Sweeping away a sector’s chaos

Sweeping away a sector’s chaos
By Ross Tieman

Published: May 28 2008 01:25 | Last updated: May 28 2008 01:25

It sounds like an apocryphal story, but Nigel Woodward, London-based director of financial services at Intel, insists it is true.

“At one of the big UK clearing banks, the core accounting system still does calculations in pounds, shillings and pence,” he says. Decimalisation was introduced in the UK in 1971, 37 years ago.

The scale of the IT transformation needed in many areas of the financial industry is mind-boggling. Cobbled-together systems are still the bedrock of a hugely expanded sector accounting for an estimated 7 per cent of global gross domestic product.

While bad systems did not cause the present credit crisis, they probably contributed. “Some big banks failed to keep track of the risks as the volumes built up,” says Intel’s Mr Woodward.

He uses the example of sub-prime mortgages. When a bank bought a collateralised debt obligation (CDO), “was the transaction recorded and tracked back to a residential property in Texas,” he asks. “The bank might already have had a full exposure to property in Texas but didn’t know.”

Technology-enabled scale allowed traders to run ahead of banks’ ability to measure risk, he says. And when regulators and auditors started demanding answers about the scale of banks’ exposure, extracting the information from fragmented systems and databases was difficult and time-consuming. Hence revisions to banks’ profit warnings, as the scale of risk was progressively uncovered.

Jeremy Badman, partner in the strategic IT and operations practice focusing on investment banks at Oliver Wyman, highlights the problem that arose with credit default swaps, a mechanism used by banks to lay off risk that has turned into a market measured in trillions of dollars.

It started as a market where people fixed deals by phone, recorded them on a spreadsheet and faxed contracts. Back-office processing was manual. But as volumes increased, settlement remained manual, and three-month piles of unmatched contracts built up – alarming regulators over uncertain risk positions.

The lesson, says Mr Badman, is that technology has to support innovation, and processes must be “industrialised” quickly when a new product is successful. The trouble is that many financial institutions find this hard, because they rely on gummed-up legacy systems.

Rudy Puryear, global head of the IT practice at consultant Bain, explains: “Many of the IT solutions have been layered on over 15 or 20 years or more. In the 1990s everyone went out and wanted to buy a best-of-breed solution and then had to bolt that on to the legacy system. Then everybody wanted web access, plus companies have made acquisitions of companies using different systems.

“Almost every organisation I have walked into has a huge amount of unnecessary complexity in IT. It drives up cost and it slows down response in terms of time-to-market. We want IT to be an enabler of change. Right now it is very often like a block of concrete, adding rigidity to organisations.”

His recommendations? “You have to recognise that you have a complexity problem and that it is bad. It is driving up cost and constraining the ability to respond to the market-place and it is using up more and more IT dollars.

“You have to start saying you are not going to introduce more complexity. You have to create a future-state view of where you want to migrate this to in, say, five years time. You need to push a lot of shared, common, off-the-shelf solutions. So, as you make incremental decisions, you can measure it against how it helps you towards your desired five-year target.”

One example of this kind of thinking in action is Oyster, a ticketing system for Transport for London, by which users pay fares with a smart card, which stores cash, and can be used to pay for travel and other services.

Jonathan Charley, head of banking, Europe, at EDS, which advised on Oyster’s creation, says it was built as a stand-alone solution because “to integrate it into an existing system would have been a huge challenge”. The system was built on an off-the-shelf package of services-oriented architecture, put together “like Lego bricks”.

Clipping on ready-made flexible units that can take over tasks fragmented across existing systems seems a promising way forward. Charles Marston, who previously worked in the interest rate derivatives operation of a bank, founded systems and software company Calypso in San Francisco in 1997 to develop a universal front and back office platform.

Today, Calypso offers an off-the-shelf system that can be used to trade a host of financial instruments, from spot foreign exchange via derivatives to equities and commodities, yet which also supports straight-through back office tasks such as settlement, and allows banks to capture the data they need for risk and capital management. About 80 institutions have bought the system, including HSBC, Dresdner and Calyon.

As Peter Van der Vorst, chief financial officer of Sybase, an integration, data management and platform company, points out, one of the biggest challenges for many financial firms is keeping pace with the need to process vast and booming volumes of information at appropriate speeds.

So Sybase has just launched a product called RAP, designed to handle algorithmic computer-based trading, service the data needs of the quantitative analysts who write the algo programmes, and deliver the data needed to monitor trades for risk management and compliance.

Retail institutions, too, are finding legacy systems an encumbrance to business development. Nationwide, a UK building society, has decided to embark on a wholesale system renewal using an off-the-shelf solution from software house SAP.

Darin Brumby, divisional director for business systems transformation at Nationwide, says shifting to a new platform will enable it to introduce new products – different kinds of account, for example, and a suite of mortgages – that the current system cannot support.

It will also allow improvements to front and back office organisation. It is tantamount to creating a new building society around the changed market and customer needs. Although it is costly, “we think there is a good first-mover advantage”, he says.

SAP and US rival Oracle believe a pre-integrated offering is the best solution. Over the past few years they have been positioning themselves for the colossal orders that are beginning to flow as financial institutions start replacing legacy systems.

Rajesh Hukku, senior vice-president of financial services at Oracle, reckons the company has spent $30bn buying best-of-breed suppliers and developing a pre-built application integration architecture.

This one-stop-shop purchase of a core banking architecture with the features of your choice that are all promised to work seamlessly has won some other big converts. Citibank, the world’s biggest with 350,000 staff, is among them, replacing 59 versions of its old corporate banking system with a single Oracle solution, in which, for example, a base in Singapore services 14 banking operations in Asia. It is, says Mr Hukku, the biggest legacy system replacement ever.

The idea is that each bit can access all the data, and off-the shelf packages of analytics, for example, will keep a bank compliant with Basel II regulations, credit risk, and liability management, while assuring the flexibility to add in regulatory changes without complicating or compromising performance. “Two plus two equals five, if not 11,” Mr Hukku says.

It sounds like nirvana. And today, maybe it is. But will it still be the best answer in 10, or even five years? “We know that things will change,” says Mr Hukku, “but the basic requirement will always be to look at core data in certain aggregations.”

David Hunt, head of technology consulting at Capgemini Financial Services, agrees on the importance of data, but cautions that the IT industry still does not necessarily deliver all the right answers. “What we are not good at, as technologists, is doing that low-cost, throw-away innovation,” he says.

Yet financial services firms need to experiment with products as consumer technology changes.

Today’s private bank customers “may be happy to come to the office and have a fat cigar, but their inheritors might want to bank on their X-box 360 or mobile phone,” says Mr Hunt.

Tomorrow’s systems won’t just need to be agile, he says. In consumer, as well as investment banking, they will need to support rapid innovation of products, and rapid industrialisation of those that succeed.

It is a far cry from the days when they wrote that program in pounds, shillings and pence. Financial businesses are learning that they cannot see far into the future. System designers must learn not even to try.

Jerry Norton, head of financial services at consulting and software group Logica, deserves the last word. A layered approach that separates fundamental systems from distribution channels can help. But fundamentally, it’s about philosophy, he says. “Most other things – consumer products, even buildings – have a design life-time.”

Sure, a general ledger doesn’t change much. But isn’t it time systems were sold with an end-of-use date warning?
Copyright The Financial Times Limited 2008