Adoption of advanced technologies is causing rapid change in the way FX markets work. How we got here, what that means today, and how that might look in the future was the matter for debate at last month’s London Summit.
As a line of business, FX has stood out as an asset class for Corvil, an IT data analytics firm that works with platform and market data providers among its FX clients.
“Last year, FX was our fastest growing business in terms of monitoring FX infrastructure…if you look at equities, for example, it took a good five, six years to really mature, but FX is doing it in maybe two years,” said Donal Byrne, Corvil’s CEO.
The distributed nature of the market, however, means the way the technology works is different. Speed still matters, but it’s on a global scale so the opportunity to arbitrage, or use technology favourably is much “wider”, explained Byrne.
In addition, the FX community is less sophisticated, although that is changing from a tech adoption point of view: “There’s a little bit of the wild west in there…it’s not as tightly controlled,” he added.
Arms race isn’t won
Also on the panel was Graeme Burnett, a high performance computing expert, who noted that the financial markets “arms race” is far from being over. He explained that technology that has been around for a while, like FPGA, which is used in the aerospace industry, is only gaining momentum.
“When you are receiving a packet as it’s coming off the wire, you are actually constructing the trade instruction going out at the same time,” explained Burnett. “You see the price, you can then flip the price in the output packet at the last second. That means there’s almost no latency on the turnaround on the trade of a very simple strategy.”
There are of course drawbacks to incorporating FPGA for trading purposes, for one, it’s very difficult to program, but the results are tempting investment and innovation towards development. “(It’s) only going to get more interesting as Intel and Altera bring out their new co-hosted chip sets with very fast interconnects,” he said.
Saeed Amen, consulting quant and founder of Cuemacro, said that barriers to entry are lowering. Using a programming language like Python, for example, makes it a lot easier to create, run and back test a trading strategy.
“There’s so many tools which are free and open source for…doing lower frequency trading in FX from a quantitative perspective,” he said. “If you are a retail trader, you can actually back test your trading strategies now without too much difficulty compared to the past, so you don’t need the same investment to get the same results.”
The FX market continues to be “a technology game”, said Victor Lebreton, director of Quant Hedge, a trading firm that deals in liquid FX pairs among other asset classes.
Big Data, News, Sentiment Analytics & NLPGo to article >>
“If you have microwaves, you are faster, if you go cloud and large big data sets, it is better for you. You get better information on the market,” he said. “So there’s still a technology race to be faster, better and aggregate flows.”
He also sees that more and more products could get caught in the net, such as NDFs and swaps: “Now with technology we can distribute it. It’s today, we see all this change. Why not go into a swap if it has this kind of ETF-like solution? You don’t need all the institutional infrastructure to trade.”
The future of advanced technologies in FX is exciting and uncertain.
Lebreton pointed to digital currencies, as well as private pools of liquidity as developments to keep an eye on. In addition, more autonomous artificial intelligence models could cause a shake-up for regulators looking to apportion responsibility for undesirable market practices.
Donal Byrne was emphatic on the future challenges faced in implementing MiFID II regulations dealing with time synchronization: “If you can’t tell time in machine time, you can’t run orderly markets. That is the situation that quite honestly our politicians, our regulators and our industry doesn’t want to grapple with.”
Cuemacro’s Amen said that the use of alternative data sets in trading is a development he’s looking forward to.
“There’s so much data being connected, for example, on the web, through our mobile phones…the internet of things; I think there’s a lot of potential there for that to be used to understand markets better and to make better trading decisions,” he said. “The key issue here is there’s so much data in question, which data sets are actually useful? That’s the big challenge I see going forward.”
Intelligence about speed
Graeme Burnett said that exchange technologies are catching up, pointing to Eurex as an example because the derivatives market now allows participants to see the timings of the individual elements of exchange operations.
“People had a much greater idea of what the inbuilt latencies of the exchange were rather than having to send fake trades to actually time the trade route, which was quite common,” he said.
Burnett also noted that analogue electronic switches are far more powerful than today’s message parsing, and expected to see more transmission mediums, such as RF, which can transit the whole globe at the speed of light, to become more prevalent over the next year.
Donal Byrne explained that intelligence about speed, rather than speed itself, is going to identify winners in the electronic markets of the future.
“Right now, FX markets are struggling to catch up from a speed point of view, but there’s already a whole new game in town…and if you’re not on that wave you are already way behind,” Byrne said. “We live in a world where there are multiple trading opportunities, multiple places across the world, and so the people who have the best information and the best analytics to take advantage of that will win.”