How complex would it be to develop an information architecture in a diversified group spread across geographies? A financial services conglomerate in the Middle East was grappling with this issue some time ago. The conglomerate had grown through acquisitions, and hence it had acquired a myriad pool of information technology (IT) platforms and systems.
A retail bank of the group based in Kuwait was using Tandem servers and associated devices for its retail, wholesale and investment banking services along with a banking solution developed by a Singapore firm. An asset management firm of the group based in Kuwait was on Unix servers while an investment banking arm in Bahrain was running its systems on Midas-Kapiti (Misys) on the AS/400 platform. The group had at that point acquired financial services institutions in Dubai, Jordan and Tunisia besides a hotel and diary business in Saudi Arabia, a boutique bank in Belarus and an investment firm in the US. The Dubai-based bank was using IBM RS/6000 servers on AIX for its retail and wholesale banking applications. The Jordanian bank had a home-grown banking application on Unix servers while the export-import bank in Tunisia was using Midas-Kapiti (Misys) on the AS/400 platform.
The group financial controller (GFC) was having a challenging time in collating timely information across the various banks and instituting metrics and triggers. Some banks were efficient while there were laggards who were even missing their local central bank reporting deadlines. The challenge for the group was that despite investing in high-end computing and well-known banking systems, the primary objective of seamless information dissemination seemed like a pipe dream. A large part of the problem was due to the disparate systems that had come together, due to the group's peculiar nature of growth which would not readily communicate with each other.
The GFC was part of the group steering committee responsible for developing operational road maps. Given the challenges faced by the group, an implementable IT road map seemed to be the need of the hour. The GFC engaged consultants from India, the UK and the US. The various broad options the GFC developed from the different pieces of advice he received was along the following lines:
nMake Midas-Kapiti banking solution a group standard: The impacts would be more than $5 million invested on the Tandem platform, the IBM/other Unix servers and banking applications across the group; long gestation and high cost conversion/integration of banking applications across at least three of the group's banks; reskilling/training of more than 100 full-time skilled employees.
nMake Unix servers-based platform a group standard: The impacts would be over $5 million invested on Midas-Kapiti contracts; long gestation and high cost conversion/integration of banking applications across at least three of the group's banks; reskilling/training of more than 100 full-time skilled employees.
nDevelop service-oriented IT architecture for the group: The impacts would be - Simple Object Access Protocol wrap-around projects could potentially exceed $5 million given the scale and diversity of systems; protect the major IT investments made by the group; add a systems group at the head office along with the required IT infrastructure.
The GFC needed a convincing assessment of the broad options to present to the steering committee. He chose a complexity modelling firm, which presented him some uncanny insights. He was told that irrespective of any of the broad choices the group made, they would need to make investment and IT reorganisation decisions in the short term that would affect customer and investor service, probably affect client acquisitions, and likely lead to legal issues.
The long-term impact of achieving an overarching IT architecture, besides significant investments, would be uncertain or partial success of implementation, and the potential lack of flexibility to acquire banks with disparate IT systems.
The complexity modelling firm had developed the above scenarios using a multi-level influence and impact modelling approach. The key intervention stages involved the starting stage, followed by a series of stages determined by the management's decision based on changing trends in business climate. They advised the GFC to hold the status quo for the next six months and just add a systems group at the head office for collation, dissemination and other coordination activities. They further advised him to take advantage of changing business decisions to streamline the IT systems across the group during the subsequent six to 12 months while at the same time incorporating a systems policy in the group's growth decision-making process.
GR Chandrashekhar
Associate professor, Complexity Research Group, Institute for Financial Management and Research
Associate professor, Complexity Research Group, Institute for Financial Management and Research