Core banking system replacement is risky and complex but is an increasingly likely scenario for banks. Like their smaller peers, larger institutions may need to opt for packaged solutions to keep costs and time-to-market down.
Replacing a core banking system is often compared with replacing the engine of a Boeing 747 in mid-air. Experts consider core banking replacement the most complex, risky and expensive IT project that any bank can undertake.
A core banking system forms the backbone of a bank’s IT infrastructure and contains records of all customer transactions and the processing of those transactions. A small error in this area can cause a bank’s entire system to crash, tarnishing its reputation in the process. So while banks are keen to offer customers the latest products and services and to use the latest technology, they are exceedingly cautious about proceeding with a core banking replacement programme.
When to replace?
Banks are driven to evaluate their incumbent core banking systems when they face operational issues such as introducing a new product requiring code changes; integration problems between the core banking system and ancillary solutions and difficulty in providing customers with a unified view of their financial information. And when a bank’s goals of higher efficiency and greater flexibility, coupled with the ability to meet compliance requirements, are not met due to the limitations of the current core banking system, it needs to replace this system.
The main reason for change at FöreningsSparbanken (Swedbank), one of the largest banking groups in the Nordic area, was obsolete technology. Developed in-house in the 1970s, the COBOL mainframe-based core system had been subject to several enhancements over the years. Along with archaic technology causing problems of inflexibility, high cost of maintenance and low efficiency, one of the main problems was the absence of adequate documentation.
Lack of documentation
“Some of the modules didn’t have any documentation. Enhancements and upgradations over 30 years had completely changed the system and we were getting dependent on people who knew how the system worked,” says bank chief information officer Gert Engman. With regard to the risks involved with replacing the solution, he says that there was much greater business risk in having a system that was not documented down to the last line of code.
Raiffeisen Zentralbank Austria (RZB), one of Austria’s leading commercial and investment banks that has grown through mergers and acquisitions, decided to replace its non-homogenous mix of core banking systems, when it realised that maintaining a diverse set of systems with separate development requirements and interface management issues was extremely inefficient and expensive. Three years previously the bank decided to replace the various systems with a single core banking solution, Globus, from Swedish vendor Temenos. “The advantages of replacing an old inefficient core banking system far outweigh the risks and costs involved,” says Heinz Weidner, chief information officer at RZB. He adds that developing a solution in-house was never a serious option. “Software development is not one of our core competencies and over time we will never be able to be as cost efficient and cutting edge as a software vendor. We want to be the best users of IT, not developers of IT.”
Changing trend
The strategy to buy rather than build a core banking solution has traditionally been popular with small and mid-tier bank, which don’t have the wherewithal to develop such a complex, mission-critical solution. However, the trend seems to be gradually catching on with larger banks. These banks, especially the largest, have traditionally had big IT departments and have focused on developing most of their systems in-house. As in Swedbank’s case, core banking systems in these large banks are a complex mass of software. Usually developed over several decades, many of these systems are at the end of their lifecycle and now need replacing or dramatic modernisation.
An issue of resources
At Swedbank, which has an 800-strong IT team, the decision to buy rather than develop a solution in-house was obvious, says Mr Engman. “Although we did evaluate the possibility of developing the solution in-house, we soon realised that it would have taken too much of our resources. A core banking system is very complicated and it would have taken us nearly two years to define the functional specs itself. We do a lot of software development in-house but that is in the area of business application development and the web, not core banking.” Opting for a packaged solution, Swedbank bought a mainframe-based core banking solution from local vendor TietoEnator.
Mr Engman believes that most banks today would make a similar choice, encouraged by the fact that core banking has matured in recent years. New generation solutions are much more robust, based on web-enabled architectures and offering the benefits of flexibility, faster time-to-market, increased cost savings, and real-time banking.
Pilot implementation
The trend seems to be picking up, though large banks maintain a mixed attitude towards the in-house vs packaged debate. Last year, HypoVereinsbank Group (HVB Group) of Germany, one of Europe’s top five banks, chose Indian banking software vendor i-flex Solutions’ core banking solution Flexcube. An initial pilot implementation at FMS Bank, one of HVB’s subsidiaries in Germany, could potentially encompass all banks in the HVB Group, including parent banks HVB and Bank Austria.
HVB Group, with its own 3000-strong IT department, evaluated developing the solution in-house – but decided it wanted a cost-effective, quick, reliable solution. The packaged solution won on cost and time-to-market benefits. According to HVB’s spokesperson: “As long as we are talking about small and medium-sized subsidiaries like FMS Bank, a packaged software has all necessary features that makes it cheaper than an in-house solution.”
While there are examples of large banks opting for packaged solutions, such as Citibank replacing its in-house developed core banking system with i-flex’s Flexcube solution and ING Bank trading in its wholesale core banking systems with Temenos’ Globus solution, but not many top-tier banks have yet followed suit.
Amit Dua, group manager, European sales at Indian banking solutions vendor Infosys, says that although many core banking systems in large banks have reached a “saturation stage”, where replacement is the best option to see any tangible benefits, it will take time for these banks to think beyond in-house development.
He says: “There needs to be a very strong champion within the bank. The HR implications of replacing a core banking system are huge, not only on the organisational side, but also in terms of the cost of re-training bank staff like tellers, etc. It is a very politically sensitive subject.”
Targeted solution
Temenos, in association with IBM, has devised a strategy aimed at these large banks. Temenos Core Banking is a solution targeted specifically at the largest retail banks worldwide that operate on mainframes. Explaining the idea behind Temenos Core Banking, marketing director Jean Michel says: “The open systems and mainframe space are completely different and we are targeting them accordingly. For example, banks that traditionally buy packaged solutions operate more often in the open systems space and Temenos T24 is targeted at them.
“Banks that operate in the mainframe world are usually banks that believe in building core banking solutions. They operate differently, have different development methodologies and Temenos Core Banking is targeted at these banks, to make their transition from in-house development to buying packaged software easier.”
Change on its way
While core banking technology vendors develop strategies to woo banks, industry experts are positive that large banks will replace their in-house developed core banking solutions with packaged solutions. According to reports by research firm Celent, approximately 70% of the top 100 and 50% of the top 1000 financial institutions in the world have internally built core banking systems in place.
Celent believes it will be almost impossible for a large financial institution to remain competitive without replacing its core system. It predicts that as banks begin to replace their systems, an increasing number will look to cheaper and less risky vendor-built solutions on non-mainframe platforms.
Celent predicts that the top 100 global institutions will spend, in aggregate, around $3bn in 2003 and approximately $6.5bn in 2005 on new core banking software (licence fees, purchasing costs), hardware and professional services.
In association with CSC
