HSBC trades from a clean IT slate

In early 2005, banking giant HSBC granted Kevin Bourne, managing director and global head of execution trading at its investment banking arm, two precious freedoms that made him the envy of his peers: the first was permission to retire the group’s aged legacy systems that so often frustrated IT innovation; and second, the clear authority to engage both business and IT in the creation of a single, global equities trading system.

In an arena where technology deployed in different regions is not normally aligned beyond basic connectivity and where high-level replication is usually thwarted by vested IT interests, this represented a simple – and rare – “attack of common sense,” he explains, “and one driven by a fundamental understanding of how financial markets are changing”.

Faced with the emergence of far-flung markets, new distributed pools of liquidity and the growth of complex asset classes, HSBC had resolved to revise the operating model of its equities business. Equities, which had functioned almost as a standalone unit, was consequently integrated into – and had its operating model aligned with – the core of the investment bank, says Bourne. “At that point, it was decided it would be opportune to look not only at the operating model for the equities business,” he adds, “but to look at the technology model to support its related products such as equity finance and equity derivatives.”

Cue Bourne’s entrance at the bank, alongside other business hires, and the birth of a uniquely empowered role in which the realms of both business and IT were effectively aligned. “HSBC said they were looking for someone to come in and manage the equities execution products and also to have responsibility for its core architecture. That person not only needed to know the electronic trading discipline but also, by default, have an understanding of how a trading system might be built that would actually deliver those highly automated needs.”

Tasked with creating nothing short of a world-class electronic trading platform that would embrace 92 different markets across 76 countries, the business and IT team was given just two years to delivery – an “extremely aggressive” timeframe considering the industry average for such a systems stretches from six to 10 years. However, the team had one key advantage: unlike most of their industry peers, they were permitted to start from a clean sheet of paper, to rip out the existing architecture of HSBC’s nine legacy equities trading systems and replace them with an all-new design.

“The first thing we decided was we didn’t need nine trading systems, we need one. It needs to be replicated, it needs to be based on four primary hubs and then distributed down to our local offices around the world, and it needs to be real-time.” A grand vision, some would argue, but in fact, Bourne demurs, “it’s not actually as difficult to implement as you would think – when you’ve got a good IT team together and they want to engage closely with the business”.

Taking a single trading application (Fidessa from equities trading software specialist Royalblue Group), HSBC has, during the course of the past two years, established those four primary hubs in London, New York, Hong Kong and Tokyo, from which the system is then deployed downstream into 20 further regional dealing centres. Where the system had to be connected to localised markets, some customisation was required to accommodate variations in regional stock exchanges. “But fundamentally,” adds Bourne, “the technology, the platform and the workflow model that we use to oversee how the traders fulfil their responsibilities is now the same globally.”

In December 2006, the last hub, in Tokyo, went live on schedule. This final and complex delivery was completed on time largely as a result of the overall architectural model which made both the deployment and the conceptualisation of the project much easier than anticipated, says Bourne. “You can run the project on a global basis centrally quite easily, because the only variant is the individual connectivity to the local market; the core operating system is the same in each country.”

The choke point

“Effectively one instance, but replicated”, the system adheres to a common data model and a common transaction model. Front-office staff in New York, for example, use the same trading system as a regional trader in South Africa, a model that brings obvious advantages where business continuity and, in particular, disaster recovery are concerned.

But of equal importance for HSBC, which as a group spent a hefty £245 million on R&D during 2006, the common architecture model allows the bank to optimise its intellectual capital, and advance its continuous product development. If a trader in Hong Kong comes up with a bright idea for using the system more effectively, that intellectual property can be captured and syndicated to the rest of HSBC’s trading systems globally. It is an approach that may jar with the individualistic spirit of some traders, but one that saves the bank significant resources by eradicating instances of duplicate development – a repeat offender in most large, IT-driven organisations.

Now, says Bourne, the bank is able to focus on a release cycle of four deployments a year, at which it completely updates its entire underlying software base, with the result that “compliance and regulatory best practices are deployed through the system globally”.

With functions replicated globally, moreover, HSBC has been able to make enhancements to the underlying technology to encompass tasks beyond the scope of the system as originally conceived.

None of which is to suggest, however, that the project was without its challenges. The most noteworthy, according to Bourne, were not, in the main, technological but managerial.

“Some people feel their region is different, but that is rarely the case,” he explains. “The demand management task therefore becomes quite a challenging one, as there are always 10 units of demand for every five units of budget, but at least the investment is monetised by staff efficiently.”

Reconciling those inevitable issues has required strict discipline, says Bourne, who was in the enviable position of having broad control – financial, structural, architectural – which helped him reconcile the demands of different stakeholders. Bourne describes his role as a “choke point” between the business and the technology sides, pointing out that his authority came from other senior managers who fully appreciate the benefits of committed, technological development.

“It’s not that we’re smarter than anybody else at building systems,” Bourne summarises, “and it’s not that the technology we’ve used is any different – we’ve used a vendor application.” The key, he concludes, was being able “to deploy that technology on an enterprise-wide basis, using an architecture model that [unlike the efforts of its rivals] allowed HSBC to start from scratch”.

Pete Swabey

Pete Swabey

Pete was Editor of Information Age and head of technology research for Vitesse Media plc from 2005 to 2013, before moving on to be Senior Editor and then Editorial Director at The Economist Intelligence...

Related Topics