Regulators may have high-frequency trading firms in their sites and real speed limits may be coming soon, but this isn’t stopping traders who have clients to please and bonuses to earn.
Speedy trading firms are considering new technologies to approach the speed of light and technology providers are offering new innovations to attain those speeds. One provider is Strike Technologies, the New York-based firm that has set up a series of ultrafast microwave towers that form a secure connection between data centers in Chicago and New Jersey. They describe StrikeNET as a "physical microwave-and-millimeter network designed, built, and operated by our RF team, connecting North American exchange data centers along near-geodesic paths. …. Built around our lightweight, RF optimized market data protocol, StrikeNet delivers ultra low latency market data feeds and bandwidth without sacrificing reliability."
We spoke with Shawn Melamed, global head of commercial products at Strike Technologies, about who is using this brave new technology.
Is this a new technology that investment firms are using right now or is it still too new and untested for trading?
Shawn Melamed, Strike Technologies:
Shawn Melamed, Strike Technologies:There are probably 20 to 30 firms using it today. It’s relatively new obviously. The technology has been around only for a year now - really live - but people have been using it. Some of the larger firms initially built it internally for their own trading purposes and not through a vendor. What you see now only recently is the rationale to build a commercial offering and we’re spearheading that (offering). The vendors who offer that tend to offer the economies of scale. It’s a very heavy expense, much harder than just buying from a teleco provider the same type of line.
Can you describe these firms? Are they large institutional investors or smaller hedge funds?
Melamed:A combination of both. We have the larger institutions and banks and you get the prop firms that are trading equities, futures, FX - it depends. A lot of the benefits of microwave technology is to provide one asset class data at another data center. The ability to get futures data, to trade equities or ETFs, is one thing - the ability to get futures data to trade treasuries is another.
What are the firms' chief concerns about adopting microwave technology?
Melamed:In general, the key concerns are threefold. The first one is reliability. Just because it's susceptible to weather conditions from where you’re transmitting the signal. The second concern is bandwidth. Unlike traditional fiber where you can turn on wavelengths and enhance terabytes of bandwidth on a given strand of fiber, with microwaves and the way physics works, and limited transmission frequency spectrum that is available and regulated by the FCC - you can only transmit on along a whole path in an order of magnitude of hundreds of megabits. This is a dramatic and significant difference compared to the capacity and bandwidth we are used to.
And the third concern is security. The problem of transmitting the data over the air is that somebody can extract it. So how do you protect your information?
How do you answer these concerns?
Melamed:On the reliability issue, the weather and rain can reduce the signal level on each of those antennas. What you want to do is create a flexible network that can read and understand the weather around it, almost like each tower having its own weather station and understanding any given point of time what is the micro-climate that exists around the tower.
And then you'd have the ability to change radio modes to either add error correction capabilities in real time when the weather is bad or drop error correction capabilities when the weather is good to improve latency. We built a sophisticated network that can compensate for weather conditions that are harsher and optimize for low latency or weather conditions.
We're talking about tens of hops, tens of different towers along the way, it can do that at different points on the line. So if there are storms all the way through Ohio, we can turn on more error correction capabilities there.
What about bandwidth?
Melamed:Well, here you can’t deal with it with a physical aspect on the appliance. You have to develop a software-based solution and we built a unique proprietary self-engineering protocol that can compensate for these issues and take the feed, say the CME Futures Fast 6 feed, and filter it, compress it, and extract only the key data that people care about.
Melamed:The third issue is privacy. For example, we encrypt everything that goes out over the line. Everything that leaves the radio is encrypted in 256-bit encryption.
Will we see transatlantic microwave transmission soon?
Melamed:Microwaves are a range in the electromagnetic wave spectrum that can transmit a signal between five to 50 miles. That’s the range. When you are looking to transmit longer range over large bodies of water, you have to look more to shortwave radios, which is closer to the AM FM spectrum.
The reality is the longer you move to this spectrum for transmission, you lose the bandwidth. So, if you move to short wave radio, you can only transmit a few bits at a time per second. So that becomes quite limiting for the data you can send over the ocean.
In addition, when you transmit over the oceans the types of radios you can’t really transmit in a direct session because of the amplitude of the wave form. So you would have to shoot that radio wave to the stratosphere and bounce to the water and then bounce back to the stratosphere. It will do that two or three times before it gets from New York to London.
So there are other challenges with wireless transmission. And the alternative of going through satellites, you are adding 60 miles or more just to go up in the air to the satellite and back to Earth.
Are we close to having drones transmitting market data?
Melamed:That would be expensive, right? I mean, what’s the Defense Department’s budget?
There are a lot of new ideas, like having balloons hang in the air that adjust their location. I don't think these are close to reality yet or justify the cost. I think more likely we’ll see very effective transatlantic lines that go through water but then have microwave connectivity at the landing sites. This can easily save you 10 milliseconds.
I think you'll see those hybrid approaches in the near future.