REINVESTMENT IN information technology infrastructure is a crucial issue for all banks. New technology options and products are changing rapidly under the weight of scientific advances.
To reinvest in information technology is very expensive and requires a vision of both the bank's long-term viability and the usefulness of the new technology infrastructure to the business. Reinvestment expenses are high because, while unit computing costs have decreased steadily for decades, the information technology infrastructure has expanded in scope even faster and now penetrates virtually every nook and cranny of every bank.
Banking is a consolidating industry suffering from overcapacity, erosion of market share to nonbanks, and long-term pressures on margins and profits. These facts, combined with the need for escalating information technology reinvestments, are leading to wide divisions among banks in how much they reinvest.
In fact, we believe all banks are now following one of three mutually exclusive information technology reinvestment strategies:
* Leaders: Banks that are quick to adopt new technology products and trends.
* Followers: Banks that apply only thoroughly proven technology and are spending only modest amounts on information technology infrastructure.
* Laggards: Banks that have essentially ceased reinvesting in information technology infrastructure and may incur future difficulties as a result.
Because reinvesting is so expensive, smaller banks are at a disadvantage.
Leaders - typically banks with $15 billion or more of assets - generally spend a minimum of $100 million yearly on information technology. They account for about 15% of the total industry spending on technology. Followers, representing a wider size range - anywhere between $1 billion and $15 billion in assets, account for 50% of total technology spending.
Laggards are found below $15 billion in assets, spend less than $100 million on technology, and account for about 35% of the industry total.
The Tower Group recently carried out an informal assessment of the information technology reinvestment strategies of eight banks to see whether this pattern held.
The participating banks have been disguised for reasons of confidentiality. We will identify them here only by letters, which have no correlation with the bank's actual name. All were over $12 billion in asset size and over $50 million in annual information technology spending. All eight had sizable branch networks but differed greatly in the size and sophistication of their wholesale and fee-service business lines.
We measured both the level of information technology reinvestment spending and the level of open systems reinvestments as two key guides to identifying which strategy each bank was following. Open systems strategy is critical because it represents such a pervasive change in how technologies are assembled to meet bank needs.
As you can see from the bar graph accompanying this story, reinvestment strategy is clearly related to bank size and business mix. Bank A, the leader, is quite large. Banks F, G, and H, the laggards, are the smallest of the eight.
Three selected new technologies - open systems, work group systems, and client-server technologies - show diversity of effort:
* Leaders had made extensive investments in these three technologies.
* Followers had made only minor investments in open systems, moderate investments in work group systems, and moderate to extensive investments in client-server technology.
* Laggards had made no investment in open systems or work group technologies and only minor investment in client-server systems.
Information technology reinvestment must be related to both the bank's business and its existing infrastructure. Leaders tend to have both a complex business mix and a complex set of platforms, hardware, networks, and applications. They also deliver information technology in a cost-effective way.
However, followers get caught in a trap. They have relatively complex product lines, but their business volumes don't match the leaders'. They've often tried to maintain various proprietary technological approaches for too long without having enough scale to really afford it.
Laggards tend to have simpler business approaches and simpler infrastructures. Because of this, their information technology spending strategy is generally more effective than that of the followers.
Four broad themes appeared in the information technology reinvestment thinking of all eight banks - leaders and laggards alike:
* Mainframe continuance. Cornerstone applications are being left on the mainframe with no current plans to migrate to another platform. The data security, distributed access, and heavy transaction capabilities continue to make it the platform of choice for these core banking products.
In addition, the software and overall infrastructure are not in place to fully exploit the symmetrical multiprocessing or massively parallel processing capabilities of some of the new hardware platforms. Lastly, the conversion costs of moving all applications off the mainframe would be horrendous. While the core banking systems continue to be mainframe driven, the user interface to these applications are often being redesigned on client-server front ends.
* Client-server penetration. Client-server architectures, local area networks, and, to some degree, open system technologies, are being employed by most of these banks to support individual departmental computing needs, not enterprisewide data sharing. In particular, workflow and decision support applications are the key candidates for transfer to these technologies. Providing the primary impetus for this move are the more sophisticated and interactive graphical user interfaces on the client-server platforms as well as the interactive nature of the data base management systems.
* Ease of access to data. To an increasing degree, data contained on the mainframe application systems are being stripped and downloaded to user friendly platforms. The line areas can massage and query the data more easily and the systems and technical staff can be relieved from excessive user demands. These data repositories are being developed for the most part in lieu of an enterprise-wide data model.
* Move towards vendor packages. Regardless of bank size, proprietary applications are being selectively replaced with vendor offerings. Internal bank information technology staffs increasingly focus on installing, maintaining, and linking these packages together.
In summary, every bank needs an explicit information technology reinvestment strategy. Leadership may make sense for the few with the scale and business needs to support it. Following may also make sense as long as complexities can be reduced. Even lagging may be appropriate for extremely simple businesses or for those to be acquired. The choice is important and should not be ignored.
Mr. Teixeira is president of the Tower Group, a consulting firm in Wellesley. Mass. Andrew Wasser, a technology analyst at the company, contributed to this article.
