It is critical that banks learn to make more productive use of information.
Consolidation, deregulation, and diversification in product lines have all contributed to intensified competition in the banking industry.
Bank services and products have become narrow-margin commodities, sold to millions of customers and supported by huge distribution networks and processing infrastructures. In such businesses, information quickly becomes the most important management tool.
The information channel that banks rely on most heavily to guide decision-making is the enterprisewide management information or profitability reporting system.
In their efforts to improve their competitive positions through better use of information, most banks have invested heavily in these systems.
Many of these banks find that, once installed, these expensive information systems actually prove inadequate for decision support. Some fail because they were designed without clear objectives, others because they attempt to serve too many conflicting objectives and end up serving none well.
The truth is that most bank executives feel that they are not receiving full value from their management information and profitability systems. The numbers they see in their regular reports are seldom the numbers they need to make new decisions or evaluate the results of past decisions. They are left with the brooding suspicion that they ought to be getting something better from their systems investments.
We can uncover the underlying cause of their discontent rather easily. Simply challenge any bank to demonstrate how its regular monthly financial reports can be used, on their own and without supplementation, to analyze and resolve a current problem.
Not many banks will be able to meet this challenge. Decision-making will almost always require additional data and analysis. Over time the amount of additional work will expand, supported by a growing staff with strong analytic and data development skills.
The critical missing link in this process is the recognition that the needs of information users evolve. Everyone knows that the demands that banks face are changing continuously, and that bankers' understanding of their business is constantly evolving. Yet banks continue to design and implement reporting systems that are extremely resistant to change.
In response to the need for continual adaptation, the first principle in designing decision-support systems must clearly be flexibility. The architecture and data base that support the information and reporting systems must allow users to restructure information, add new information, or perform new calculations whenever necessary. These capabilities, in turn, require that the data base underlying the system maintain detailed account-level data and that the information and reporting system allow users to create information structures that will suit their evolving needs.
In other words, drill-down functionality alone is not enough. Users must be able to rebuild their information from the lowest level. A drill-down analysis of falling profit center profitability might, for example, reveal that underpriced loans are the cause of declining yields. Is this exception pricing illustrative of poor controls on pricing, or might there be some other reason?
A build-up analysis of the underpriced accounts might indicate that the low-rate loans were, in fact, parts of very profitable customer relationships that span several profit centers. Clearly the build-up analysis casts a very different light on the situation, and suggests different action than would the exception pricing information on its own. This same example also illustrates how a data base limited to aggregated data will undermine the ability of users to restructure their detailed information to meet new decision-making needs.
The second principle of systems design is that all new development must be guided by clear prioritization of system objectives. Banks may have several objectives for their monthly reporting systems. Some of these objectives, however, will require data, methods, or assumptions that are inconsistent with those required to satisfy others.
For example, suppose that a bank's highest priority for monthly reporting is to track actual performance of business units against budget and plan goals, and that its secondary priority is to assess the economic value generated by business units such as branch offices. The methods and assumptions required for these two objectives are substantially different.
Consider the difference in the ways the bank would allocate costs in these two cases.
In the first case, tracking actual performance, the bank might only allocate controllable costs (those that the performing manager or department can influence) or standard unit costs that are uniform for all managers or units being reviewed. In the second case, the bank needs to allocate all costs that could be eliminated if the business unit were shut down, even though many of these costs may be far beyond line management's control.
The total costs allocated under these two methods would be very different. So might many other quantities, such as revenues, numbers of accounts, or volumes used to compute unit rates. A reporting system would require different assumptions, definitions, and algorithms to serve each of these priorities correctly.
When second-priority objectives require different assumptions, methods, or data from the primary objective, a bank has two options.
The first is to implement all the methodologies required, and use each one as appropriate for its particular decision-making purpose. A reporting system built on this approach might produce several different sets of business unit profitability reports, each set using different data, assumptions, and methods. The second viable option is to implement the methodology it takes to meet the highest priority objective, and accept the need for manual data modification efforts to accomplish all secondary objectives.
Unfortunately, too many banks opt for a compromise between these clearly defined choices. They implement a set of hybrid assumptions, methods, and data in a doomed attempt to serve all purposes with a single solution. In trying to meet all objectives partially, they undermine the validity of results for each single objective, and ensure that every result produced by the system will require inefficient manual modification and validation.
To many bankers, these considerations may seem academic. After all, they've been using their reporting systems for years, supplementing reported results with experience and good business instincts. They may still view the vast quantities of information buried in the bank's data processing infrastructure as being relatively inaccessible.
Other industries have learned to exploit the value of information much more effectively. Manufacturers, for example, have harnessed information to drive "just in time" inventory practices and flexible production techniques. Retailers have capitalized on point of sale tracking capabilities to offer buyers improved selection at reduced costs.
Catalogue operations have learned how to track and evaluate customer preferences, leveraging the information through improved cross-selling efforts and sales of mailing lists to third parties.
Few banks have comparable abilities. For instance, few banks can readily answer questions like:
*How much of a business unit's profitability is derived from products or services sold in the current year (or month)?
*What is the real profitability of a total customer relationship?
*How does the profitability of a product vary with channel of distribution?
*Does cross-selling multiple products actually increase overall profit?
The banks that can master the use of information in guiding decisions and answer questions such as these will be the banks that survive and prosper. Information technology has already brought dramatic changes to the banking world, and even more dramatic changes are sure to come. By observing these two simple but powerful guidelines, banks can insure that their profitability analysis and reporting systems will allow them to meet the challenges of the future:
*Design and engineer the system for change.
*Define and prioritize business objectives for the system clearly.
One of the best places for a bank to institute these guidelines is in the design of its regular profitability reporting system. No other reports reach so many decision-makers. And, at most banks, no other system needs improvement quite as urgently.
Mr. Crandon is a principal of Treasury Services Corp., Santa Monica, Calif.