chants. This was long before the days when 49 of every 50 bank depositors were completely covered by deposit insurance. In that era, the majority of bank customers could, and probably did, read reports of condition to decide whether the bank looked safe or whether it might be advisable to shift to a stronger institution. It is important to bear in mind also that in those days the bank statements so published in accordance with law were practically the only statements that were published at all.

In our lifetime the significance of required reports of condition has changed greatly. Today, I venture to assert, only a tiny fraction of bank customers pause to read items headed "Report of Condition of XYZ State Bank * * *" published in accordance with call made by whomever it would be; instead, they relax calmly in the shelter of Federal deposit insurance. Those who are interested in the condition of a bank-such as the treasurers of corporations with millions on deposit are seldom misled by window dressing. They know it exists and make necessary allowances, checking against the surprise reports, and often they can directly ask banks for the information they want. They are not fooled.

For bank supervisors also, call reports are less important today as instruments of supervision. With extremely rare exceptions, the general condition of a bank does not alter substantially from month to month. Furthermore, supervisors have developed better alternative sources of information. Reports of examiners today are much better than they used to be. During my 30 years of bank supervision, I do not recall a single instance in which a dangerous trend calling for corrective action first came to our attention through a call report.

But reports of condition today serve one important purpose that hardly existed in 1900. In economic analysis and planning and particularly in the formulation of monetary policy, reliable bank statistics are a principal tool. Reports of condition, I venture to say, are the No. 1 source of these statistics for the American banking industry.

For statistical information of this kind, standardization of reporting dates is of great value. In many bank asset and liability categories, seasonal, even intraweekly, variations are astonishingly large. If reports were called for as of December 26 in one year, December 6 in the second, and December 16 in the third, even our own skilled statisticians would not be able to measure, with reasonable accuracy, movements in such basic items as deposits, business loans, and many others. Moreover, since most nonbank statistics to which banking data must be related are end-of-month figures, variable bank reporting dates detract considerably from the suitability of banking data for analyses of this character. Even the accuracy of the actual data reported would be better under fixed-date reporting than under surprise calls. Bankers have repeatedly informed us that it is most difficult to reconstruct an accurate report of condition retroactively for items not regularly covered in their daily trial balances. Because of these difficulties, many banks resort to estimating procedures that are often subject to a troublesome margin of error. With fixed-date reporting at the midvear and yearend, banks could arrange in advance for an accurate tally for each reported item as of the reporting date.

Against this background, I return to the fallacy that might be called the surprise call panacea. Plainly stated, this is the argument that

all the benefits of call reports would be retained and perhaps even enhanced and the evils, particularly window dressing, would disappear, if all calls were made on a surprise basis. What I have already said suggests some of the weaknesses of that argument. However, to evaluate it effectively, understanding of the actual call report situation is essential.

Under section 7 of the Federal Deposit Insurance Act, almost all banks in the United States are required to make four reports of condition annually to their Federal supervisors. The date of such balance sheets is the same for all banks. In actual practice, the regular custom has been to call for two of these reports, each year, on unexpected dates, usually in the spring and fall. The remaining two ordinarily are called for on or about June 30 and December 31 of each year. The latter are the principal source of the financial statistical series that I have mentioned. We see, then, that ordinarily there are two surprise calls every year. To the extent that publication of reports of condition called for unexpectedly are of benefit to the public, such benefit is derived from these calls. (However, in the opinion of at least one Federal supervisor, it appears, publication of those surprise call reports is of little benefit to the public. The Comptroller of the Currency has authorized national banks to omit current publication of those reports, requiring publication only as an adjunct to the reports that are customarily called for as of June 30 and December 31.)

The surprise call fallacy amounts to a contention that all benefits of call reports would be retained and all detriments eliminated if June 30 and December 31 were avoided as dates for the remaining two calls. Actually, however, this is far from the case. As I mentioned, a "call" as of December 18 would not prevent banks from window dressing, if they cared to, in preparation for a widely publicized yearend voluntary statement. In other words, even if the supervisors called for a report as of December 18, and in that report— because of its unexpected date-there was little or no window dressing, the report that actually comes to the public's attention, and which is the basis of the semiofficial size rating of the bank, would be a voluntary yearend statement, which would contain just as much or as little window dressing as the bank might wish. From that viewpoint, the only result of the December 18 call would be to impose upon all banks (including the vast majority that do not indulge in window dressing) the work and expense involved in the preparation and publication of two yearend statements instead of one.

Occasional issuance of an end-of-year call as of a date other than the last business day of December has at least one advantage, however. It helps to reveal to supervisors and other interested persons the extent of window dressing. This was demonstrated in 1962. Reports of condition were called for as of Friday, December 28, which, for most banks, was just 1 business day before the end of the year. It can be assumed that since banks are accustomed to December 31 calls, the December 28 figures were not appreciably inflated by window dressing. Most banks, particularly large banks, published_not only their call reports as of December 28, 1962, as required by law, but also voluntary reports as of December 31, and the latter were in more eye-catching form.

24-624-63- -2

Comparison on the figures of these two statements, just 1 business day apart, was instructive. The hundred largest banks in the country, according to a compilation published in the American Banker, furnish a striking example, although similar conditions exist in smaller banks as well. Those hundred banks showed total deposits of $121 billion on December 28, and by the end of December 31, 1 business day later, this had increased to almost $129 billion, a difference of more than 6 percent. Among the hundred, nine banks showed deposit increases, in 1 business day, of more than 10 percent. Individual figures ranged up to a high of 34 percent expansion. It is obvious that most of these increases were due to window dressing. Not all of them; some of them. Consequently, last year's experience provided a good picture of the magnitude of the problem.

Although window dressing sometimes has been attacked for the wrong reasons-mainly the loss of benefits to the supervisors-the undesirability of the practice must not be underestimated. It is not a negligible problem. The aggregate volume of window dressing, I suspect, does not vary so greatly, from year to year, as to distort seriously the bank statistics we need, but I have no doubt that, although allowance may be made for window dressing, the figures on which we depend would be more realistic and reliable if window dressing could be done away with. In addition, there is inequity in a system that enables a bank to pretend to the public that it is the largest in the community or State, when in fact it is not. Personally, I do not believe banks gain or lose much by this "numbers game," but some banks take it very seriously indeed, and that is why window dressing sometimes threatens to get out of hand as banks try to outmaneuver each other. One is reminded of what the Red Queen said in "Through the Looking Glass":

***here, you see, it takes all the running you can do to keep in the same place. If you want to get somewhere else, you must run at least twice as fast as that.

Window dressing, then, is an undesirable practice-an untruthful, unfair, wasteful, and misleading device. There is little doubt as to how a good banker would react to a borrower's statement that was inflated to show a more liquid position or a larger volume of business than in fact existed. To the degree these efforts succeed, they result in deceiving the public. And, to the degree they are recognized and discounted, they result in raising doubts as to the reliability of bank statements and of bankers' statements.

I must apologize for quoting from myself here. This is a part of the statement which I made to the Illinois Bankers Association earlier this year.

All bank supervisors, and most banks as well, would like to see the last of window dressing. But the "all surprise call" approach clearly is not the answer. As I have said, it would impose additional burdens on banks, would not materially improve public understanding, and might weaken essential statistics rather than improve them. In fact, an all-surprise program actually would eliminate the only existing penalty for window dressing. At present, a bank that window dresses its yearend report pays a larger deposit insurance premium, and this deterrent to window dressing, for whatever it is worth, would be lost if the required reports, on which the insurance assessment is based, were called for as of December 18, let us say, rather than December 31.

As the committee may know, a number of efforts actually have been made to diminish window dressing. Perhaps the most effective has been moral suasion-efforts to convince bankers that the practice is morally unworthy, that it could be injurious to the prestige of the banking industry-that is, to public confidence in the ethics of banksand that the game simply is not worth the candle, in the long run.

In my judgment, moral suasion is not only the most promising avenue toward the elimination of window dressing, but also the most desirable. Because of the complicating factors I have described, I am inclined to believe that the problem could not be solved by governmental fiat without an excessive degree of regulation and control. Actual experience indicates that bankers are prepared to stop this practice "if the other fellow will." Many are convinced that competing in window dressing is an unprofitable tug of war, but each participant hesitates to let go for fear that his opponent will carry off the prize. I believe that if the bank supervisory authorities, acting vigorously and simultaneously, would request banks throughout the country to quit window dressing, the likelihood of success would be excellent. But I emphasize that such an effort would certainly fail unless it was based on complete cooperation and coordination, most careful preparation, and determined face-to-face discussion with the bankers, State or National, in every city where the practice prevails. And after intial success, the supervisors would have to remain alert to chop off any new sproutings of this unhealthy growth.

A supplementary line of approach would be to require call reports to include daily average figures for important items. This might also be helpful from the statistical viewpoint, although it would add to the reporting burdens of banks-including the majority which do not window dress-and it would not, by itself, prevent window dressing in the yearend balance sheet, and that is where it is principally used.

I should like to summarize my ideas on this subject. Window dressing is an undesirable practice. Every reasonable effort should be made to eliminate it. Calling for all reports of condition on surprise dates is not a satisfactory answer. I believe that bank supervisors are in a position to develop a program, based on moral suasion, that will enable American banking to rid itself of this detrimental practice. If that is the answer, or if some other effective answer is found, it will be transmuted from a hope to a reality only through painstaking study of this complex question, with full and frank interchange of ideas and criticisms leading to cooperative action, among bank supervisory agencies and the industry itself. I hope that these conditions will prevail, so that efforts to solve this problem, and the many other problems that confront bank supervisors, can take place under conditions that offer the greatest likelihood of success.

(The attachment follows:)

DEVICES EMPLOYED TO "WINDOW DRESS" BANK CONDITION STATEMENTS CALLED FOR BY SUPERVISORY AUTHORITIES

1. Roundrobin exchange of interbank deposits among three or more banks which increases both deposits and cash-equivalent assets to make the bank appear larger and more liquid than it normally would. At least three banks must participate, since reciprocal deposits between two banks are required to be reported "net" in official condition reports.

2. Short-term reductions in borrowings, which member banks may offset by larger borrowings on other days of the reserve-computation period to maintain the required level of average reserves. This does not inflate the report's figures, but it does show a debt-free condition in published statements of the borrowing bank, although the payoff of the borrowing may be in the mail on the statement date and the loan account of the lending bank may not be reduced. 3. Arrangements with large depositors to increase their deposits temporarily by drawing drafts against their accounts at other banks. These drafts are credited to the customer's account immediately but are in the process of collection on the statement date and are not charged against the account at the other bank until after the statement date. This transaction may be reversed immediately after the statement date, so that there is no change in the allocaion of the depositors' balances in the long run.

4. Very short-term loans to cooperating customers, the proceeds of which are credited to the customers' accounts on the statement date and repaid immediately afterward. Similar results may be obtained by purchase of bank acceptances or open-market paper from brokers or nonbank dealers or by shifting of loan participations among banks. Payment is credited to the seller's account and the drafts used in payment are in transit on the statement date so that both loan and deposit totals are inflated.

5. Delayed processing of items presented for collection, or of interoffice clearings in a branch system. This is a simple and practically undetectable way of inflating total deposits and liquid assets and can be accomplished by holding back only a relatively few large checks without disturbing normal processing arrangements and without resorting to collusion with other banks or with customers.

INFLATION OF FIGURES IN VOLUNTARY PUBLISHED STATEMENTS BY METHODS THAT ARE NOT PERMITTED IN OFFICIAL CONDITION REPORTS

1. Voluntary statements may include reciprocal interbank balances which are required to be reported "net" in official reports of condition.

2. Voluntary statements may incorporate the assets and liabilities of foreign branches, which must be excluded from official condition reports.

3. Loan and investment totals and capital accounts may include bad debt reserves and other valuation reserves. They are required to be excluded from totals in the official condition reports of most banks.

Now, Mr. Chairman, I would be very glad to try to answer any questions that you or the members of the committee might have.

I have with me not only my lawyer but my statistician and they can provide any help.

Mr. FASCELL. I am going to nominate you the modern man, Mr. Robertson, since you are accompanied by the right people.

Mr. ROBERTSON. That always helps.

Mr. FASCELL. Thank you, Mr. Robertson. Your statement has been very candid, and you are certainly clear enough in your position.

I think as far as moral suasion is concerned, it is always nice to think that everybody is going to police himself, but I am afraid the facts of life are that that usually does not happen; whether it is in the banking industry, in Congress, or some place else. So, it seems to me, although moral suasion is, of course, the primary method, you need a little help.

may

Mr. ROBERTSON. May I give you an example, Mr. Chairman, of moral suasion that I think has worked in this field?

Mr. FASCELL. Yes, sir; by all means.

Mr. ROBERTSON. I might just say that over the years, and I have had a lot of experience in this supervisory field, there has never been a time, when determined interest on the part of the various agencies of the Federal Government was focussed on the same problem at the same time. One agency would be interested at one particular period, and the other one could not be bothered.