with the more rapid expansion that has occurred in their liabilities during the past 30 years. Some reduction in reserve requirements would not necessarily impair the liquidity and safety of banks' assets if banks are prudent in the use of additional funds obtained. Any substantial reduction in requirements, however, might raise questions about the adequacy of safety or liquidity in the asset structure of banks, unless offset by other additions to liquidity.

To the extent necessary to avoid undue credit expansion, reserves released by any reduction in requirements could be absorbed by Federal Reserve sales of securities in the market. This would in effect shift earning assets from Federal Reserve banks to member banks. The present System portfolio is adequate to permit a substantial reduction and still leave enough to provide sufficient earnings to cover necessary expenses, as well as for current purposes of policy. Any decrease in requirements, however, should leave the Federal Reserve with a portfolio adequate to cover possible future contingencies, such as a large inflow of gold or economies in the use of currency that might add reserves in excess of appropriate needs. In view, moreover, of the growing international liabilities of this country, the reserve base of member banks, as well as that of Federal Reserve banks, should be maintained at a level that would permit further reduction of requirements if needed to cover a future drain on our gold reserves.

While normal growth of the economy may require some increase in the supply of money and in needs for additional reserves, there may be large potentialities for economies in the use of money, which make any projections of future needs unreliable. Reserves should be released only as needs actually arise and not in anticipation of possible needs.

Any substantial release of reserves within a short period of time can have disturbing effects on credit markets and on the economic system generally. Although the amount of any such release can be offset by adsorption of the same amount through Federal Reserve sales of securities, the initial uses that might be made of the actual reserves released cannot be controlled. Increased leverage for credit expansion permitted by lower reserve requirements may also present problems of adjustment.

For these various reasons, any reductions in reserve requirements have to be made in relatively small amounts over extended periods of time. They should be made only at times when money market and credit conditions are such that undue credit expansion in some lines would not occur before it could be brought under control. The Federal Reserve should have no mandate or commitment to reduce requirements to some specific level within a definite period of time. Legislative authority should be sufficiently flexible that changes can be made in a manner, in amounts, and at times that do not conflict with the needs of monetary policy.

CHAPTER 4. BASIS FOR DIFFERENTIALS IN RESERVE REQUIREMENTS

In considering the question of an appropriate system of reserve requirements for commercial banks, one of the most difficult and elusive problems is a determination as to what should be the basis or criteria for reserve requirements from the standpoint of both monetary theory and institutional practice. This problem is concerned not only with the level of required reserves in the aggregate, but also with differences in requirements for different types of deposits and classes of banks. Differentials in reserve requirements as between types of deposits and among classes of banks determine the distribution of reserves among individual member banks. These differences, moreover, contribute to the effectiveness of reserves and reserve requirements as a fulcrum for regulation of the supply of money and bank credit.

As previously pointed out, it is now commonly recognized that the principal function of bank reserve requirements in modern monetary systems is to serve as a means whereby the central bank can exert control over the volume of deposit money. If volume of deposit money were the only concern of monetary regulation, then reserve requirements should presumably be a ratio of total deposits with no distinction by type of deposit and be identical for all banks. The existing system of reserve requirements for Federal Reserve member banks contains differentials by classes of banks and by type of deposits. These differentials imply some principle of distinction as to the money quality of deposits of different types and in different banks.

The principal reason for maintaining such distinctions, rather than having a single percentage figure applicable to all deposits, is the significance of varia

tions in what may be called the money quality of different deposits as means of payments. The effect of monetary holdings upon economic activity is determined by the use of money in economic transactions, not merely by the holding of idle balances; the outstanding volume of money has significance as an indication of the availability of money for use, i.e., as an index of the liquidity of the economy. Variations in the economic contribution of monetary holdings are roughly indicated by differences in the rate of use or the velocity of turnover of different deposits. Changes in the rate of turnover of existing deposits may have as much, or more, influence on economic activity and prices as variations in the quantity of money, and hence are of concern for the effectuation of monetary policy. Later in this chapter there is a discussion of the difficulties of measuring turnover for this purpose.

INTERBANK DEPOSITS

Historically, differentials in reserve requirements, though applied according to location of banks in particular cities, were based primarily upon liquidity or convertibility concepts. The more volatile deposits bore higher reserve requirements. Under the National Bank Act, reserves consisted of balances with correspondent banks, together with vault cash holdings. Higher requirements against deposits in cities where banks held large amounts of balances due to other banks were designed to limit the possibility of bulding up a pyramid of reserves upon a narrow base of specie or money of ultimate convertibility. With a banking system in which balances with other banks can be counted as reserves against deposits, requirements that banks hold reserves in the form of currency or balances with the Reserve banks agains their balances due to banks serve to limit total credit expansion. In such a system higher requirements against interbank balances help to limit potential credit expansion on the basis of a given amount of basic reserves.

Under the Federal Reserve Act as amended, interbank balances have not counted as reserves since 1917, but the classification of cities for reserve purposes has been to a large extent based on holdings of deposits due to other banks. The only eligible reserve assets since 1917 have been balances with Federal Reserve banks, and it is now suggested that vault cash holdings be made eligible. The aggregate of these generally cannot be increased except through the use of Federal Reserve credit. Interbank balances can be built up through interbank lending or other reciprocal arrangements. Balances with other banks still serve as reserves for banks that are not members of the Federal Reserve System and thus to a limited extent may provide the basis for multiple credit expansion.

Correspondent balances at banks serve as liquid secondary reserves for the banks making the deposit because they can be readily drawn upon to meet reserve needs, but the deposit and withdrawal of such balances do not affect the total supply of reserves, although they may result in shifts of available reserves among banks and stimulate expansion or contraction by banks gaining or losing such deposits. In essence, such balances reflect the transfer of the task of lending or investing from the depositing banks to banks holding the balance. Interbank balances are in effect merely an intra-system operation with no direct or separate monetary effect; they are part of the process of putting money to use and do not need to be subjected to extra reserve requirements. The present system of permitting the deduction of balances due from banks in computing net demand deposits subject to reserve requirements eliminates duplication of requirements. Banks obtaining the funds for lending or investing must supply the reserves required to be held against them.

The effect of such balances on the total supply of reserves for member banks, therefore, is negligible, although changes in the amount and distribution of such balances among banks of different reserve classes affect the total volume of required reserves. There does not appear to be any basis in principle for imposing higher reserve requirements on interbank balances or on banks that hold such balances than on other deposits or other banks. An exception to this conclusion might apply in the case of balances due to nonmember banks, which can serve as the basis for multiple credit expansion.

DEMAND AND TIME DEPOSITS

Under the Federal Reserve Act, requirements against time deposits have been lower than those against demand deposits, although there was no difference

under the National Bank Act and some countries having reserve requirements, e.g., Canada, make no distinction. The basis for this distinction from the standpoint of monetary regulation is that time deposits are presumed to represent savings which are entrusted to banks for investment, while holders of demand deposits generally use them more actively in financing the flow of goods and services. In effect, such differentials in requirements give recognition to the function of money as a medium of exchange, as well as to the store-of-value function. The rates of turnover of demand deposits, as measured by debits against them, average between 12 and 30 times a year at most banks, with a number of the largest banks showing turnover rates of over 30. Scattered information available indicates that withdrawals of savings deposits at commercial banks have in recent years averaged about once every 2 years or one-half times a year, although there is some variation among banks, and at times in the past the average rate has been higher. Information is not available as to with

drawals of other types of time deposits at commercial banks.

While it is evident that true savings invested through financial intermediaries, including banks, do not make the same contribution toward effecting payments as do more actively used types of money, the line of distinction between savings and current money is not a clear and narrow one. The gradations are not always distinctly definable. A large portion of time deposits represent liquid funds that may be drawn upon frequently, enabling holders of such readily available deposits in banks to keep their demand balances at a minimum. Moreover, some demand deposits are relatively inactive and might be viewed as savings.

If the margin of difference between requirements against demand and time deposits is too wide, commercial banks, holding both demand and time deposits and using both in the process of credit creation, have a greater inducement to attract funds into time deposits through payment of higher rates of interest or by other means. Many of these funds might otherwise have been held in demand accounts, and the result of the shift would be a lower level of reserve requirements, thus permitting additional credit and monetary expansion, with little or no significant change in owners' attitudes as to their money holdings or in their actions with respect to the use of money. Such a tendency developed in the 1920's, when time deposits had fewer restrictions on withdrawals and showed somewhat higher rates of turnover than has been the case in subsequent periods. Subsequent changes in legislation and regulations placed more restrictions on time deposit withdrawals. Since the end of 1956, however, when banks have been in a position to pay attractive interest rates on time deposits, there has been a considerable growth in time deposits at commercial banks, re flecting some shifting from demand accounts. There are indications that a significant portion of the funds moving into time deposits represents liquid holdings that might readily be shifted into other uses, rather than the accumulation of savings.

Whereas lower reserve requirements against time deposits than against demand deposits are appropriate because of differences in monetary significance, some reserves should be held against time deposits at banks. Such deposits, payable at face value on short notice, perform the store-of-value function of money and, like demand deposits, they play a role in the process of bank credit expansion. Any institution handling such funds runs the risk of fluctuations in values of the assets in which they are invested and should maintain a minimum margin of cash for liquidity or safety. Commercial banks should not be exempted from this requirement or be permitted to maintain a lower degree of liquidity than other savings institutions may find necessary, nor should they be permitted to depend on reserves or liquidity held against demand deposits to cover their time deposit needs. Available information indicates that the ratios of withdrawals to outstanding balances in deposits at mutual savings banks and in share accounts at savings and loan associations are somewhat lower than corresponding ratios against savings deposits at commercial banks. These institutions maintain some cash balances which serve the function of reserves.

For these various reasons, although some differential in requirements between deposits of different types is reasonable and desirable, the present statutory limits of 3 to 6 percent for reserve requirements against time and savings deposits at member banks should not be lowered or narrowed. It is essential that the Board retain at least this much flexibility in order to adjust to changing forces that may affect the distribution and use of such deposits.

CLASSES OF BANKS

Differentials in requirements by classes of banks, although perhaps not specifically established for such a purpose, also in practice conform roughly to differences in the rate of use of deposit money. These differentials by class of bank, as well as those by type of deposit, imply some principle of distinction as to gradations in the money quality of deposits.

Classification of cities for reserve purposes under the Federal Reserve Act has continued to be based upon the system followed under the National Bank Act, which was related to holdings of interbank deposits, even though balances with correspondent banks no longer count as reserves. Banks in the designated central Reserve and Reserve cities are required to maintain higher reserves against all their demand deposits, not only against interbank balances, than are banks in other cities. As explained in chapter 1 and appendix B, criteria used by the Board in classifying cities for reserve purposes, particularly since 1947, have been based primarily upon relative holdings of interbank balances.

In practice, this basis of classification does not work out completely illogically because banks with large amounts of interbank deposits generally also hold relatively active nonbank accounts. The higher requirements, however, apply to all banks in the city, regardless of nature of business, although under existing law the Board may permit banks in outlying sections of a central Reserve or Reserve city to carry reserves at the lower requirement level specified for Reserve city or country banks. These exceptions are granted on the basis of differences in character of business of the individual banks, as well as in location. Nevertheless, many individual banks classified as central Reserve or Reserve city banks hold relatively small amounts of balances due to other banks.

Table 6, which shows the distribution of member banks in each reserve class by amount of total deposits and by amount of demand deposits due to other banks, indicates that in general banks with the largest amount of interbank deposits are now classified as central Reserve city banks and that nearly all other banks with interbank demand deposits of over $10 million are classified as Reserve city banks. There are a few country banks that hold a significant volume of interbank deposits, although the bulk of country banks have little or none. About a fourth of the central Reserve city banks and over half of the Reserve city banks have relatively small amounts of interbank deposits, and most of these are also relatively small in size as measured by total deposits. To some extent these are banks located in the central portion of cities and cannot under present law be permitted to carry lower reserves, and some of them may engage in other types of business which are similar to and directly competitive with banks carrying substantial amounts of interbank deposits. To a small extent they are banks in cities that have chosen to continue as Reserve city banks although they could be given a lower classification under the present regulation. It may be concluded that even tested by the distribution of interbank balances, which is the present underlying basis of classification, the existing classification of banks is to some degree haphazard and unsatisfactory.

TABLE 6.-Distribution of member banks by reserve classes, by volume of total deposits and by volume of interbank deposits, June 30, 1956

To provide a broader basis for classification of banks for reserve purposes, consideration needs to be given to other aspects of the banks' activities than merely interbank deposits. It may be said that in a general way deposits of active commercial, industrial, and financial businesses contribute more to economic activity than do deposits in personal accounts and those of smaller less active enterprises. Differentials of this nature have a bearing upon the effects of monetary policies designed to influence the level of demands for goods and services. It may be considered appropriate, therefore, that banks the bulk of whose deposits consists of more active balances should be required to carry larger reserves than banks whose deposits consist largely of less active balances. Measurement of such differences presents difficult problems. For the country as a whole the overall rate of use of money is broadly indicated by the relationship between the volume of money outstanding in any period to the total volume of transactions in the exchange of goods and services. For individual banks or separate deposit accounts, the only feasible method of measuring activity of deposits is from records of checks drawn against outstanding accounts, but many checks represent transfers of funds that do not involve exchanges of goods or services or other significant economic transactions. The proportion of such transfers, moreover, varies considerably among different accounts. Hence debits against deposits accounts do not always accurately measure significant variations in the money quality of different deposits. Allowance has to be made for other characteristics of the accounts, such as the nature of business of the depositors.

Activity as measured by ratios of checks debited to average balances outstanding, nevertheless, do provide some indication of fundamental differences in the nature of deposit holdings of individual banks. Table 7 shows the distribution of over 1,000 of the largest member banks by reserve classes, by amount of total deposits, and by annual rate of turnover of deposits of individuals, partnerships, corporations, and State and local governments (i.e., all demand deposits except those of banks and the U.S. Government). The figures relate to the situation as of June 30, 1956, and cover nearly all banks for which debit figures are available.