In the preceding chart, as well as in charts that follow, the numbers do not match those in the relevant appendixes because the charts were adjusted for nonresponses and for those who responded, "no basis to judge."

In another section of the bankers' questionnaire, we requested that bank officials indicate the degree of importance they think examiners place on the major areas they review during the course of an examination. These areas can also be divided on the basis of cause versus effect. Reviews of such indicators as liquidity and earnings can identify the effects of bad management and policies while evaluations of management itself might trace the root of a bank's problems. The responses to this question, therefore, give some indication of whether bankers feel examiners emphasize cause-oriented or effects-oriented areas.

By stratifying the responses to this question by supervising agency, we found that the Comptroller was perceived by bankers we surveyed as placing more emphasis on evaluating banks' managements. As was pointed out in a previous GAO report, 1/ the Comptroller has revised his examination procedures to place greater emphasis on evaluating a bank's management than either the Federal Reserve or FDIC. Accordingly, in some instances the proportion of national bank officials who thought the Comptroller's examiners emphasized cause factors was larger than that of Federal Reserve- and FDIC-supervised banks. As the following figure shows, the Comptroller was seen to emphasize evaluations of management competence and of a bank's policies and procedures more than were the Federal Reserve and

This difference was not as apparent in the three agencies' evaluations of banks' internal controls, however.

1/GGD-81-12, page 35.

But the bankers apparently did not equate attention to effectiveness. As the following figure shows, the percentages of national bank officials rating the Comptroller as being effective in performing cause-oriented evaluations were higher than percentages for FDIC's and the Federal Reserve's banks, but there is less difference among the agencies here than in the previous figure for cause-oriented areas.

Although agreeing with our observations, officials at one or more of the bank regulatory agencies gave us three primary reasons for not criticizing management before a bank's condition deteriorates. First, it is hard to convince a bank's board of directors that managers are doing anything wrong if the bank is in good condition. Second, until a bank's management is challenged with adversity, it is hard to tell how good it is. Today's economic conditions are giving bankers challenges they have hitherto not had to face. Finally, regulators do not want to appear to be interfering with management decisionmaking prerogatives.

Although it may be difficult to evaluate or challenge a manager's performance, that, nonetheless, is one of the regulatory agencies' responsibilities. Statements in examination policies developed by all three agencies indicate the importance of assessing the quality of a bank's directors and officers. Waiting for the effects of bad management to manifest themselves before criticizing managers--i.e., letting the bank deteriorate--is not the best way by which to fulfill supervisory responsibilities.

We agree that regulators should avoid unduly interfering with bank officers' legitimate prerogatives. But since the agencies do have the responsibility to evaluate the safety and soundness of the banks they supervise, and since the key to the condition of a bank is its management, then it is incumbent

upon them to make an objective, independent evaluation of management as one part of their overall assessment of the bank. Such an evaluation need not interfere with managers any more than would the adverse classification of a loan. One expresses the regulator's opinion of processes and the other an opinion of the results.

SURVEILLANCE SYSTEMS HAVE LIMITED VALUE

AS EARLY WARNING OF BANK PROBLEMS

Computer-based offsite surveillance systems are inherently limited in their ability to provide early warning of bank problems. Though not designed to replace onsite examinations, the systems were touted as between-examinations indicators of developing problems. Experience to date demonstrates limits on their value for this mission because they do not identify management weaknesses until manifest as changes to a bank's condition. Systems also cannot evaluate the quality of individual loans, an important indicator of a bank's potential condition.

Computer-based systems were originally conceived as one way, but not the only way, to spot bank problems before they become serious. Specifically, they were to help spot potential problems between onsite examinations, perhaps allowing a longer interval between examinations.

Our study disclosed that, although the systems did detect manifestations of problems before they became too serious to correct, offsite surveillance did not identify them until they had affected banks' financial conditions. In 12 of our 17 detailed case studies, surveillance systems indicated or could have indicated problems in the banks before those banks had to be placed on special attention lists. In only one of those cases did surveillance--a special regional analysis--cause the agency to change a bank rating and consider a bank to be of particular concern, independent of an examination. The other 11 cases exemplified surveillance's more usual role in problem identification--as a complement to the examination process, confirming conditions already detected during onsite examinations or soon to be detected by a routinely scheduled examination.

Because, as agency officials pointed out, surveillance systems analyze financial data but do not directly evaluate management or the quality of assets, they cannot pick up management weaknesses until they have adversely affected a bank's financial condition. This is why bank examiners we surveyed felt that the systems were least effective in identifying problems at an early stage or helping prevent problems from developing at a bank. Only 44.5 percent felt systems were effective or very effective

at identifying current problems at an early stage, and only 29.4 percent thought they were effective or very effective in helping prevent problems from developing. These percentages are considerably lower than the perceived effectiveness for other surveillance functions. (See app. III, quest. 12.)

The

Moreover, agency officials told us that their staffs do not rely on surveillance systems to monitor the progress of problem banks because they follow those banks closely, anyway. officials feel that the monitoring of banks with problems through surveillance is duplicative of the other supervision given to these banks. Once a bank is identified as having problems, other methods are used to keep the agency informed of the bank's condition. These include informal methods, such as progress reports and bank visits, and formal actions, like written agreements and cease and desist orders. Through these means the agencies are able to obtain information that is more current and more tailored to the bank's specific problems than the data supplied by surveillance.

In 16 of our 17 detailed case studies, at least one of the above methods was used to monitor the banks' progress. The agencies generally requested progress reports from the banks, and examiners visited them to check on specific areas of concern. Using these tools, the agencies are able to obtain and review bank information on a monthly or bimonthly basis--more often than the surveillance systems allow.

CONCLUSIONS

Even though they pay more attention to bank management as a cause of problems, agencies still equate the quality of management with the condition of a bank. Our analyses of sample data showed that examiners often cited problems at a bank without criticizing management practices, but as the problems worsened, the examiners began criticizing management more. Agencies rarely criticized management practices unless problems had already developed. One reason this occurred was that, in the view of bankers themselves, examiners placed a bit less emphasis during an examination on evaluating management. But the major reason was that the agencies are reluctant to criticize a bank's managers unless the bank's financial condition is deteriorating. Although it is hard to criticize management if a bank is doing well financially, our concern is, as it was in our 1976 study, that a bank's management should be evaluated separately from its financial condition. This is because a dip in the economy could have an unnecessarily severe impact on an institution that is poorly managed.