The core issue

FST. Legacy core-banking systems based on mainframes are the bane of most financial services companies. What are the key issues that you see bank’s still trying to address in this space?

SS. Many banks are constrained by their current inflexible, mainframe-oriented IT infrastructures as they look to increase customer growth, improve services, reduce costs and meet new regulatory requirements such as Basel II and Sarbanes-Oxley. This drive for improvement means there is a need for customers to replace their existing core banking applications and associated hardware systems.

FST. Providing more customer-centric systems is the goal of every bank, what are the best ways of approaching this challenge?

SS. With our adaptive enterprise methodology of simplification, standardization, modularity, and integration, and our understanding of the banking world, HP has developed a framework to assist in the transition from legacy to open and modular systems.

The approach we take is to evolve the traditional banking model from a complex framework that limits options and incurs high costs to a revolutionary new service-oriented architecture (SOA) based on a common service platform (CSP) called HP OpenBank, offering manageability and connectivity throughout the entire organization.

This architecture framework integrates banks from the front office through the mid office to the back-end processing systems. It provides multi-channel
integration and operational data store capabilities via an SOA, and focuses on improving customer service.

The OpenBank framework enables banks to:

• Integrate services and data from multiple product silos.
• Achieve a single view of the customer across all channels.
• Reduce the cost of supporting disparate or duplicated systems and applications.
• Provide real-time visibility into key business events and performance indicators.

By moving their core banking functionality to such a flexible, scalable, and modular open architecture framework such as, financial services institutions position themselves to become dynamic and competitive organizations.

FST. Migrating to more scalable, open-standard based servers can lead to greater choice and lower total cost of ownership. But how can bank’s manage this change without upsetting day-to-day operations?

SS. Depending on their unique situation and business goals, financial services companies can follow one of three deployment strategies to make the transition from their legacy mainframe system to a lower-cost, lower-risk solution built on an open, adaptive framework.

• Extend and enhance: If business change triggers the need for additional functionality but does not cause a major impact on the current core system, the institution can simply deploy the new capability on cost-effective platforms and integrate the new applications with the existing mainframe-based core banking system. HP and its partners offer a variety of FSI solutions and enterprise application integration (EAI) tools and services to extend and enhance an institution’s core-banking applications.
• Offload: If a key part of the current core banking application is not meeting the organization’s business needs, and it is too difficult or expensive (or both) to enhance it ‘in situ’, the bank could retire the obsolete part of the application and deploy a new up-to-date solution on a separate platform.
• Replace: If resources required to maintain and enhance the current core banking systems are too expensive, and each addition to the existing applications causes unexpected consequences and increases risk, it’s time to replace the entire core system and deploy a new solution based on an agile, industry-standard infrastructure with lower total cost of ownership (TCO).

FST. HP has promoted the use of modularity and SOA to address some of the issues of Core Banking. Why have you gone down this path?

SS. For several years, IBM has based much of their claimed performance leadership on strong published Power5 TPC-C benchmark results. However, banking customers have found that in comparisons using real workloads, IBM does not measure up to this supposed performance superiority. What IBM has done is take a niche hardware architecture and exploited certain loopholes in both the TPC-C and TPC-H benchmarks to hide weaknesses. HP’s SOA is supporting open systems and is defining common new standards, which is incremental in Banking to lower ROI.

The HP Integrity server platform is designed to effectively handle large batch or query processes in addition to handling low-latency transactional workloads. The Power5 and UltraSPARC chip architectures were designed for higher-latency transactional workloads. The Power5 architecture with its slower, more numerous, shared cache locations and a centralized memory bus, contribute to increased latencies as the size of the workload increases. In fact, IBM’s Power5 Servers are tuned specifically to maximize performance of workloads that are most similar to the TPC-C benchmark, which are more reflective of the work as done 15 years ago in a pre-internet, pre-business intelligence world.

FST. What’s the kind of ROI on investments in CB systems? Is it a quick pay off for what can be risky projects?

SS. Core banking replacement is the most complex, risky and expensive IT project that a bank can undertake. The cost of replacing a large core system is estimated by Celent to be US$400 million over 3-5 years. And there is a fear that promises of the upgrade won’t be achieved, such that banks often won’t replace until the feel they have to.
But when it comes to performance, customers have found that HP Integrity servers have balanced performance and consistent low-latency transfers for running banking applications, giving both excellent transactional and excellent batch performance when compared to IBM systems. Even more interesting, on the occasion of customer benchmarks on Temenos, Infosys, Open Solutions, IFLEX and SAP software, HP Integrity servers have out-performed IBM p-series by a significant margin. It also showed, that it is not about CPU power. Key success factor for good ROI is configuration and SOA compliance.

FST. Many financial executives take the view ‘if it aint broke. . .’ Why do they need to fix the core banking systems that have served them for so many years?

SS. While mainframes have supported core banking systems well in the past, today’s infrastructures demand simplicity, agility, and low-cost efficiency that older mainframes simply cannot offer.

Tower Group points out that, while there may be no immediate urgency for large FSIs to transform their core systems solely for IT cost reasons, since they can easily tolerate billion-dollar IT budgets, ‘they cannot afford to sacrifice competitive advantage, incur operational losses, tolerate unnecessary stock price vulnerability, and sink opportunities for top-line and bottom-line growth’.

Core system migration is a business issue. Financial institutions around the world are realizing that there are sound business reasons to transform their aging mainframe environments into robust, standardized infrastructures that better align with their business objectives – and that to delay doing so may well prove riskier than taking the initiative to begin the migration process.

However, long-term industry trends make core banking replacement inevitable. High maintenance costs cannot be justified. Competitive differentiation will come from improved operational efficiency, which requires open architecture and shared IT services and modern customer insight and risk management tools that need detailed, timely data with a combination of best third party application vendors such as Infosys, I-flex, SAP, Temenos and Open Solutions in combination with a Microsoft.net compliant front-end and supportive database applications from Oracle.