Speed of light
Written by Elaine Graham
Elaine Graham looks at the rapidly changing world of trading systems and highlights how technology can help win competitive advantage
A seismic shift is taking place in the trading landscape. New high-volume, high-frequency systems based on computerised algorithms are driving up transaction speeds across the trade lifecycle, from placing an order to clearing and settlement. Measured in milliseconds, the goal is to achieve the lowest possible latency (the time it takes for a trade to go from initiating computer and back), to generate alpha and capture best-price liquidity. With over 50 per cent of wholesale FX trading now algorithmic and other asset classes catching up, how is technology supporting these trends?
One way to reduce latency is by ensuring a straight-through process across the trade lifecycle. Automated trading specialist, OpTier, provides transaction management technology to investment banks, asset managers and other financial services organisations. It monitors the end-to-end workflow across some 15 touchpoints, alerting the client CIO to a problem before it hits the organisation’s trading desk. Colin Rowland, senior VP, worldwide sales, says: “We can show exactly where a transaction went, how long it spent there, when it slowed down and its impact on performance.”
In addition to speed, the market wants a wider choice of trading venues, lower transaction costs, greater transparency, better compliance tools and more efficient risk management. In response to these demands, and as a result of MiFID’s deregulation of the exchanges, a new generation of fast,
flexible, low-priced equity exchanges are springing up: the Multiple Trading Facility (MTF), or central counterparty clearing houses (CCPs).
Alasdair Haynes, CEO at Chi-X, one of the new trading venues, comments: “We are attacking the exchanges by pushing down connectivity, data, execution and post-trade costs.” At the heart of these efficiencies is the Chi-X-built real-time matching engine (also used by Nasdaq). “The real skill is how you use computers to lower latency and get the fastest trade - now measured in microseconds and soon in nanoseconds. A platform that puts 20,000 trades through at the same time another can process a single trade is clearly better, but not to the point of degradation. At some point the system will break and then it becomes about reliability.” In the race for speed, Haynes thinks the industry is reaching critical mass. “When we started four years ago we were in a car, and others were on horses. Now everyone is in a car - so what then is your differentiator?”
The exchanges strike back
Faced with the threat from the MTFs, traditional exchanges are hitting back. Some, like the London Stock Exchange (LSE), which acquired a 51 per cent stake in share trading platform Turquoise in 2009, are buying up these rivals. Others, like Deutsche Boerse and NYSE, are joining forces to increase volume, create economies of scale and extend global footprint. Many are also looking to expand into downstream services like clearing and settlement. Not least, they are investing in new technology platforms to win competitive advantage.
The LSE’s new trading system, Millennium Exchange (bought from Sri Lankan provider, MillenniumIT), went live on Turquoise in November 2010 and the LSE in February 2011, with its Borsa Italiana due to follow. This followed a year of testing, including dress rehearsals with LSE’s 500 member firms.
Antoine Shagoury, CIO at LSE Group, comments: “This migration is a crucial step forward in our drive to offer best in class trading services and marks a key milestone in the introduction of tightly integration transaction technology across our markets, providing exceptional levels of performance, functionality and capacity.” According to the Group, Turquoise is now the fastest trading venue in the world, with similar latency statistics for the LSE. For margin-challenged exchanges, selling its proprietary technology also offers a new revenue stream. For the LSE, this means a focus on emerging markets like Africa and Mongolia.
While wholesale/institutional investors may benefit from these competitive improvements, what about private traders? Enter LMAX, which claims to be the first retail-based MTF. Up to now the retail sector - typically private investors, day traders and smaller entities using automated trading applications - were required to go through a broker who managed the credit and liquidity risk. This, however, adds to latency and traders often mistrust brokers because of their lack of cost transparency. Martin Thompson, CTO, comments: “By using LMAX they can bypass the dealing desk and link directly into our realtime risk engine which feeds automatically into the MTF. A trade is placed and as the market moves we track what is needed in the trader’s account so they don’t default, with an open window on the order book at all times. And because all of this is happening so fast there is no added latency.”
Banks with large hedge fund, asset management and other institutional investors are also focused on improving trade execution speeds and cost-efficiencies. According to Gaurav Sachdeva, business development, banking and capital markets Europe at Infosys: “Banks are looking at projects like multi-asset class platforms, integrating middle/back office systems, consolidating data centres and sharing data across different lines of business such as reconciliation, clearing and settlement and corporate actions. The drive is to have a leaner front office by creating common operations.” Sachdeva’s view is that latency is not the only - or even the most - important issue for the banks: equally critical are governance, data compliance requirements like anti-money laundering and know your customer (KYC) and counterparty risk management.
Trading room telephony has not been left behind. Traditional dealerboard vendors like IPC, BT and Orange Business Systems (OBS) are creating integrated solutions that can link voice trading turrets (some with up to 200 lines each) with applications on the sell-side, buy-side, hedge funds, derivatives, FX, bonds and more. Other applications enable voice calls (including mobile) to be recorded, stored, retrieved, monitored for compliance and trader performance, and integrated with emails, SMS and video content. IPB’s Jonathan Morton, VP of product management, says: “Most of our clients are large institutions who use a combination of low-touch (automated) and high-touch (voice/manual) systems. While a lot of execution is moving from voice to electronic, traders still need to communicate with each other, research analysts and portfolio managers so they can understand what’s going on in the market and get opinions in realtime.”
IPC, which works with over 100 exchanges, also manages high performance point-to-point connectivity infrastructures.
The company recently upgraded its London-Frankfurt route which boasts round-trip latency as low as 8.8 milliseconds. “Microseconds matter more than ever and latency remains a highly sensitive issue in the financial services industry. We optimise speed by using the most direct links between the most trafficked routes like London-New York, Chicago-Toronto,” explains Morton. “If one route fails, it can be quickly switched to another.”
Sometimes access to multiple touchpoints, preferably simultaneously and in real-time, trumps the need for point-to-point speed. To facilitate global collaboration, both IPC and OBS are developing cloud-based extranets and British Telecom’s Radiance is the world’s largest financial services cloud with 15,000 organisations. Howard Boville, head of financial markets and MD of trading systems at BT, says: “Radiance is a high-availability platform that enables everyone from Goldman Sachs to a small hedge fund to share data across the entire trade cycle.”
What does the future hold for the latency-driven trading community? As Chi-X’s Haynes points out: “Ultimately you can only go so fast before reaching a point of diminishing returns.” While this point may arrive sooner than later, no doubt by then the industry will be grappling with challenges that lie beyond the speed of light.