« PreviousContinue »
examinations for all banks, though the period between examinations for sound banks is being lengthened.
Field office officials told us that they make relatively minor adjustments to examination schedules based on surveillance data. Schedules in one agency, for example, are established 6 months or more in advance. Since a bank's condition is considered when the schedules are set, and since the field offices usually know their banks reasonably well, surveillance results may cause a bank to be examined sooner than originally planned. For instance, after a September 1979 IMS report and a December 1979 call report showed serious problems in an FDIC-supervised bank, a planned August 1980 examination was rescheduled for April. The one exception to this pattern that we observed was a case in which one of the Comptroller's regions performed a specialized examination to investigate capital problems detected by an NBSS test that the region itself had designed.
As resource constraints lengthen the time between examinations, the regulators hope that surveillance can substitute for some onsite work. The Comptroller's staff said it will probably use more offsite analysis as a supplement to fewer examinations to pinpoint areas of concern. According to one official, NBSS will supplant onsite examinations to a certain extent when it becomes more of a forecaster and a risk assessor. Similarly, the Federal Reserve is studying the possibility of limiting onsite examination time by making more use of offsite surveillance and financial analysis techniques. However, according to one official, the surveillance timeliness problem must be solved before more offsite analysis can supplant onsite work.
The agencies' belief in the necessity of regularly scheduled onsite examinations will preclude them from using surveillance as a primary resource allocator. In response to a previous GAO report, 1/ the FDIC stated that it is
"* * * not aware of any monitoring system that is
1/GGD-81-12, page 63.
In the same report, the Comptroller stated that the increasing rate of change in the banking system and the economy may make more frequent but shorter overview examinations preferable.
FEDERAL COUNCIL WILL NOT
The Federal Financial Institutions Examination Council, established by the Congress in 1978 to better coordinate policies of the Federal financial regulators, is addressing some of the concerns we have raised here about surveillance systems. But prospects for an early successful completion of this task are not sanguine.
The Council was created in Title X of Public Law 95-630, and part of its function is to prescribe uniform principles and standards for the Federal financial institutions regulators. As one of its projects, the Council established its Task Force on Surveillance Systems to help formulate goals, objectives, and priorities for facilitating the development and sharing of computer-based and other surveillance procedures.
However, according to the task force's chairman, the Council will not soon find solutions to the problems we have raised. He said his task force cannot establish a uniform surveillance system until the agencies agree on a common examination philosophy. Another task force working on that question, we found out, is well behind schedule and is finding it difficult to reach agreement on common philosophies. Until they do, according to the surveillance task force chairman, his group cannot decide what they want surveillance to do and how it should complement onsite examinations.
Though the agencies have designed, developed, and implemented computer-based surveillance systems, there is as yet only an incomplete demonstration of their use or usefulness. The agencies feel that surveillance systems do offer advantages, and we have noted evidence of some of those advantages. But, it is not clear just how well those advantages can and will be integrated into supervision.
Moreover, we are concerned that the development process took place without the necessary justification studies being performed, and no cost-benefit data has been collected.
It is reasonable to assume that computer-based surveillance could play an important role in an era that will see fewer onsite visits to banks due to resource limitations by the regulators. However, the time is past due for a formal and rigorous evaluation of the capabilities and role of surveillance systems. The Federal Financial Institutions Examination Council has a task force working on this problem, but major philosophical differences among the agencies inhibit an early satisfactory solution to this issue. In the meantime, we believe each of the Federal regulators should study the cost effectiveness of its surveillance system.
Officials at FDIC and the Federal Reserve are cognizant of the reluctance by some examiners to use the surveillance systems Much of this resistance is due to incomplete training. The two agencies have introduced training courses to aid in the acceptance of the systems by agency personnel. At the time of our review, this training had not been given to all of the examiners that should be using the surveillance systems. Agency officials recognize that this deficiency exists, and they should continue and extend the training pro
so that the examiners requiring the surveillance knowledge receive it.
RECOMMENDATIONS TO THE BANK
We recommend that the Comptroller of the Currency, the Chairman of the Board of Governors of the Federal Reserve System, and the Chairman of the Federal Deposit Insurance Corporation formally assess the costs and benefits of their surveillance systems using recognized methods of systems analysis. Much of the data for such an analysis may already have been collected through the routine operations of the systems. The formal study, by focusing the agencies' attentions on this specific aspect of the supervisory process, could accelerate the proper integration of surveillance and onsite examinations.
Neither the Federal Reserve nor the FDIC was willing to perform a cost-benefit analysis of its surveillance system. They are committed to using them, although the FDIC is now going to "* * * overhaul the surveillance system operating procedures, and precisely define how and where the system fits into [its] overall supervisory mission *
(See app. V, p. 96.) The Federal Reserve said "* * * the role and benefits of monitoring and surveillance have been under continual review * * *." (See app. VII, p. 99.)
The Comptroller, on the other hand, agreed that cost effectiveness evaluations are desirable and also agreed to perform them as system improvements are made. (See app. IV, p. 90.)
As we previously stated, the process used by the Federal bank regulatory agencies to develop their surveillance systems did not meet accepted systems development criteria. result, the systems are evolving by trial and error.
We agree, in fact, that these systems may offer distinct advantages. But instead of perpetuating the processes that led to the current difficulties, the agencies should be using a more formalized, disciplined approach that requires surveillance systems to meet the same rigorous test that all systems should: the benefits accrued by using them must be proven to justify their development and operations costs.
We are pleased that the Comptroller has agreed to perform evaluations on future improvements. However, this does not substitute for a comprehensive cost-benefit study of the surveillance system in relation to its role in the bank supervisory process.
SCOPE AND METHODOLOGY
Since the use of statistical sampling techniques and computer analyses was integral to this review, we are presenting in this chapter details of our scope of work and methodology used, including limitations and interpretive comments. Our work was divided into three segments. First, from a universe of all cominercial banks supervised by the Federal Reserve, Comptroller of the Currency, and FDIC, we selected random samples to receive questionnaires surveying bankers' opinions about various aspects of supervision. Second, we sent another questionnaire to senior Federal commercial bank examiners at the same agencies asking some different questions and some that were similar to those asked of the bankers. Third, from the universe of commercial banks added to the agencies' lists of institutions warranting special attention between July 1978 and June 1980, we selected a sample for case studies by our evaluators.
SURVEY OF COMMERCIAL BANKERS
We chose two disproportionately stratified random samples of bankers to participate in the survey. The first was a general sample drawn from the universe of all commercial banks as it existed on November 3, 1980. The second consisted of banks randomly selected from the most current (December 1980) lists of special attention banks maintained by the agencies. These banks were not eliminated from the general universe and therefore a few appeared in both samples.
Because the three supervising agencies supervise differently and because different size banks behave differently, our samples take these factors into account. We disproportionately stratified both bank universes first by supervising agency and then by five asset size categories, ranging from assets less than $25 million to assets greater than $1 billion. This stratification assures that projections can be made to agency and bank size subpopulations as well as to the universe. It resulted in each universe being divided into 15 different strata. We then calculated separate sample sizes for each stratum.
The following figure lists the universe, sample sizes, and number of respondents for all strata in the general sample. The next figure shows information for the special attention bank sample, summarized by agency. Although we had also stratified our special attention sample by asset size, we have not presented those figures in order to avoid the inadvertent disclosure of bank identities pursuant to Public Law 95-320.