Non-display fees

[kmiklas]

Well, that's sort of good news, since at least IB will probably be left untouched. Their feed is not real time, but might be enough for many algos. I'm not really planning to do HFT nor anything similar.

Is there anyone here day trading succesfully using only IB Market Data?

(Raises Hand).

I was going to buy a data feed, and looked at a few vendors, but was stopped cold by these non-display fees. Let's not forget that there's more fees from the vendor for the data, including setup and monthly fees. It's just too expensive.

IB's data feed is usable, but I had to slow down. With a slightly delayed VWAP feed, It is impossible to compete against someone using real-time tick data. I was pushed into day/swing trading, limited by the quality of the data. Also note that you will not be top of book; others will get orders matched before you. You get what you pay for.

 

[Shreddog]

What makes you so sure IB's consolidated data is OK? Couldn't NYSE argue the data is "derived"?

 

[Joan]

(Raises Hand).

I was going to buy a data feed, and looked at a few vendors, but was stopped cold by these non-display fees. Let's not forget that there's more fees from the vendor for the data, including setup and monthly fees. It's just too expensive.

IB's data feed is usable, but I had to slow down. With a slightly delayed VWAP feed, It is impossible to compete against someone using real-time tick data. I was pushed into day/swing trading, limited by the quality of the data. Also note that you will not be top of book; others will get orders matched before you. You get what you pay for.

Do you trade manually or with algorithms through the API?

 

[Joan]

Joan:

https://www.ctaplan.com/publicdocs/ctaplan/notifications/trader-update/CTA Network A Pricing - Jan 1 2015.pdf

Bottom of page 3:

Non-Display Use refers to accessing, processing or consuming data, whether delivered via direct and/or redistributor data feeds, for a purpose other than in support of the datafeed recipient’s display or further internal or external redistribution. It does not apply to the creation and use of derived data.

IANAL. But the above sentence says to me that if you "derive data" then use it to trade, that is "use of derived data", and the "non-display" thing does not apply to it.

Thanks terr. I don't really get that last sentence. Algorithms do read Market Data, derive another data, and create orders based on that derived data. It looks to me that sentence wanted to say that you can calculate your custom indicators and use them to help you trade, the same way you watch a moving average or something similar.

 

[kmiklas]

What makes you so sure IB's consolidated data is OK? Couldn't NYSE argue the data is "derived"?

I never said that I was sure!!! I'm not a lawyer, and don't work for IB. I'm just relaying what I've been told, and giving my opinion on the matter.

Do you trade manually or with algorithms through the API?

I use TWS and the API; specifically, the C++ version.
http://interactivebrokers.github.io/tws-api/#gsc.tab=0

 

[terr]

Thanks terr. I don't really get that last sentence. Algorithms do read Market Data, derive another data, and create orders based on that derived data. It looks to me that sentence wanted to say that you can calculate your custom indicators and use them to help you trade, the same way you watch a moving average or something similar.

That's my point exactly. So, based on that, you can file your "non-display" declaration with NYSE, truthfully saying that your use is not "non-display".

 

[kmiklas]

I was so aggravated by all this that I actually started building a new exchange, a "lit" pool, where data in all its forms--market, deep book (Level 2,3,4,...n), and queue data alike--are all free. A level playing field, where we can openly develop and run algos, without all this non-display crap.
http://www.decentralizedstockexchange.com/

It's obvious that these fees are discriminatory.

 

[Joan]

I was so aggravated by all this that I actually started building a new exchange, a "lit" pool, where data in all its forms--market, deep book (Level 2,3,4,...n), and queue data alike--are all free. A level playing field, where we can openly develop and run algos, without all this non-display crap.
http://www.decentralizedstockexchange.com/

It's obvious that these fees are discriminatory.

Could you not just graph the ouput of your algos on your screen and just hit the BUY button manually yourself? It's a bit crappy, but it's the same regular traders do, isn't it? You derive data that is in support of your display, but it's ultimately you who decides when to send the order. Even though the visual output would tell you exactly what to order, prices, limits, etc... That would be really really hard to distinguish for NYSE, I guess.

 

[lspdtrdr]

That's my point exactly. So, based on that, you can file your "non-display" declaration with NYSE, truthfully saying that your use is not "non-display".

If your data is coming from a broker like Lightspeed with an integrated data/order API, then you can expect an audit if you declare display and are pulling data via API. If you don't mind going through an audit process and have nothing to hide, then it should be fine.

Before you say, not gonna happen, I have a friend who was audited by NYSE recently for pulling data from a vendor with NO broker connection. They're getting WAY more strict about this and at 100k+ yearly fees minimum for SIP data, they can afford to audit everyone that's remotely suspect.

 

[kmiklas]

Could you not just graph the ouput of your algos on your screen and just hit the BUY button manually yourself? It's a bit crappy, but it's the same regular traders do, isn't it? You derive data that is in support of your display, but it's ultimately you who decides when to send the order. Even though the visual output would tell you exactly what to order, prices, limits, etc... That would be really really hard to distinguish for NYSE, I guess.

Nope. You're playing games... and you won't wiggle out of it that way.