In the world of high-frequency trading (HFT) a saving of 1ms on the round-trip latency of a transaction could equate to a saving of GBP100 million (USD157 million).1 The promise of such financial rewards has stimulated the transatlantic cable market for the first time since the turn of the century. Operator Hibernia Atlantic is investing in the construction of a new high-speed, low-latency fibre-optic cable. The success or otherwise of this venture could lead to renewed cable-laying activity between London and New York.
HFT is the large-scale turnover of numerous financial positions within a very short timeframe, making a small return on each. It relies on the use of sophisticated computer algorithms and low-latency, fibre-optic data networks.
There have been numerous media stories putting a negative spin on HFT, some of which even cite it as the reason for the recent volatility in financial markets. However, while the practice is allowed to continue, the demand from traders for ever lower end-to-end transaction times between London and New York seems to be feeding a new market in transatlantic telecoms.
For much of the past decade, the perception within the market has been that of over-capacity on transatlantic routes. This was the result of two main factors: the large amount of capacity that was installed towards the end of the last century, and the ability of operators to drive more capacity from these assets using increasingly sophisticated driver technology. As a consequence, no new transatlantic cable has been built since the Apollo project in 2003.
The established transatlantic estate is more than adequate in terms of capacity. However, thanks to the age of the fibres and their topologies (driven by the technology available at the time of their construction), these cables cannot be retro-engineered to reduce latency to the extent required for HFT.
The overall performance of a fibre-optic cable is dictated by its link budget – transmission power less receiver sensitivity less losses incurred through cable splices, splitters and connectors. So although the transmit power can be forced higher by the use of more-powerful laser technology, the topology of the established submarine cables have too many negative factors within their link budgets because of the many splices and connectors along their paths.
In response to this lack of sufficiently low-latency transatlantic links, as part of its Global Financial Network (GFN), Hibernia Atlantic has embarked on ‘Project Express’ – a new low-latency London–New York fibre-optic cable that, it claims, will knock 5ms off the route, giving a less-than 60ms round-trip delay (RTD). Hibernia Atlantic claims that Project Express will provide the shortest route from New York to London, initially being lit with 40Gbit/s technology that can be upgraded to 100Gbit/s in future.2 The business case is based on the premise that the lowest-latency path could command a substantial premium from the HFT community and Hibernia Atlantic plans to extend the reach of this service to financial centres in continental Europe as part of its GFN.
Provided that the financial regulators allow the practice of HFT to continue, the requirement to achieve ever-decreasing RTDs on transatlantic cables could spur other cable operators to build competing services to Project Express. Hibernia Atlantic’s competitors will surely be monitoring the success of the project with great interest – potentially leading to a new transatlantic fibre-optic building boom, decreasing international bandwidth prices and improving capacity. Such a trend would ultimately benefit both business and residential users, as well as the financial community.
1 Mike Saunders, Vice-President of Business Development, Hibernia Atlantic – on BBC Radio 5 Live’s Wake Up to Money, 12 September 2011.
2 http://www.hiberniaatlantic.com/documents/ProjectExpress-PRFINAL-JSA.pdf