base your opinion on. My statement that CQG is unfiltered is based on:
1. Personal experience in continually monitoring and analyzing many data feeds that NinjaTrader integrates with
2. Confirmation and explanation from CQG Executive that their real time data is unfiltered
This data is continually monitored to filter out bad ticks.
Ray, so how do you real time filter for bad ticks without inducing delay?
I know they use a algo because I have used CQG since the 80's and I had the formula they used and made a user function to filter bad ticks on stocks when Omega Research depended on other data feeds back in the early 90's.
I don't blame you for what you think or the current people working at CQG think cause anyone who really knew anything at that company is long gone.
CQG Data
CQG receives its data directly from the exchanges. Immediately after CQG IC has been enabled, the system begins collecting and displaying real-time data from the CQG network.
This data is continually monitored to filter out bad ticks. When bad ticks are detected, corrections are immediately processed. Likewise, when inserts, deletions, and fast market indications are received, the system immediately sends them to your system.
In the event of an interruption in the operation of the software or the data source, the system automatically corrects the data and fills in the gaps that occurred as a result of the interruption.
Ray, so how do you real time filter for bad ticks without inducing delay?
I know they use a algo because I have used CQG since the 80's and I had the formula they used and made a user function to filter bad ticks on stocks when Omega Research depended on other data feeds back in the early 90's.
I don't blame you for what you think or the current people working at CQG think cause anyone who really knew anything at that company is long gone.
CQG Data
CQG receives its data directly from the exchanges. Immediately after CQG IC has been enabled, the system begins collecting and displaying real-time data from the CQG network.
This data is continually monitored to filter out bad ticks. When bad ticks are detected, corrections are immediately processed. Likewise, when inserts, deletions, and fast market indications are received, the system immediately sends them to your system.
In the event of an interruption in the operation of the software or the data source, the system automatically corrects the data and fills in the gaps that occurred as a result of the interruption.
NinjaTrader_Ray
ET Sponsor
Mark,
What I am attesting to is that data delivered from the server is unfiltered. Internally verified against other unfiltered data feeds (such as but not limited to Kinetick) that NinjaTrader works with and furthermore, confirmed by CQG.
I am not debating that there is tick filtering client side by CQG IC desktop.
I respect your opposing opinion. This will be my last post on the subject.
What I am attesting to is that data delivered from the server is unfiltered. Internally verified against other unfiltered data feeds (such as but not limited to Kinetick) that NinjaTrader works with and furthermore, confirmed by CQG.
I am not debating that there is tick filtering client side by CQG IC desktop.
I respect your opposing opinion. This will be my last post on the subject.
at the tick level the best cqg can do is retard the live feed by one tick. they pre-filter the data before they broadcast it actually then making their real time data delayed.
hypothetically speaking because i would not want to disclosed trade secrets there are footprints you can see filtering.
the way you can know this is when there is a tick far out of the norm they will replace that tick with the price of the previous tick.
mark
i am not saying that your company is not delivering directly what cqg is sending you.
I won't jump into the TT versus CQG fray - I've used them both.
I thought I would mention that quite a few trading desks use Bloomberg Time & Sales in terms of data feeds. There's a reason why MATLAB has a Bloomberg datafeed interface - it's been an industry standard for a couple decades now. Of course, a Bloomberg terminal is an expensive proposition (the contract term is killer).
https://www.mathworks.com/help/datafeed/bloomberg-1.html
I thought I would mention that quite a few trading desks use Bloomberg Time & Sales in terms of data feeds. There's a reason why MATLAB has a Bloomberg datafeed interface - it's been an industry standard for a couple decades now. Of course, a Bloomberg terminal is an expensive proposition (the contract term is killer).
https://www.mathworks.com/help/datafeed/bloomberg-1.html
This is from automated trading perspective, human trading any of them should be ok.
CQG is slowest. they dont even have colocation endpoints in the exchanges.
Rithmic API is faster.
TT is faster
Rithmic Diamond API is fastest of these retail apis.
Vela
Stellar
Rival Systems
are all faster again
half the time your trades are going to take longer than your data feed because you have slow risk checks that clear before your order makes it to the exchange. so your poor man trade bounces around the countryside. bring 1 million and you can skip those checks.
CQG is slowest. they dont even have colocation endpoints in the exchanges.
Rithmic API is faster.
TT is faster
Rithmic Diamond API is fastest of these retail apis.
Vela
Stellar
Rival Systems
are all faster again
half the time your trades are going to take longer than your data feed because you have slow risk checks that clear before your order makes it to the exchange. so your poor man trade bounces around the countryside. bring 1 million and you can skip those checks.
Last edited: