Feeling the strain of high-volume trading

Rising high-frequency trading volumes has left a number of banks feeling processing strains at the back office over the past year or so in particular. At our FX Invest congress in Chicago earlier this year, one of the main themes that ran throughout was that back-office systems at banks were not always equipped to support this new way of trading foreign exchange. In some cases, some prime brokers were said to process 50,000 tickets a day.

If platforms can try and help prime broking banks manage this volume through netting services, that's surely a good thing. Traiana's system developed for spot broker EBS proved to be successful. During the pilot with ABN Amro, the service NetLink cut ticket volume by 95% by netting dozens of prime customer trades into single tickets.

Certainly, as the market expands and opens up to different styles of trading foreign exchange, it should also evolve to support those changes. EBS and Hotspot appear to be doing so, while settlement platform CLS is still looking into it.

Earlier this year, the settlement platform was asked to undertake an industry-wide review of launching a netting service by the CLS board. The platform said the growth of high-frequency trading meant a small group of CLS member banks were facing capacity constraints, with systems not designed to deal with seven-nanosecond computer-based trading models.

However, the issue with CLS offering this service is that netting then settling wasn't part of the deal when banks signed up. The system was established as an industry initiative to eliminate settlement risk, and the charging structure for this worked on a volume basis – the higher the volume, the lower the ticket costs.

By netting, the ticket volume decreases, therefore making further cost reductions less likely. And all this for what CLS said was a small – but significant – group of member banks. The results of that research were due at the CLS board meeting at the end of April. Needless to say, given the complexity of the issue, no decision has been made.

It seems the platform is between a rock and a hard place. While algorithmic trading is unlikely to reach the levels seen in equities, it is possible more banks will pick up these traders as clients in the near future. So, should CLS come up with a solution now, or hold out until the concept gains critical mass? You decide. Cast your votes at www.fxweek.com.

Saima Farooqi, Editor