[maxima]

In other thread I tried to ask Urma on what his experiense with data feeds above retail quality is... However the answer was Bite me!... I chose not to bite but rather ask around. I know Urma is not the only trader who saw real trading desks. Also I have a suspicion that he has no idea what am I asking about.... Anyway.

What I am trying to find out is - how CME distributes the market data - trades and quotes.

Is there one single level of quantity/quality of data distribution or there might be several tiers?

Do companies like CQG, TT and DTN for instance have the same input data on their servers as say Lehman or Merrill?

[james_gsx]

Originally Posted by maxima »

In other thread I tried to ask Urma on what his experiense with data feeds above retail quality is... However the answer was Bite me!... I chose not to bite but rather ask around. I know Urma is not the only trader who saw real trading desks. Also I have a suspicion that he has no idea what am I asking about.... Anyway.

What I am trying to find out is - how CME distributes the market data - trades and quotes.

Is there one single level of quantity/quality of data distribution or there might be several tiers?

Do companies like CQG, TT and DTN for instance have the same input data on their servers as say Lehman or Merrill?

What a great way to start a post, by dragging non-sense from another thread. This isn't ET, maybe you got confused

As to your question, I think it depends on what you want to do. CQG is definitely one of the best, and an old member (not sure if he still posts, I forgot his name) used CQG at his firm and loved it. As far as quality goes, I think you already know the answer to that, you won't get the same level feed from Zen-Fire as you would TS or E-Signal.

As far as data input goes, I'm not really sure what you mean by that. As far as big shops that do a lot of HFT and provide liquidity - they will obviously get their data directly from the CME since it all comes down to speed at that point. I don't think the CME would limit that information.

You can check out this link and it might be of some help if by "input" you mean something completely different.

Information for Distributors

And then theres this...

CME E-quotes


Hope it helps.

[Spydertrader]

Nanex.net

http://www.dtn.com/files/Trading/SS_NXCORE_1107.pdf

[brownsfan019]

Originally Posted by james_gsx »

What a great way to start a post, by dragging non-sense from another thread. This isn't ET, maybe you got confused

As to your question, I think it depends on what you want to do. CQG is definitely one of the best, and an old member (not sure if he still posts, I forgot his name) used CQG at his firm and loved it. As far as quality goes, I think you already know the answer to that, you won't get the same level feed from Zen-Fire as you would TS or E-Signal.

As far as data input goes, I'm not really sure what you mean by that. As far as big shops that do a lot of HFT and provide liquidity - they will obviously get their data directly from the CME since it all comes down to speed at that point. I don't think the CME would limit that information.

You can check out this link and it might be of some help if by "input" you mean something completely different.

Information for Distributors

And then theres this...

CME E-quotes


Hope it helps.

How can you possibly know that James?

As a matter of fact, just last weekend or the weekend before Zenfire was showing trades on their data on a Saturday on the CL (when the market is closed)...

Please show some evidence of Zenfire's feed being superior to others; otherwise you are just giving an opinion w/ no basis of fact.

[BlowFish]

In my experience. Zenfire is far far far far superior to tradestation. Far superior. Esignal I have not used for a while (probably a couple of years) last time I did it could seriously lag in fast markets. Guess it islikely improved now.. Zen does use UDP but that is not really an issue if your infrastructure is good and poor infrastructure will effect any feed. Zen and ninja I am not so sure about. Sadly seems like the ball has been dropped with NT7.0 so I guess I wont be looking at it again until 7.5 whenever that is.

I guess it all depends what you mean by "high end"The two big things are timeliness and completeness. The method I am currently using (and have used in the past) are fairly robust (robust in respect to the quality of the data that they require). Having said that I am thinking about rolling in a couple of extras (order flow and inventory type stuff) for this I am leaning towards Zen or TT for live data with DTNIQ for history.

Edit brown I think James was saying Zen is inferior. I did not keep my test results I just monitored in real time with code that recorded differences in feeds. Zen was clearly the most timely and complete. Multicharts is quite a good platform to compare feeds.

Edit2 if you use different platforms for different feeds you obviously have another variable. For example if using ninja/zen and .net does garbage collection that will cause a hiccup that is clearly not due to the feed.

[brownsfan019]

Inferior or superior, doesn't matter - just curious how anyone would know this. If the issue is based on lag, that's one thing but not necessarily what defines a feed better than others as it could purely be on your end.

I'm just curious how one defines a feed better than others and substantiate it.

[maxima]

sorry guys... I didnt ask who is superior or not. And I am certainly not interested in Zen nor IQFeed nor eSignal less of all in TS.

Let me put it simple - will Thompson or eSpeed have MORE data than DTN NxCore or CQG...

Can one have MORE granular ticks or SOONER than others from CME ....

The other version of the question - do some shops have advantage over others?

[BlowFish]

Originally Posted by brownsfan019 »

Inferior or superior, doesn't matter - just curious how anyone would know this. If the issue is based on lag, that's one thing but not necessarily what defines a feed better than others as it could purely be on your end.

I'm just curious how one defines a feed better than others and substantiate it.

As I said before the criteria I used was timeliness and completeness. I also gave an idea of how I tested those (some rudimentary code to measure delta time). My end is pretty good though only 20 meg (cant justify 50 meg beyond bragging rights) having said that the lag across it will be the same for the data feeds I am comparing.