IB : the end of excessive messaging?
- Thread starter TraDaToR
- Start date
maybe one reason nobody is jumping in is that people like you do not precisely represent the most desirable clientele for most brokers. Whatever promises execution parties make, someone who changes orders xxx-times more often than the number of executed trades is not someone any broker wants to work with. Especially knowing that most exchanges dont tolerate such behavior either, unless serious money backs such operations and make it lucrative for the exchanges, I doubt you are one of such parties. But I thought that was rather common sense. Mark my words, in 2-3 years nobody, not the last worm on earth will talk about HFT anymore. If you trade sub-minute (but not HF) and need to re-submit orders into the market to the tune of 100+times before you execute a trade then there is something seriously wrong with your automated strategy.
Quote from shortie:
thank you for asking an excellent question, shortie. i will bump it up for you.
IB only supports exchange-supported calendar spreads, but not butterflies, strips, crush, crack, brent/crude...I don't know why. I suppose it would be hard to integrate to TWS and few IB clients would be interested.
Calendar spreads can be automated though and orders are matching FIFO, unlike their futures option spreads which are not filled 'til yours orders are marketable for MMs...
With every BD I have ever been with, this is typical. I have researched this extensively, meeting with many firms in Chicago. Same answer, there is a ratio of messages to execution they want. The reasons are that they have to run very expensive pipes to the exchange to handle the enormous amount of data. In addition, the number of messages explodes at the exchanges, causing them to have to run a continuous battle of being able to handle more and more messaging. I have even been with firms that ran private lines to an exchange to give it's traders a technological edge or at least to play on a level playing field, and it was horrid. Even money is not enough, you have to know what you are doing too.
It always angered me also when I was told the same thing no matter where I went. The only solution is to run your own private lines into an exchange or colocate at one of these places that aggregates this business, a very expensive option not just because of the pipe and fees, but because it usually also entails custom software. Expenses start to pile up. And then, the BD or clearing firm has to allow you to have direct (naked) access, something that they won't do unless you have a big account.
Trading, and in particular trading on the technological edge, is meant for those that are very well capitalized. In addition, many of these BDs are in the same business you are attempting to take advantage of. They don't want you doing it on their pipes. For 99% of the people on ET, do yourself a favor and stop trying this sort of trading. You are banging your head against a brick wall.
It always angered me also when I was told the same thing no matter where I went. The only solution is to run your own private lines into an exchange or colocate at one of these places that aggregates this business, a very expensive option not just because of the pipe and fees, but because it usually also entails custom software. Expenses start to pile up. And then, the BD or clearing firm has to allow you to have direct (naked) access, something that they won't do unless you have a big account.
Trading, and in particular trading on the technological edge, is meant for those that are very well capitalized. In addition, many of these BDs are in the same business you are attempting to take advantage of. They don't want you doing it on their pipes. For 99% of the people on ET, do yourself a favor and stop trying this sort of trading. You are banging your head against a brick wall.
Cheaper and cheaper is correct, but it isn't infinite bandwidth you know.Quote from Cdntrader:
from an outsiders perspective I'm somewhat surpised this is even an issue in this era of cheaper and cheaper technology/bandwidth.
What am I missing? or is it that it just isn't cheap enough yet?
Let me give you a somewhat concocted example. Say you had a business that leased F1 engines, you know, the ones that they use to race at the Indy 500. Clearly this is a finite resource. Now, say you had a driver that came to you and said, I am a great driver, I want to make use of your engine. You explain that you charge per laps around the track in races. You say great. You give this race car driver the engine, only to find out that he spends most of his time on the dyno, tweaking his design against this engine, testing breaks, shift patterns etc. He finally races a couple of times a year. In the meantime, you aren't getting payed. This guy is dicking around pretending to be a racer. You lose.
Imagine now that racing took off and now they had every Tom Dick and Harry coming to you with the same proposition. You see where this leads? If the engines were that cheap, they wouldn't need your service of leasing F1 engines, just go off and buy twenty F1 engines and do your thing.
It is important to split the issue in two though. There is the dissemination of market data, and then there is the sending of orders to the matching engines. They are closely related, but only the very well capitalized has raw direct data from exchanges and the leased lines to get it, and send orders to the matching engine directly. The question you may not realize you should be asking is, is data etc that profitable for exchanges? The answer is yes, and that is where some of the problem starts (I am not even talking about exchange fees for matching a marketable order - that is another revenue source for them separate from market data). Putting in a circuit costs money. Maintaining it costs money. Putting in servers, routers switches firewalls that can handle the traffic costs money. Staff to monitor it all costs money. The matching engine costs money. Programmers cost money. Storing the data for regulatory requirements costs money. Redundant everything costs money. The list is almost endless.
I firmly believe that the data does not belong to the exchanges. The data itself should be free, and I believe it will eventually be free when 10GB fiber to every house is as common as tap water is into a house. It will probably be an option on your cable bill to get market data multicast to you. What a dream. It will be ten years plus though for this to happen. What belongs to the exchanges is the matching engines and the right to charge to send orders into them. But not the data itself, imo, since there is no intellectual property in it.
Well, as usual the bottom line is in fact the bottom line. Exchanges, prop firms and clearing firms get paid for executed orders. Even a HF algorithm strategy that derives income from exchange rebates by providing liquidity is in fact executing orders.
So, you are dealing with a supply chain that survives and thrives on executed orders - as the end user (trader) you are going to have to deal with that fact no matter where you go, who you clear, or who backs you on a proprietary basis. Inescapeable fact of life, so in the end it's best just to deal with it and move on. There is no getting around "the man" in this case.
Sure, you can move from Globex to ELX, and maybe they'll let you message ad infinitum (for the time being), and you get some trades, but I promise you that the bid/ask spread width and volume won't justify the change - the very first pioneers are the ones with the arrows in their backs lying face down in the mud. You want to be one of the later pioneers in terms of risk/reward, and by then (big "if", nobody outside of the founding partners appears to want to clear them) ELX will certainly have a message-to-fill ratio policy.
So, you are dealing with a supply chain that survives and thrives on executed orders - as the end user (trader) you are going to have to deal with that fact no matter where you go, who you clear, or who backs you on a proprietary basis. Inescapeable fact of life, so in the end it's best just to deal with it and move on. There is no getting around "the man" in this case.
Sure, you can move from Globex to ELX, and maybe they'll let you message ad infinitum (for the time being), and you get some trades, but I promise you that the bid/ask spread width and volume won't justify the change - the very first pioneers are the ones with the arrows in their backs lying face down in the mud. You want to be one of the later pioneers in terms of risk/reward, and by then (big "if", nobody outside of the founding partners appears to want to clear them) ELX will certainly have a message-to-fill ratio policy.