10.11.2013
By Terry Flanagan

Exchanges at Center of HFT Debate

High-frequency trading is becoming a regulatory as much as a technology issue, with exchanges at the center of the debate.

Under the German High Frequency Trading Act, which came into force on May 15, 2013, trading Participants must flag orders and binding quotes generated by them through algorithmic trading and identify the trading algorithms used.

Eurex and Deutsche Boerse earlier this month published rule changes relating to the German HFT Act. Under the new rules it will no longer be enough simply to differentiate between algo and non-algo orders.

Starting April 1, 2014, all orders sent by an exchange member to a German venue must include a RegulatoryID referencing the history of any algos involved in creating that message. It will become necessary for trading firms to define a taxonomy of algos, according to Christian Voigt, product manager at Fiedessa.

For example, looking at the same automated trading system, some firms might argue that one single algo is in use with many different parameters, while others may claim that they run many separate algos in parallel with each fulfilling a distinct function, Voigt said in a blog posting.

This raises questions, such as: How would the new rules be applied in those circumstances? Also, the observed behavior of a particular algo in the market could differ depending on whether it was run on older hardware or on the latest server technology in exchange co-location. Would that justify different RegulatoryIDs?

“These questions will undoubtedly remain high on the global regulatory agenda for some time to come,” Voigt said. The Exchange Supervisory Authority of the State of Hesse in Germany is planning to publish further guidance, and similar Algo IDs are being discussed as part of both MiFID II and the CFTC’s concept release.

The technology aspects of HFT are keeping pace with the regulatory.

For example, high-frequency traders can take advantage of more consistent and reliable trading performance thanks to a new technology platform launched by Supermicro.

The Hyper-Speed SuperServer line has been upgraded with accelerated Ivy Bridge processors and system firmware optimizations that lower median latency by 15% and decrease jitter by 94%, the company said.

“HFT has become synonymous with speed and in a world where being first is everything and second is nothing, every element of performance matters,” said Mariano Salomon, FSI business manager at Supermicro UK. “But from a technology perspective, having the fastest technical capabilities does not automatically lead to the highest profits. It is no good having the fastest gigahertz if your competitor has a solution that delivers a more consistent overall trading experience.”

The launch of the Hyper-Speed comes at a time where high speed trading continues to be a highly prominent form of market making, with global HFT trading volume set to increase by more than 20% by the end of 2013, according to some estimates.

Traditionally dominant across equities and commodities, the practice is starting to become more influential across other asset classes such as foreign exchange (FX), where HFT now accounts for over 40 per cent of global currency trading.

“With HFT now powering such a large percentage of trading volume across multiple asset classes, market participants can ill-afford to ignore the importance of ensuring continuous uptime,” Salomon said. “Any drop in performance could lead to significant losses at the end of the trading day. This is why we are seeing more and more firms invest both time and capital in the technology that underpins this complex form of market making.”

Related articles

  1. Agency broker moves beyond execution to offer a broader suite of services.

  2. Algorithms have become more prevalent in the spot FX market.

  3. Congress Unlikely to Act on HFT

    QB’s Algo Suite for futures market trade execution is also being co-located to HKEX.

  4. Breaking data silos is key to deploying automation beyond 'nuisance' orders.

  5. They can be used on quantum hardware expected to be available in 5 to 10 years.