The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.
FEDERAL DEPOSIT INSURANCE CORPORATION W A S H IN G TO N , D. C . 20429 Remarks of K. A. Randall, Chairman Federal Deposit Insurance Corporation Washington, D. C. before the National Automation Conference of THE AMERICAN BANKERS ASSOCIATION in New York, New York Monday, May 8, 1967 This conference will be concerned during the next several days with almost every aspect of automation— from hardware and software re quirements and specifications to the broad implications of automation for banking. Rather than delve into the mysteries and intricacies of the automation revolution, however, I would like to concentrate today-if only briefly--on a few of the problems and issues deserving of particular emphasis from our viewpoint and worth some repetition. Then I would like to tell you a little about what the Corporation is trying to do in this area of computer applications to banking and why the Corpo ration has decided to devote a significant portion of its efforts in this direction. This conference affords a unique opportunity to exchange experiences in automation— especially in the application of the com puter to banking problems--as well as the opportunity to assess the long-run prospects for new developments in this exciting and challenging field. As we become increasingly involved in the use of computers in banking, however, we must be careful not to lose sight of some of the fundamental constraints inherent in this field of activity. The very newness of the subject matter tends to engender a spirit of optimism. Tasks that are almost unbelievably difficult look mi giPifl/Hngiy easy--especially in what might be termed the early versational” stages. con The distant goal more often than not appears attainable within months or a year or two at most when, as a matter of fact, it may very well be purely a time illusion. As long as this - 2 - situation is recognized, the task becomes less discouraging and hopes are not as readily danpened by the lack of immediate, concrete results. Optimism and enthusiasm are necessary ingredients in motivating efforts to apply technological advances in data processing to problems in banking and finance, but they alone will not and cannot solve all of the problems. The computer imposes a hard and exacting discipline of its own. The interval between the initiation of a project and its effective operation is usually much longer than expected. Even the solution of rather simple problems entails much planning, testing, and redesigning before the operation can be reduced to a routine process. The amount of lead time between the conception of a project and its final operational form can be discouragingly long--as it has been to the Corporation in seme instances and to many of you. This is a basic fact that must be faced by all of us, however. As the problems become more intricate, moreover, it seems quite likely that the lead time interval will become longer rather than shorter because mastery of techniques may not be able to keep pace with the increasingly complex problems and relationships that can be put on the computer. So, not only long lead times, but significant initial startup costs in terns of equipment and manpower are needed. There is another and even more important variable in this automation equation that must constantly be borne in mind— and that is the human variable. It is easy to be impressed with the versatility - 3 - and adaptability of computer technology— but it should also be obvious that its value is heavily dependent on its creators and users* It is the human element that exercises selectivity in computer applications and the element that must have a basic understanding of the advantages and disadvantages of this modern tool. It is trite but true that the quality of the end-product can only be as good as the quality of the input; no startling metamorphosis takes place within the computer itself* On a more prosaic level, sharp and insistent questions must also be asked about the quality of all of the data processed by a com puter. For example, when banking statistics were compiled by old- fashioned methods, the results were oftentimes primarily of historical interest only— or used to embellish a speech such as this one and occasionally as support after the fact for rationalizations of theory* But if we intend to base important administrative decisions on these results, it is essential to have data that reflect reality as accurately as possible. Perhaps the most serious charge that can be leveled against the computer today is that much valuable time is devoted to the pro cessing of data that fail to meet minimum qualitative standards. From this standpoint, it is almost impossible to overstress the need to the highest quality data possible or the necessity to abandon data that do not measure tip to standards. The relative importance of what is put on the computer is a question that deserves more careful consideration at this time than - k - ever before. Expenditure of time and effort on computerization of projects that are not worthwhile is one of the major pitfalls that can beset computer applications. This is particularly true as total information systems with computers as an integral element are developed. Seeing a man sitting next to the paraphernalia that gives him remote access to a computer suggests important and meaningful activity. The value of the computer, however, is determined less by the fact that it is being used than the use being made of it. The computer is a demanding and costly tool that should be reserved primarily for sig nificant work. It can be helpful in performing routine chores, but its greatest potential lies in its use as a problem-solving device and management aid. Whatever contribution the computer can make to assist the executive or administrator maximizes its usefulness. At the present time, however, because of its newness, it may well be that, in the field of management science, development of the hardware and technical side of automation may very well have outrun the adaptation of computer capabilities to the needs of top management. In the process of adapting computer techniques to management decisions, it might be appropriate at this stage to mention again the valuable intellectual discipline that the computer imposes upon the user. The computer can force the mind to analyze problems more system atically and to view a project in its entirety— proceeding from a logical beginning, through an orderly process to a clearly stated objective. - 5 - This concept of problem solving— •which possibly has to be more clearly stated and more rigorously applied when computer facilities are employed— can be counted as one of the side benefits of the computer revolution. The Federal Deposit Insurance Corporation is also devoting its attention to automation— from two somewhat different but related directions. The primary area in this field to which we are allocating our resources is the determination of the contribution that automation nnri computers can make toward the strengthening of bank supervision. The second area of interest concerns the progress of automation in banking itself and an attempt to evaluate its impact on banks and bank management and on bank supervision. The Corporation's interest derives from its concern with the total environment of banking. Our supervisory responsibilities demand that we be cognizant of banking practices or policies that affect banking or its structure. Initially, the Corporation is using the computer to assist in the task of maintaining information on banks— essentially a house keeping chore. Since the beginning of bank supervision over a hundred years ago, quantitative data covering the regulated banks have been accumulated through reports of condition and exam ination. The information was consolidated and analyzed in an attempt to identify characteristics of the entire banking community or certain specific segments. Because the process was time consuming, the information had only limited utility for supervisory purposes and was largely of historic interest. - 6 - Application of computer processing to this information makes possible the production of more timely and, therefore, more useful data for both supervisor and banker. The qualities of speed and precision that make the computer such an attractive tool to the banking industry likewise make it very promising for bank supervisory purposes. This, however, is not the most important application of modern technology to the supervisory process. It is analogous to computerization in your own banks of the bookkeeping for deposits. Our horizons and interests go beyond merely facilitating the handling of records to an investigation of the factors that influence bank management decisions--and therefore bank supervisors’ decisions. To the extent that computers have played a role in the decision-making process of banks, we are concerned with their impact. But even more important is the possibility that computer techniques will give the Corporation the capability to conduct studies that will increase our knowledge of an individual bank's operations and the relevant variables. The computer could modify the form of bank supervision— as well as make our decisions more timely and relevant. In contrast to much of our approach in the past where we dealt largely with aggregates, current research efforts are being directed in a micro- rather than macro-economic direction. In other words, with greater computer capability, we can feasibly study indi vidual banks other than through the examination process alone. the research projects that we axe contemplating or are already If - 7 « sponsoring in the area come to fruition, we as bank supervisors will have a much clearer picture of the impact of economic conditions, interest rate ceilings, and changes in market structure, for instance, on individual banks than has heretofore been possible. As a by-product, we foresee also the potential to generate information that could be valuable in turn to banks in making management decisions— that is, the development of a worthwhile and workable information feedback system. In attempting to increase our knowledge of the hanking process, some additional information will probably have to be collected from banks--as well as possibly some new types of information, which you may already use internally. The development of central data files and management information systems can constitute a real contribution to individual bank m anagem ents as well as bank supervisors. The benefits are not confined exclusively to one or the other— and, to the extent possible, we hope to make available to students of banking and the interested public much of the results of our studies as our contribution to a better public tinderstanding of banking. We are presently trying to determine what our role should be in this process— as well as the kinds of the information and the means of transmission that will be most useful to us and to banking. We are concerned that the product and efforts be commensurate with the time and money expended. We are working conscientiously to understand banking and the changes that are constantly occurring, and we are trying our best to be - 8 - as knowledgeable and helpful as we can be. The Corporation can best fulfill its responsibility to the public by acting as a positive force In improving our understanding of banks and the banking process. We are searching not only for the technological answers to banking questions but trying to determine the most effective role of the supervisor in the evolution of the bank-supervisor relationship. The Corporation's efforts in the field of computerization, management science, and the development of a management information system will help to upgrade the quality of management as well as bank supervision without injecting the Corporation into areas of responsibility that properly belong to bank management. So far, despite some problems, our efforts in these new directions seem to hold much promise. to a successful conclusion. We are committed to seeing it