I think in the next 5 years we'll be down to a handful of true Ultra low latency, high frequency traders and they'll probably be the ones you know, TradeBot, Getco, Citadel, etc.
I think I'm very fortunate not to have to compete in that space, it allows us to concentrate more on the algorithms than the execution.
It's very hard to get started in it anymore. I work for a decent sized firm and we can get laughed out of meetings when we ask about some of the more desirable fiber routes. I can't imagine what it would take to get any microwave spectrum space:)
The incumbents have the spectrum and fibre locked down and the bigger players like Citadel pay the retail brokers for their retail flow so they can trade against the "dumb" money before anyone else. I've heard that most Canadian banks sell their flow to a big US fund that has a chance to trade against it before it reaches the market.
If the CN tower was available for renting out microwave signals I'm sure someone would have rented out the entire spectrum and held it for themselves. Its in a very good location for Inter-listed arbitrage between Toronto and Chicago and Toronto and New York:)
This is a good article about the cutthroat business of being in the HFT space and the one I think the post was talking about.
> "There are rumors in the industry that people have bought tower space or have bought frequencies that they don’t use,” Cumberland said. “We call it ‘frequency squatting’ or ‘tower squatting."
> "Traders need multiple towers because signals run the risk of breaking up beyond 100 kilometers (62 miles). Controlling a single tower wouldn’t benefit a trader unless the goal was to force rivals to go around it by using a slower, less direct route, or hoarding the frequencies that companies rely on to transmit financial data, said Colt’s Cumberland."
Meta note, if anyone wants to talk trading algos or technology feel free to reach out, email in my profile.
Why is that distateful? My impression is selling flow is the only way for retail brokers to keep competitive prices while complying with SEC rules. Instead of executing orders themselves according to the rules, they sell the order then price it to the customer as if it had been executed by the rules. I've never worked in equities so correct me if I'm wrong.
My concern is that the investor doesn't understand what is happening. Someone goes out to hit a bid, but actually a prop group has paid to see their intention before it goes to the market, and can choose to participate or not.
But I'm a futures guy, so my understanding of payment for flow is minimal.
You are being overly pessimistic about access to these microwave paths. In some cases in the U.S. you can purchase wireless market data directly from the exchanges†. I am less familiar with the Canadian markets so I can't speak to what is available there.
Agreed -- however, it's worthwhile to note that in my experience, you get what you pay for in terms of wireless connectivity. There are superior products to the Nasdaq offering.
How profitable are the large HFT firms? My understanding is that as more competitors entered the space profitability has gone down significantly. I remember hearing one HFT guy explain how in the early 2000s, before HFT became big the bid-ask spread could very large but it is now virtually non-existent due to so much HFT activity.
http://www.businessweek.com/articles/2013-06-06/how-the-robo...
I think in the next 5 years we'll be down to a handful of true Ultra low latency, high frequency traders and they'll probably be the ones you know, TradeBot, Getco, Citadel, etc.
I think I'm very fortunate not to have to compete in that space, it allows us to concentrate more on the algorithms than the execution.
It's very hard to get started in it anymore. I work for a decent sized firm and we can get laughed out of meetings when we ask about some of the more desirable fiber routes. I can't imagine what it would take to get any microwave spectrum space:)
The incumbents have the spectrum and fibre locked down and the bigger players like Citadel pay the retail brokers for their retail flow so they can trade against the "dumb" money before anyone else. I've heard that most Canadian banks sell their flow to a big US fund that has a chance to trade against it before it reaches the market.
If the CN tower was available for renting out microwave signals I'm sure someone would have rented out the entire spectrum and held it for themselves. Its in a very good location for Inter-listed arbitrage between Toronto and Chicago and Toronto and New York:)
This is a good article about the cutthroat business of being in the HFT space and the one I think the post was talking about.
> "There are rumors in the industry that people have bought tower space or have bought frequencies that they don’t use,” Cumberland said. “We call it ‘frequency squatting’ or ‘tower squatting."
> "Traders need multiple towers because signals run the risk of breaking up beyond 100 kilometers (62 miles). Controlling a single tower wouldn’t benefit a trader unless the goal was to force rivals to go around it by using a slower, less direct route, or hoarding the frequencies that companies rely on to transmit financial data, said Colt’s Cumberland."
Meta note, if anyone wants to talk trading algos or technology feel free to reach out, email in my profile.