--Are the changes working as intended?

--Does applying the changes improve the way supervisors
identify and deal with bank problems as compared to
our observations in our 1976 task force study?

If the changes in laws, procedures, and systems are working as designed or intended, and if we observe on a case-by-case basis better handling of problems than we did in 1976, then we can conclude that supervision has improved.

We used a three-step process to choose the changes on which to focus. First, we identified all the changes that had occurred as of the date of our review from other GAO work, annual reports published by each agency, and lists prepared for us by the agencies. Next, we categorized them by common agency supervisory function as described above. Finally, we selected for emphasis in our case studies and questionnaires those changes that had the most significant impact on those functions.

We used two methods to study how the three Federal bank regulatory agencies employed the supervisory changes we chose. First, we sent questionnaires to a statistical sample of bankers and to all senior bank examiners. The bankers' questionnaire was designed to elicit from high-level bank executives their perceptions of certain aspects of Federal supervision. In the other questionnaire, we asked senior bank examiners at all three agencies to describe how they used new supervisory tools and procedures and to evaluate those tools. Second, we studied a sample of cases to evaluate supervisors' handling of banks warranting special attention. Our case studies began with statistical analyses of a sample of 105 banks on Federal bank supervisors' lists of institutions warranting special attention. From this sample we selected 17 cases for a detailed review of the history of problems and agency actions pertaining to those banks.

Our review was performed in accordance with GAO's current "Standards for Audit of Governmental Organizations, Programs, Activities, and Functions."

Caveats

Because of resource constraints, the scope and methodology we chose for this review imposed certain limitations on our analysis. Those limitations are related to inherent characteristics of questionnaires and to the number and type of bank cases we selected to study.

A basic characteristic of questionnaire results is that they elicit opinions and observations of other parties without the presence of our own staff. Responses are, therefore, more subject to the interpretations and biases of the individual respondents. In order to minimize the effects of this characteristic, we pretested both questionnaires, in the presence of our staff, with bank officials and examiners.

Unlike our 1976 task force study, in this review we did not look at a sample of cases drawn from all banks; we concentrated our analysis on banks requiring special attention by supervisors. We did so for two reasons: first, the staffing requirement for case studies is quite high, and second, we were most interested in supervisory changes pertaining to the discovery and correction of bank problems.

However, we recognize that choosing this methodology excluded our reviewing cases in which the supervisory agencies may have identified and helped banks correct emerging problems before they became serious enough to warrant placing the banks in a program of increased supervision. In addition, it should be pointed out that only a small percentage of banks--3.91 percent as of June 1981--warrant the concern such as that given the banks in our case sample universe.

We attempted to amplify the results of the case studies through our questionnaires. The questions we directed to bankers asked them to evaluate bank examination for all banks, not just special attention banks. Our analyses distinguished between answers from banks warranting special attention and those from banks in general. In addition, the questions we asked of bank examiners concerned the way they examine banks in general, not just special attention banks.

CHAPTER 2

DATA GATHERING IMPROVED, BUT PROCESS

COULD BE MORE EFFECTIVE, LESS BURDENSOME TO BANKS

Overall, regulators now get a better and more complete picture of the banks they examine than they did during our task force review. Our case studies showed that examiners are getting adequate data on banks that are developing problems. The financial analyses made by new computer-based surveillance systems are more sophisticated than those that were previously available using noncomputerized methods. The Federal Reserve is inspecting holding companies with more regularity, using standard procedures, obtaining data not previously collected, and using surveillance to make more extensive financial analyses than before.

Agencies have attempted to streamline their information collection by modifying the scope and procedures used in onsite examinations. Modified examinations have been shown to effectively utilize agency resources, and the agencies are not yet using them as extensively as they should. However, potential difficulties with regard to examiner training may limit the extent to which agencies can expand use of modified methods.

Information collected on insider transactions and reported by banks exceeds supervisors' requirements. Under a 1978 law prompted by the revelation of abuses, banks are required to maintain information and report to Federal supervisors on loans to executive officers and principal shareholders. Federal regulators have recommended simplifying these requirements, and our review indicates that one report filed with regulators by banks is not actually necessary for examiners to obtain adequate information on insider loans.

CHANGES IN DATA COLLECTION AND ANALYSIS
HAVE IMPROVED INFORMATION ON BANKS

Compared to the period covered by our 1976 task force study, the Federal banking regulators are getting better and more complete information on the institutions they regulate. Agencies still use the same basic sources for information they did in 1976--onsite examinations and periodic data reported by banks. However, the scope, types, and frequencies of examinations have changed, and the periodically reported data is now analyzed by computer-based surveillance systems. Whereas in 1976 we found little had been done to obtain complete information on bank holding companies, new reports, procedures, and systems are being employed in this area, too. A general trend in gathering information exists: an increased emphasis on offsite data gathering and an associated decrease in onsite examinations.

Surveillance systems give supervisors better data on banks financial conditions

Compared to what was available before the surveillance systems were developed, agency personnel now have a better picture of a bank's financial condition and performance trends. The systems offer a much greater level of sophistication and are perceived as a valuable tool.

Supervisors were using financial ratio analyses at the time of our 1976 study to evaluate various aspects of a bank's condition, such as capital adequacy, income, and liquidity. These ratios had been calculated by hand, thus limiting the extent to which financial profiles could be developed. However, even then the agencies had begun to create computer systems to calculate the ratios and perform analyses to provide a more extensive, in-depth picture of each bank .

Since that time, each Federal banking agency has developed its own surveillance system, though an attempt is being made through the Federal Financial Institutions Examination Council to standardize them. The systems vary in design and concept, but each attempts to use financial ratio analysis to evaluate a bank's financial condition, show performance trends, and spot potential problems.

The Comptroller's National Bank Surveillance System (NBSS), primarily designed for the early detection of banks requiring special attention and as a supervisory and administrative system, consists of three computer-based elements. The Bank Performance Reports are produced from a data base obtained from reports submitted periodically by banks and bank examiners. The Bank Performance Reports show financial data and ratios calculated for each bank and for peer groups against which each bank's position is compared. The Anomaly Severity Ranking System, the early warning device for NBSS, monitors banks between examinations to detect those banks which show abnormal positions when compared to their previous status and to their peers. This is accomplished by a computerized scoring system which allocates the highest score to those banks having the most abnormal positions, changes, and trends in performance or composition. Those banks receiving the highest scores receive greater supervisory attention. The third element, the Action Control System, is a management information system that records the problems identified in a bank and shows the progress of the efforts being made to correct these conditions. Bank conditions which have been entered into the system for monitoring cannot be removed until

corrected. Once removed from the active file, the conditions are retained in a historical record.

FDIC's Integrated Monitoring System (IMS) was developed to supplement the examination process by providing a method of monitoring bank performance. By monitoring banks between examinations, the IMS can alert FDIC to the presence of a deteriorating situation before it reaches a serious level, facilitating a faster response by the agency. The primary component of the IMS, Just A Warning System (JAWS), acts as an initial screening device for data supplied by the banks. This data is furnished to staff via computer terminals, before FDIC corrects bank reporting errors that often occur, so they can immediately begin to determine the validity of the data as well as to perform preliminary analyses using the JAWS tests. JAWS consists of certain selected tests intended to measure capital adequacy, liquidity, asset-liability mix/growth, and profitability. The bank "fails" the tests if its calculated financial ratios do not meet certain thresholds. The tests are then analyzed and forwarded to each region to assist in completing review and analysis. Another IMS component, the Comparative Performance Report, was designed to provide bank management with a statistical and analytical tool for use in evaluating performance and generating decisions. The report is produced from information submitted by the banks and compares individual bank data to peer group data.

The objective of the Federal Reserve's Bank Surveillance Program is to identify member banks with current or potential financial problems. The Program was developed to detect deteriorating financial conditions of member banks, to aid in scheduling examinations for State member banks, and to support the examination process through bank performance reports. The screening component of the system identifies existing and emerging financial problems through the analysis of changes in financial ratios and a composite score relative to peer group statistics. The information for these screens is submitted by the banks and is generated by the Comptroller for the Federal Reserve. The Comptroller's NBSS supplies Bank Performance Reports for State member banks. Reports for those banks that failed ratios are sent to the Federal Reserve banks for further review and analysis.