the need for speed in financial services

Sandra O'Boyle from Current Analysis explains why algorithmic trading has made network speed essential for investment banks.

The cliche "every second counts" has become irrelevant to the world of investment banking, as measuring progress in terms of seconds is of little value to traders who need to capitalize on pricing and trend movements measured in milliseconds. There are situations where traders would rather not trade at all than trade late. As more and more trading is electronic and automated, any weak links that create delays are increasingly the difference between profit and loss. The key driver in this market now is algorithmic trading, which has created exponential growth in trade volumes.

automated trades

Algorithmic trading is a process whereby factors such as timing, price or quantity of the order are determined by machine rather than man. In its most extreme form - known as high frequency trading - orders are initiated electronically based on information received electronically, with trends identified and acted on more quickly than is humanly possible.

Regulators frequently express concern that a large percentage of these trades is never actually completed and thus distort the market, but in an industry where shaving fractions off share prices can yield massive returns, algorithmic trading is here to stay.

This means that low network latency is very important for traders. Latency rates should remain the same regardless of the level of traffic (e.g., which can spike when the U.S. market opens) or problems with the primary route - by possibly switching to a preconfigured secondary route. Distance is another key consideration - the closer the customer is to the exchange (in reality, the distance between the data centers hosting the customer's and the exchange's servers), the better.

data demand

The relentless growth of electronic exchanges is another important factor influencing network choice. Electronic trading platform provider Chi-X has been rapidly expanding in Asia and now provides trading services in Japan - the world's second largest trading center - and Australia. Chi-X and other multilateral trading facilities sell themselves on the basis of lower fees and faster trades compared to the established bourses.

Then there is growing interest in emerging markets, particularly those with liberal trading regimes such as Brazil's BM&F Bovespa. In Southeast Asia, the Kuala Lumpur, Bangkok, Singapore and Philippine stock exchanges will be linked, and the Vietnamese and Indonesian exchanges are expected to follow suit in a development that is being closely monitored in Japan and South Korea.

Such developments highlight the impracticality of expecting any single network provider to have major data center facilities at every key location in the developing as well as the developed world - which is where fast and reliable access to markets' data comes into its own as well as partnerships between operators to share data center facilities.

Sandra O'Boyle is Service Director, Business Network and IT Services at Current Analysis. Find out more about Current Analysis at www.currentanalysis.com.