NANEX- FOMC announcement shenanigans by HFT

[Syprik]

AustinP/Clubber Lang:

Sigh. CQG IC (Advantage Futures backbone) and IQFeed 5.1 had only a ~300ms differential shift per live collect on TY & ES tick, and were nearly identical on historical data parse. Both incorrect? Right. I used this data to make above commentary. I've attached a CQG/IQ ES 100-tick bar sequence with corresponding time slices for you to think real hard about. Like I said, after 6 seconds of news, price appreciated only 5-6pts.

If it was unclear, I did not imply one could read release, then manually react. We had set session range levels pre-release, one reacts to price context to tell you what Fed txt depicted. For me, that happened to be a TY scalp on hard break of ES open balance. As stated above, there was ample time to interact with ~80% of the 27 pt move over that next 60-mins of trade. Get a professional platform if you could not react within initial 5-6pts of that move.

Thanks for the reminder why I rarely venture here. The "can't cut it, so lets blame everything but oneself" attitude is rampant. Then act derogatory towards someone where you have zero insight into their capabilities and long trading experience. Interesting how it is often the chatty high post count ET ambassadors who single handily destroyed this forum. Sorry to have interfered in your angry little world.

 

[loopquantumgirl]

Quote from NetTecture:

Little stupid today?

Who tells you their data is stamped 1ms? The fact that the data you BUY from them has a 25ms granularity?

The tape granularity is their decision. Their own (raw) data has a lot more granular timestamps.

You don't seem to understand my point. These colorful plots that they've been making have never been at a higher resolution than 25 milliseconds. The analytics that they've been trying to drive at you are sometimes at the 100 ms resolution:

e.g. http://www.nanex.net/Research/EMini1/Emini1.html

Or sometimes at the 1 second resolution:

e.g.: http://www.nanex.net/StrangeDays/05192011.html

Well, they're free to say whatever they want but to put this in perspective with an example that is easier to understand (magnification factor of the human eye), this is like someone going to the court claiming that he witnessed a murder on the ground from 22,000 meters up in the air. That someone actually believes them, I'd attribute to either genius marketing or a self-selected group of consumers.

I don't feel a need to insult you though - if you are defending them, the chances are that you probably have much worse already and I don't want to add salt to a wound.

 

[NetTecture]

Quote from loopquantumgirl:

You don't seem to understand my point. These colorful plots that they've been making have never been at a higher resolution than 25 milliseconds. The analytics that they've been trying to drive at you are sometimes at the 100 ms resolution:

e.g. http://www.nanex.net/Research/EMini1/Emini1.html

Or sometimes at the 1 second resolution:

e.g.: http://www.nanex.net/StrangeDays/05192011.html

Well, they're free to say whatever they want but to put this in perspective with an example that is easier to understand (magnification factor of the human eye), this is like someone going to the court claiming that he witnessed a murder on the ground from 22,000 meters up in the air. That someone actually believes them, I'd attribute to either genius marketing or a self-selected group of consumers.

I don't feel a need to insult you though - if you are defending them, the chances are that you probably have much worse already and I don't want to add salt to a wound.

No, little not so smart girl.

I just use Nanex myself and I am very aware that they for example work on providing 1ms tapes on request.

Thhey store all data in very high resolution internally. The 25ms is a definition for the generated output of their sytem. They can actually generate more detailed tapes - 1ms is the current technical limit without changing the programming, as the timestamp is in ms resolution - and internally they may swell store gps clock granularity.

You basically are the clueless person standing in a mcdonalds believing cows come in the form of the little bricks of cow meat naturally because that is what McDonalds sells you.

And yes, you need the insult. Looking in a mirror and realizing how utterly stupid it is to assume a company uses the same format internally like in the product they sell may be a wake up call. You do NOT have to demonstrate that level of ignorance publicly.

 

[dtrader98]

Quote from NetTecture:

No, little not so smart girl.

I just use Nanex myself and I am very aware that they for example work on providing 1ms tapes on request.

Thhey store all data in very high resolution internally. The 25ms is a definition for the generated output of their sytem. They can actually generate more detailed tapes - 1ms is the current technical limit without changing the programming, as the timestamp is in ms resolution - and internally they may swell store gps clock granularity.

You basically are the clueless person standing in a mcdonalds believing cows come in the form of the little bricks of cow meat naturally because that is what McDonalds sells you.

And yes, you need the insult. Looking in a mirror and realizing how utterly stupid it is to assume a company uses the same format internally like in the product they sell may be a wake up call. You do NOT have to demonstrate that level of ignorance publicly.

There are 1ms charts as well. It's possible the 1ms plots might sometimes obscure some insight better viewed from a zoomed out view (forest vs. tree).

 

[loopquantumgirl]

Quote from NetTecture:

I just use Nanex myself and I am very aware that they for example work on providing 1ms tapes on request.

Again, you're missing my point. The analytics that they're feeding you is carried out on 100 ms, 1 sec resolution (and edit: sometimes 1 ms, dtrader98). Get over it. I get the feeling that people like to compensate for their frustrations in real life by venting on others on a forum - by implication, you do seem to have many frustrations and I apologize for your experience.

 

[promagma]

I'm just excited there's a girl on the forum, I don't even know what she's saying.

The market reaction to news can be perverse at every level, we imagine it as guaranteed profits, but even if the move happened over 1 hour time instead of a few seconds or minutes, I bet half of all traders would still manage to lose money.

 

[Makis]

Quote from promagma:

I'm just excited there's a girl on the forum, I don't even know what she's saying.
...

I think what she tries to say (based on the ferrari vs tructor example) is that it is ironic for a company that provides 1sec exchange timestamp resolution, to talk about HFT.

Nanex is 10 months past due their promised support of 1ms exchange timestamps (and going...). As NetTecture mentioned, they may have 1ms resolution internally, but still, as this thread proves, it does not look good. Specially to those people that know how much nanex is struggling to deliver exchange ms resolution to their clients.

 

[ogarbitrage]

Nanex and its retarded owner Mr. Hunsader yelling about the ghost in the machine again.

 

[Clubber Lang]

Quote from ogarbitrage:

Nanex and its retarded owner Mr. Hunsader yelling about the ghost in the machine again.

Nanex should be commended for pointing out blatant front running of massive economic news and exposing HFT for the manipulative predatory sham that it is.

 

[nitro]

Quote from Clubber Lang:

Nanex should be commended for pointing out blatant front running of massive economic news and exposing HFT for the manipulative predatory sham that it is.

People are mixing problems. There is nothing inherently wrong with HFT. What is wrong is disseminating the information asymmetrically to market participants. If we all get the data at the same time, then God Bless you if you can somehow take advantage of it then.