Variable Reserve Requirements Against Commercial Bank Deposits

FIXED LEGAL OR CUSTOMARY RESERVES against deposits have long been employed for the purpose of assuring the liquidity or solvency of commercial banks. These reserves, of course, limit the availability of bank credit and thus have important implications for monetary policy. As long as the reserve ratios are not subject to change by the monetary authorities, however, reserve requirements cannot be actively used as an instrument for carrying out a stabilizing monetary policy flexibly adjusted to changing conditions. In the past two decades, and particularly in the postwar period, the potentialities of variable reserve requirements as an instrument of monetary policy have come to be widely recognized, and many countries have given the monetary authorities power to vary reserve requirements against commercial bank deposits.


FIXED LEGAL OR CUSTOMARY RESERVES against deposits have long been employed for the purpose of assuring the liquidity or solvency of commercial banks. These reserves, of course, limit the availability of bank credit and thus have important implications for monetary policy. As long as the reserve ratios are not subject to change by the monetary authorities, however, reserve requirements cannot be actively used as an instrument for carrying out a stabilizing monetary policy flexibly adjusted to changing conditions. In the past two decades, and particularly in the postwar period, the potentialities of variable reserve requirements as an instrument of monetary policy have come to be widely recognized, and many countries have given the monetary authorities power to vary reserve requirements against commercial bank deposits.

FIXED LEGAL OR CUSTOMARY RESERVES against deposits have long been employed for the purpose of assuring the liquidity or solvency of commercial banks. These reserves, of course, limit the availability of bank credit and thus have important implications for monetary policy. As long as the reserve ratios are not subject to change by the monetary authorities, however, reserve requirements cannot be actively used as an instrument for carrying out a stabilizing monetary policy flexibly adjusted to changing conditions. In the past two decades, and particularly in the postwar period, the potentialities of variable reserve requirements as an instrument of monetary policy have come to be widely recognized, and many countries have given the monetary authorities power to vary reserve requirements against commercial bank deposits.

The purposes of the present study are (1) to examine the functions of variable reserve requirements; (2) to consider objections to variable reserve requirements and limitations on their proper use; (3) to describe the conditions in which reserve requirements have been, or may be, changed; and (4) to examine technical questions relating to the form and operation of variable reserve requirements. The status of the reserve requirements in various countries is summarized in Appendix II.

Functions of Variable Reserve Requirements

Limitation of the quantity of bank credit

Reserve requirements determine the maximum amount of loans and investments that each commercial bank, and the banking system as a whole, may maintain in relation to deposits. Thus, if the reserve requirement is 20 per cent of deposits, loans and investments (of the bank’s own choosing) may not exceed 80 per cent of deposits.1

Reserve requirements therefore limit the total expansion of bank credit and the secondary expansion of bank deposits that can occur on the basis of any primary increase in deposits.2 Reserve requirements also have the effect of limiting the reduction in bank credit and deposits that is forced upon the banking system by a primary decrease in deposits. Looked at from another point of view, it can be seen that reserve requirements limit the extent to which bank loans and investments can, or must, change when there is an increase or decrease in the cash resources of banks as a result of rediscounts at the central bank or a change in another nondeposit item in the balance sheet. This is true because the extension or contraction of bank loans and investments results in the creation or extinction of deposits and hence in an increase or decrease in the amount of required reserves.

These functions can be performed by several types of reserve requirement, but they are clearest in connection with a cash reserve requirement that the commercial banks discharge by maintaining deposits at the central bank. Attention will first be given to this type of requirement, which is not only the simplest but also the most common.

Primary changes in bank deposits come about most often as the result of a balance of payments surplus or deficit, or the extension or contraction of central bank loans or investments. These phenomena create or destroy “reserve money”—which consists of currency and deposits at the central bank and which has the capacity to satisfy the cash reserve requirements of the commercial banks. The government can influence the quantity of reserve money by shifting its deposits between the commercial banks and the central bank. In some countries the government can also increase or decrease the amount of reserve money by issuing or withdrawing treasury currency, but at the present time in most countries the amount of treasury currency is small compared with central bank currency. Bank deposits may change, without any change in the amount of reserve money, if the public decides to hold more or less deposits relative to currency.

In addition to legal reserves, a second important limitation on the capacity of the banking system to expand on the basis of a primary increase in deposits arises from withdrawals of currency by bank customers. Ordinarily, the public wishes to maintain a fairly constant relation between its deposits and currency holdings. When deposits increase, the demand for currency also increases. This may be called the currency drain. The commercial banks can obtain currency to pay out to customers only by drawing down their reserve deposits at the central bank or by using till money. Till money is the currency that banks keep on hand to satisfy day-to-day needs. In some countries till money is included, along with deposits at the central bank, in legal reserves, but in other countries its provision places a third limitation on the capacity of banks to expand loans and investments. Till money, however, is usually small relative to deposits and it will not be further considered in the following exposition. Another limitation on the growth of money supply that is highly important in many countries is the “external drain.” This is the withdrawal of money from circulation that occurs when a balance of payments deficit is caused by a rise in imports and a fall in exports resulting from an expansion of credit and money income.3 In this paper it is considered convenient to analyze the effects of reserve requirements without allowance for the external drain, since the control of the balance of payments (and hence indirectly of the external drain) is usually an objective of monetary policy and can be furthered by measures that influence the quantity of reserve money and the secondary credit expansion that can be based on it.

Since bank deposits are a large part of the money supply in virtually all countries, reserve requirements have an important influence on the extent to which the money supply can expand as a result of a balance of payments surplus, a central-bank-financed fiscal deficit, or any other development that brings about an increase in reserve money. The higher the required reserve, the smaller the maximum expansion. A high required reserve likewise diminishes the extent to which the money supply must contract as a result of a reduction in reserve money. A high ratio of currency to total money supply also restricts the capacity of the banking system to expand loans and investments and thus to increase the money supply on the basis of an injection of reserve money or an increase in the banks’ cash resources through rediscounting with the central bank. These relations are set forth in a formal way in Appendix I. Table 1 shows how the size of the potential expansion of money supply in response to an injection of reserve money (expressed as a multiple of the increase of reserve money) is determined both by the magnitude of the minimum required reserve ratio and by the proportion of the actual increase in money supply that takes the form of currency.4 It will be observed that, whenever either marginal reserve requirements or the ratio of currency to money supply is 100 per cent, no secondary expansion of the money supply is possible.

Table 1
article image

The authorities have little if any control over the proportion of cash balances that the public chooses to hold in the form of currency, but in many countries they can control the banks’ reserve ratio. By imposing a high required reserve ratio, the authorities can limit the inflationary effect of a government deficit, a balance of payments surplus, or any other development that brings about an increase in reserve money. The existence of a high reserve ratio can cushion the shock of a reduction of reserve money, because a large fraction of the loss will be reflected in a decrease in required reserves rather than in bank credit, and the contraction of money supply will be smaller than it would be with a low reserve ratio. A high fixed reserve requirement, therefore, can be an important stabilizing force by virtue of the fact that it reduces the secondary credit expansion or contraction that occurs on the basis of a primary change in reserve money. A fixed reserve requirement, it should be emphasized, does not prevent variations in the money supply; it merely decreases the extent of variation by lessening secondary changes in bank credit. Thus, for example, a fiscal deficit covered by borrowing from the central bank would cause an expansion in money supply even if the required reserve ratio for commercial bank deposits were 100 per cent (at the margin), but in this case there would be no secondary expansion.

By changing reserve requirements, the authorities may be able to take action that will not only lessen the secondary monetary effect of a domestic or external development but will partly, or wholly, offset the primary change in reserve money. If, for example, reserve requirements are raised at a time when there is a balance of payments surplus, the contraction of domestic credit forced on the banks may offset the expansionary effect of the purchase of foreign exchange by the central bank.

Changes in reserve requirements, however, cannot assure monetary stability unless they are accompanied by other appropriate actions. For example, it may not be feasible to stop an inflation by raising reserve requirements if the authorities continue to feed large quantities of reserve money into the banking system by extending central bank credit to the government to cover a fiscal deficit. In extreme circumstances, the amount of central bank lending to the government may be so great that the inflationary effect cannot be fully offset by any tolerable reduction in commercial bank lending to the private sector.

On the other hand, a reduction in reserve requirements permits but does not assure an expansion of bank credit and money supply. If the demand for funds on the part of qualified borrowers is weak, the banks may see few attractive lending opportunities and may hold excess reserves. Hence a cut in the required reserve ratio may do little to counteract incipient deflation unless it is accompanied by other measures.

Changes in required reserves have less effect in the underdeveloped countries than in the more industrialized countries. The currency drain in underdeveloped countries is heavy, and it limits the size of the secondary expansion or contraction that is permitted or forced by a primary change in reserve money. Reserve requirements, nevertheless, may be especially useful in financially less advanced countries because open market operations are not possible there and because changes in the central bank’s rediscount rate are often less influential than in countries with more highly developed financial institutions.

Reserves in forms other than cash deposits with the central bank or cash held in the vaults of the commercial banks can also serve the purpose of limiting the expansion of bank credit and the money supply. If the reserves take the form of government bonds or other securities that must be obtained from the central bank, their credit-limiting function is identical with that of cash reserves. In Mexico, for example, an important element in the restrictive monetary policy of the period following the 1954 devaluation was the application of high marginal reserve requirements which were met in part by the banks through the purchase of government securities from the Bank of Mexico. The principal difference between an ordinary cash reserve requirement and one that can be met by buying securities from the central bank is that the latter allows the commercial banks to earn interest on their required reserves.

Any reserve requirement may limit credit expansion if the available quantity of the asset that is eligible to serve as a reserve is fixed or is limited in quantity by the authorities. Reserves in forms other than cash, however, often have as their primary purpose the influencing of the allocation of credit among different uses.

Allocation of credit

Although other forms of reserve requirement involve a more detailed control over the allocation of credit among uses, even a cash reserve requirement may have an important influence on credit allocation. When a monetary expansion occurs, the increase in commercial banks’ reserve deposits and in currency in circulation constitutes an increase in the central bank’s liabilities and is reflected in a corresponding increase in the central bank’s assets. Given the currency drain, reserve requirements determine the proportions in which the increase in monetary liabilities (and in bank loans and investments) will be divided between the central bank and the commercial banks. A high reserve requirement means that, within the limits of any specified monetary expansion, the central bank can increase its assets more, and commercial banks can increase their loans and investments less, than would be possible with a lower reserve requirement. In most countries, this will tend to divert credit from the private sector to the government sector since the central bank’s domestic assets usually include a large proportion of loans to the government and government securities. When a country has a balance of payments surplus, the growth of the central bank’s foreign assets will account for part of the credit expansion.

The general relationship between reserve requirements and currency drain and the allocation of credit is examined in Appendix I. This relationship is illustrated in Table 2, which shows the proportion (expressed as a percentage) of any monetary expansion that is reflected in the liabilities and assets of the central bank; these are shown for the same ranges of reserve requirements and currency drain as were used above to show the possible variations in the expansion of money supply in response to an injection of reserve money. The remainder of the expansion will be reflected in an increase in commercial bank loans and investments.

Table 2
article image

A more explicit and detailed influence on the allocation of bank credit is exerted when reserve requirements can be met by certain types of investment or loan rather than by deposits in the central bank. The most usual provision is to allow commercial banks the option of meeting part, or all, of a legal reserve requirement in cash or assets such as government securities, certain private securities, and loans. Unless risks or inconveniences that bankers consider disproportionate to the return are attached to the loans and investments that may be used to meet reserve requirements, the substitution privilege is likely to be fully used. It is clearly advantageous to the banks to make the substitution, since a return can be earned on the alternative items whereas central banks usually pay no interest on reserve deposits.

The noncash requirements under consideration here are more an instrument of selective credit control and allocation than a means of limiting total bank credit. The banks can continue to expand credit as long as they acquire the specified assets. The amount of credit available for the government or other borrowers designated by the authorities is increased and interest rates for these borrowers are held down; the amount of bank credit available to other borrowers is limited and interest rates for these borrowers may be raised.

Some limitation on total bank credit may be achieved, however, through the selective control and reallocation of credit. The favored borrowers are the government and others that are judged to merit high social priority. These borrowers are often able to obtain all or a large part of their credit requirements even in the absence of a special reserve scheme. By diverting credit to the high-priority borrowers, the reserve scheme may enable the authorities to limit the total expansion.

A special reserve scheme that was adopted in a few countries after World War II was intended to prevent the commercial banks from disposing of their abnormally large holdings of government securities, or at least to delay the reduction of these holdings. This was the origin of the secondary reserve requirements in Belgium, Italy, and France. These plans had some success in keeping down interest costs of the government debt while minimizing the volume of central bank purchases needed to support government security prices. With the passage of time, however, undesirable consequences became more noticeable in some countries. The Belgian secondary reserve scheme was substantially modified in 1957 (see page 35).

Supplement or substitute for other instruments

A considerable degree of stability of the cash or liquidity ratio of commercial banks is a prerequisite to effective control of bank credit and money supply. If the banks were prepared to reduce their reserve ratio whenever the monetary authorities acted to reduce the quantity of reserve money or to restrict its growth, the actions of the authorities might be nullified. In some countries banking tradition is strong enough to ensure that commercial banks will observe customary reserve ratios, but in other countries it has been found advisable to establish a legal reserve requirement. A firm customary reserve ratio or a legal reserve requirement provides a fulcrum for the leverage exerted by changes in the volume of rediscounts and open market operations.

When the monetary authorities have the power to vary reserve requirements, changes in requirements can be used to supplement other instruments of credit control in order to influence credit conditions more quickly and to a greater extent than would otherwise be possible. In countries in which the traditional instruments are not available, variable reserve requirements may serve as a partial substitute for them.

An increase in reserve requirements, like open market sales of securities by the central bank, may be employed as a means of making the central bank’s rediscount rate effective. In the absence of supplementary action, an increase in the rediscount rate may have little effect on the availability of credit or its cost to nonbank borrowers. This may be the case if the commercial banks have excess reserves or are not accustomed to rediscounting at the central bank; but even where these conditions do not exist, there may be considerable delay in transmitting the effect of a change in the rediscount rate to the operations of the commercial banks. An increase in reserve requirements, however, can force the commercial banks to borrow from the central bank if they wish to expand their loans—and may force the more extended commercial banks to borrow immediately in order to meet the higher reserve requirement. A high rediscount rate can thus be made effective at once and may raise interest rates and limit the availability of credit. In other circumstances the authorities may wish to ease credit. A reduction in required reserves increases the liquidity of the banks at once and may stimulate them to expand loans and investments, provided that required reserves are still sufficient to meet customary standards of solvency. A reduction in the rediscount rate may not have the same influence if the commercial banks consider borrowing from the central bank appropriate only for emergencies, and therefore in ordinary circumstances a reflection on their soundness.

Changes in reserve requirements can also act as a supplement to open market operations. An illustration of this is provided by a case in which the commercial banks have large excess reserves, which may be the result of a balance of payments surplus, a capital inflow, or a government deficit. Although it may be possible to eliminate these excess reserves and to bring about a contraction of credit through open market operations, it is likely to take a considerable period of time to achieve this. In very few countries are the financial markets so large and so well organized that they can absorb a large volume of sales of government securities and other paper by the central bank within a short time. An effort to sell an unusually large amount of securities from the central bank’s portfolio may in these circumstances demoralize the market and produce unnecessary confusion and capital losses. In many cases, moreover, the central bank will not have enough securities in its portfolio to allow it to conduct open market operations on the scale that would be required to eliminate excess liquidity on the part of the commercial banks and the general public. If the excess liquidity of the commercial banks is first substantially reduced, or eliminated, by an increase in required reserves, open market sales of securities in modest amounts may bring about a prompt contraction of bank credit. Obstacles to open market purchases of securities by the central bank, when the authorities wish to ease credit, may be less serious but may delay or limit action.

Open market operations are not feasible, even on a limited scale, in most countries, because there is no active securities market in which the central bank can buy and sell. In these countries variable reserve requirements may be especially attractive because they are the only instrument by which the authorities can change bank liquidity entirely on their own initiative. A change in the rediscount rate or rediscount ceilings is, of course, initiated by the central bank but depends for its effectiveness on the reactions of the commercial banks. Where there is no securities market and where the commercial banks are traditionally reluctant to rediscount their loans at the central bank, variable reserve requirements may be the principal instrument available to the authorities. In many such countries, however, as has been indicated above, the potential effectiveness of changes in reserve requirements is less than in the financially developed countries because the currency drain is large.

Objections to Variable Reserve Requirements

Although reserve requirements can be an important instrument of monetary policy and many countries have given the authorities the power to impose and vary these requirements, this power has been used in most countries to only a limited extent. In a number of countries the required reserve ratio has never been changed; in certain other countries changes in the ratio have been comparatively frequent, but in some of these countries there has been considerable opposition to variations in reserve requirements.5

The principal objections made to variable reserve requirements are that (1) reserve requirements are a form of direct control and hence are subject to arbitrary use; (2) an increase in required reserves that works a hardship on some banks may have little effect on banks that are in a more liquid position when the increase is adopted; (3) the imposition of reserve requirements on commercial banks discriminates against them, in contrast to other banks and nonbank financial intermediaries that are either free of reserve requirements or subject to less stringent requirements; (4) high reserve requirements impair bank earnings; (5) high reserve requirements may result in channeling an excessive proportion of total bank credit to the government sector; and (6) reserve requirements involve administrative problems. These objections are more fully stated below, and some counterarguments are mentioned as well as measures that can be taken to avoid harmful effects of reserve requirements.

Reserve requirements as a form of direct control

Those who object to reserve requirements because they are considered a form of direct control usually stress the advantages of the price system as a means of directing economic activity and allocating credit. Open market operations and changes in rediscount rates, in contrast to reserve requirements, depend for their effectiveness on the voluntary response of commercial banks to economic incentives arising from the actions of the central bank. These instruments are therefore less susceptible of being employed in a way that will work a hardship on individual banks. Granted that this point carries weight, it still may be argued that general reserve requirements involve less detailed and arbitrary regulation of the affairs of the banks and their customers than such actions as the establishment of credit ceilings and the control of borrowing, which may be the only alternative measures available to the authorities. Furthermore, a stabilization policy carried out through the regulation of the volume of bank credit—even when obtained by application of direct controls to banks—involves much less detailed interference with private economic activity than attempts to suppress inflation by price and wage controls.

Differential impact of changes in requirements

Perhaps the most influential objection to changes in reserve requirements is that different banks are affected in quite different ways. An increase in requirements will be much harsher on banks with low liquidity ratios than on banks with high ratios and indeed may have no immediate effect on some banks. This objection is a specific form of the general objection to direct controls.

The authorities will always wish to take into account the positions of individual banks when an increase in reserve requirements is contemplated. But it is not necessary to hold the increase to the amount that will just absorb the excess liquidity of the least liquid commercial bank; if that were so, the usefulness of the instrument would be very limited. It is quite possible for the central bank to help commercial banks to meet an increased reserve requirement by rediscounting loans or buying securities from them. To be sure, the banks may suffer losses if the current rediscount rate is higher than the yield on their outstanding loans and investments. This is regrettable but it does not seem to be an adequate reason for refraining from an increase in reserve requirements which is indicated on other grounds. Certain alternative policies may be even less desirable from the point of view of private bankers. Furthermore, practically all monetary or fiscal actions inconvenience some persons or firms more than others. For example, tightening of credit is always more inconvenient to necessitous borrowers than to those who have less urgent need for funds, and an increase in interest rates may cause heavy capital losses for holders of long-term securities.

A more important point is that liberal rediscounting by the central bank can nullify the contractionary effect of higher reserve requirements. It is not necessary, however, that the whole effect of higher reserve requirements be offset by an increase in central bank credit; some liquid resources of the banks can nearly always be absorbed immediately. The central bank, moreover, will be able to exert continuous pressure on the commercial banks to reduce their borrowing. Thus over time the higher reserves can be used to bring about a decrease in the outstanding volume of bank credit. A central bank that is too weak to combine rediscount policy and reserve requirements in this way is not likely to be strong enough to carry out another program for limiting credit.

The differential impact of an increase in reserve requirements can also be softened by announcing in advance the increase, perhaps to take effect in several stages. The time interval gives the commercial banks an opportunity to accumulate cash by liquidating loans and investments or by allowing them to run off. It is difficult, however, to foresee exactly what requirements will be appropriate over an extended period of time. The technique of advance announcement is, therefore, more suitable for putting into effect a new system than for adjusting to changes in the economic situation.

A marginal reserve requirement—that is, a requirement related to increases in deposits after a certain date rather than to total deposits—also avoids embarrassing banks that are illiquid at the time the requirement is raised. Marginal requirements, however, discriminate against growing banks and against banks whose deposits were unusually small for any special reason at the base date (see pages 39-40).

Discrimination due to incomplete coverage

In several countries the authorities seem to have hesitated to place great reliance on reserve requirements because the requirements cannot be applied to all commercial banks or cannot be applied to nonbank financial institutions that are considered active competitors of the commercial banks. The United States offers the clearest example of the first limitation: the Federal Reserve Board has authority to apply reserve requirements only to members of the Federal Reserve System. Although all the large commercial banks are members, many smaller banks are not members. The Board has been reluctant to penalize membership in the System by raising reserve requirements so high that a competitive advantage is given to nonmember banks. In several countries the fact that the reserve requirements do not apply to the banking department of the central bank has aroused the resentment of commercial bankers. Although considerations of this nature have great practical importance, they do not arise in acute form in most countries and they do not suggest any inherent limitation on the successful use of reserve requirements. If political agreement can be obtained, the requirements can be extended to cover all commercial banks, conventionally defined.6

A more difficult question of principle arises in connection with building societies, savings banks, and other financial institutions, which are often fairly direct competitors of the commercial banks in certain activities but which differ from commercial banks in important respects. The fact that these institutions are not subject to reserve requirements may cause the authorities in some instances to hesitate to raise requirements for commercial banks. Some further comments on this problem are made below (pages 27-29).

Adverse effect on bank earnings

High reserve requirements reduce bank earnings unless provision is made to allow a return on the reserves. It is true that the restriction of the supply of credit brought about by reserve requirements may cause a rise in the interest rates earned by banks on their loans and investments, and this conceivably could fully offset the effect on bank profits of a reduction in the volume of lending; in view of the stickiness of bank lending rates and their conventional nature in some countries, however, the offset seems likely to be minor in most cases. Moreover, other measures of credit restraint may also reduce banks’ gross profits.7 This is clearly the effect of credit ceilings. An increase in the central bank’s rediscount rate may curtail opportunities for profitable relending of funds obtained by rediscounting. Open market sales of securities by the central bank may cause a contraction in commercial banks’ deposits and earning assets (when the sales are made to the nonbank public), or they may involve the substitution of comparatively low-yield government securities for higher-yield commercial loans (when the sales are made directly to the banks). Changes in the rediscount rate and open market operations, nevertheless, are generally more favorable to bank profits than variations in reserve requirements or the imposition of credit ceilings; and they leave more freedom of choice to individual banks.

Banks, understandably, dislike measures that obviously limit their profits. The matter is also one of general public concern in view of the socially important services of the banks in providing a convenient and safe means of payment and in acting as an intermediary between depositors and borrowers. The question whether bank earnings are adequate can be decided only in the light of an investigation of relevant facts at the time and place. In principle, it is quite possible to separate the question of adequacy of bank earnings from that of the appropriate level of required reserves. The central bank may pay interest on reserve deposits or allow the requirements to be met partly or wholly by interest-earning government securities or central bank securities. The interest rate on the deposits or securities can be adjusted as required. Admittedly, however, this approach may jeopardize the independence of the commercial banks and may introduce additional political issues into a field that is already highly controversial in many countries.

Excessive credit to the government

As explained above, high reserve requirements may have the effect of diverting credit to the government sector, either because the requirements can be met by holding government securities or because the central bank is enabled to grant larger credits to the government without bringing on inflation. This may be considered objectionable on the grounds that it allows the government to increase its expenditures at too rapid a rate or because it is thought that too little credit is left for the private sector.

The argument that high reserve requirements may allow government expenditures to increase too rapidly depends on the assumption that government spending is in fact limited by the availability of finance and that in the absence of the reserve system the government would be unable or unwilling to find money by inflationary borrowing from the central bank or by raising taxes. The availability of noninflationary finance is a factor that nearly always influences the amount of government expenditures, but the strength of this influence varies greatly from country to country and with the urgency of the government’s needs. Public discussion in Belgium, for example, suggests that in recent years financial limitations may have been an important check on government spending in that country and that the existence of a reserve requirement that could be met by holding treasury bills allowed larger government expenditures than would otherwise have been possible. In many cases, however, it must be assumed that government expenditures will increase regardless of whether conventional means of finance are available. In these circumstances, higher reserve requirements may offer an opportunity to finance in a noninflationary way additional government expenditures that would occur in any event.

The quantitative importance of this point may be illustrated by reference to Table 2, and to the formulas upon which the table is based (Appendix I). If there is no change in the central bank’s foreign assets or other investments and loans, the bank can make available to the government an amount of credit equal to the increase in central bank liabilities (consisting of currency and bankers’ deposits). As Table 2 shows, the increase in the central bank’s liabilities will be a larger proportion of the increase in credit and money supply when reserve requirements are high than when reserve requirements are low. Suppose, for example, that the marginal ratio of currency to money is 25 per cent and that the authorities consider it safe to allow a 5 per cent increase in money supply. With a marginal reserve ratio of 10 per cent, the maximum amount of credit that the central bank can give to the government without causing inflation is a sum equal to 1.6 per cent of the money supply, leaving 3.4 per cent for the private sector. By raising the marginal reserve ratio to 50 per cent, the authorities can almost double the amount of noninflationary finance that the central bank can grant to the government—at the expense, of course, of a corresponding decrease in credit for the private sector.8

A drastic reduction in credit for the private sector may be the only way of providing noninflationary finance for the government in an emergency. In such circumstances, a temporary increase in bank reserve requirements, even to a high level, may be an appropriate measure. The maintenance of high reserve requirements over a period of years, however, in order to make possible continued noninflationary central bank financing for the government, is much more questionable. This policy would severely limit the capacity of the commercial banks to provide credit for the private sector and might interfere with the growth and efficient operation of the economy.

Administrative problems

The simpler reserve requirement plans ordinarily present no great administrative difficulties, but some of the more elaborate systems involve complications for the supervisory authorities and the commercial banks. Even the simplest system requires some reporting and supervision to assure compliance. This is not needed in connection with open market operations and may not be necessary if reliance is placed on simple changes in rediscount rates. The administration of reserve requirements, however, is likely to be no more complicated—and often less complicated—than the operation of credit ceilings, progressive discount rates, and other credit-rationing devices.

Changes in Reserve Requirements

Occasions for changes in requirements

There is fairly general agreement that changes in reserve requirements may be an appropriate means of compensating for major changes in the volume of reserve money which would otherwise lead to large fluctuations in the supply of bank credit and the quantity of money. Large and sudden movements in a country’s gold and foreign exchange assets, especially those due to speculative international capital movements, have disturbing effects on both the country receiving the inflow and the country suffering the outflow. In the receiving country the central bank may be unable or unwilling to engage in open market operations on the scale that would be needed to offset the monetary effect of the inflow and may attempt to deal with the problem by raising bank reserve requirements. Thus, in 1936-37 the Federal Reserve authorities in the United States raised the required reserve ratio, in three steps, until it was double what it had been before, mainly to counteract the monetary effect of a large inflow of gold. In 1957 the Federal Republic of Germany imposed higher reserve requirements for bank deposits owned by foreigners, in order to reduce the inflationary effect of the capital inflow that the country was then experiencing. Countries suffering a capital flight are likely to be reluctant to take compensatory monetary action because the decrease in domestic liquidity associated with the gold outflow may help check the capital flight. If, however, compensatory action is considered appropriate, a reduction in the required reserve ratio is an obvious measure.

There is perhaps less general agreement, but still no great dispute, about the propriety of altering reserve requirements to help compensate for major changes in credit demand and supply attributable to factors such as war or major cyclical variations in business activity.

Less support can be found for a policy of varying reserve requirements to compensate for smaller or more temporary changes in the quantity of reserve money due to fiscal deficits or surpluses, changes in the balance of payments, minor cyclical or ordinary seasonal variations in domestic business activity, flotation or retirement of government loans, etc. Two separate questions are involved. One is whether any kind of compensatory monetary action is appropriate. The answer to this question depends not only on the magnitude of the fluctuation but also on the country’s general economic policy, the adequacy of its foreign exchange reserves, and other considerations which cannot be discussed here. The other question is whether a change in reserve requirements is the most suitable instrument for bringing about any monetary action that is considered desirable. The answer to this question turns mainly on how serious the objections to variable reserve requirements discussed in the preceding section are considered and on the availability of other compensatory instruments. Most countries have not altered reserve requirements to deal with minor or temporary situations, presumably because it has been felt that rediscount operations and open market operations are sufficient to achieve any essential compensatory action and perhaps because the traditional instruments have been considered more acceptable to the commercial banks and hence more conducive to cordial relations between the central bank and the financial community.

Some countries have, however, changed reserve requirements to deal with rather minor fluctuations as well as major variations in the monetary environment. Australia and New Zealand, for example, have altered reserve requirements frequently to deal with changing balance of payments conditions. Ecuador, Egypt, Greece, and New Zealand have changed reserve requirements to meet seasonal variations in the demand for credit; in March 1957, however, the Reserve Bank of New Zealand announced that in future it would try to avoid “substantial and frequent” changes for this purpose. The Netherlands has altered reserve requirements on occasion to compensate for the monetary effect of changes in the balance of payments and also to facilitate the flotation of government loans. In the United States the Federal Reserve authorities have refrained from adjusting reserve requirements in the light of day-to-day or seasonal changes in monetary conditions, but they did reduce requirements during the moderate business recessions of both 1948-49 and 1953-54, as well as during the more severe recession of 1937-38 and again in 1957-58.

Measures accompanying changes in reserve requirements

As mentioned above, the authorities may have to adopt transitional measures that will enable banks with low liquidity ratios to meet an increase in reserve requirements without excessive hardship. Generally, it is a better practice to provide rediscounts on an individual basis than automatically to extend central bank credit to all commercial banks that are short of liquid resources or to allow reserve deficiencies to be incurred. By relying on ordinary rediscount operations, the authorities can bring the positions of individual banks under closer scrutiny and can exert firmer pressure on the banks to reduce their loans. The use of rediscounts sets a time limit on the central bank’s assistance, which can be extended only by a new action by the central bank. If reserve deficiencies are tolerated—even though penalties are imposed—the commercial banks may be able to postpone for a long time compliance with higher reserve requirements.

Rediscounting or other assistance from the central bank may lessen the immediate contractionary effect of an increase in reserve requirements, but if properly limited this policy will not nullify the change in reserve requirements.

Technical Questions Relating to Reserve Requirements

Voluntary versus legal reserves

As long as only a fixed minimum reserve is contemplated, it may be possible to rely on banking custom to assure maintenance of the minimum reserve ratio. The United Kingdom is an outstanding example of a country in which banking traditions are so firmly established that a legal reserve requirement is considered unnecessary.9

The majority of countries, however, have preferred a legal requirement even when an invariable minimum is prescribed. A preference for a legal reserve requirement may reflect the absence of a strong banking tradition, the existence of many small banks rather than a few large banks, or general legal and governmental attitudes. In some countries in which reliance had formerly been placed on customary reserves—for example, Belgium and Germany—banking traditions were impaired during World War II or the early postwar period, and a legal reserve requirement was imposed. Other countries formerly relying on voluntary reserves have seen fit to impose legal reserves in connection with the establishment of a central bank, possibly in recognition of the fact that the existence of this institution increases the effective liquidity of individual banks and thus would permit a reduction in customary reserve ratios if this were not prevented by law or regulation.10

The most that can be expected of a truly voluntary reserve ratio is that it will provide a fulcrum upon which discount rate policy and open market operations can be based. If the authorities wish to supplement other credit controls with changes in the minimum reserve ratio, it is necessary to depart from the purely voluntary principle, since the banks cannot be expected voluntarily to increase the reserve ratio when that may be desirable from the point of view of monetary policy. Most countries have therefore given the monetary authorities power to prescribe and to vary minimum reserve ratios.

In a few countries, including some of those in which legal authority to impose reserve requirements exists, the central banks have merely “suggested” desired reserve ratios to the commercial banks or have entered into “gentlemen’s agreements” with them. This practice has been followed in Austria, Sweden, and Switzerland, for example. It seems to have been preferred by some central banks because it avoids the appearance of detailed control over the private banking system and is more in accord with tradition. The private banks may be willing to accept such “voluntary” reserve ratios because they believe that, if they do not, legally enforceable requirements will be imposed and may prove to be more rigid and less subject to discussion than the “voluntary” arrangements.

Credit institutions subject to reserve requirements

Reserve requirements are most commonly applied to commercial banks, as defined by general banking law. When reserves are required for other financial institutions the intention seems usually to be to assure solvency rather than to carry out general monetary policy. A few countries, however, have imposed regulations on nonbank financial institutions for the purpose of controlling the allocation of their loans and investments and incidentally of influencing the total volume of credit. Lately there has been public discussion in several countries of the advisability of extending reserve requirements to financial institutions that compete most directly with the commercial banks in the sense that their liabilities are close substitutes for bank deposits and their loans and investments resemble those of banks.

Among banking institutions, narrowly defined, the banking department of the central bank is sometimes exempt from reserve requirements, but this exclusion is now less common than formerly. In practice, the exemption has generally led to unfavorable results. Many of these banking departments conduct a business that is directly competitive with that of the private banks, and the latter have regarded exemption of the central bank as grossly unfair. The exemption has been defended, in part, on the ground that the central bank has the national interest in mind in expanding its loans and therefore should not be bound by the restrictions placed on the private banks, which are in business primarily to earn profits. This argument is weakened by the fact that it is often difficult to find any substantial differences between the lending policy of the banking department of the central bank and that of the private banks. In some instances the banking department has been a leading contributor to an excessive expansion of credit.

Small private banks are sometimes excluded from reserve requirements, largely for purposes of administrative convenience, since as a rule they hold only a small percentage of the total assets of the banking system.

Another important class of exempt institutions is composed of specialized banks, such as development banks, cooperative banks, and agricultural banks. The rationale behind the exemption of this group is that these institutions provide specialized forms of credit which are desirable from an economic point of view and which are not obtainable from other sources. Although these institutions may not create monetary deposits and may supply a type of credit that is rather different from that supplied by commercial banks, their operations do increase the total volume of credit and the demand for goods and services. Furthermore, the operations of the specialized institutions lead to an expansion of deposits in commercial banks and thus provide the basis for a secondary expansion of nonspecialized credit. The exemption of the specialized banks may or may not be justifiable on grounds of social policy, but it should be recognized that when such institutions are excluded more severe restraints must be placed on commercial banks, and their customers, than would otherwise be necessary to hold the total amount of credit within acceptable limits.

Where commercial banks are subject to reserve requirements, the requirement usually applies to savings deposits or time deposits as well as to demand deposits, although the minimum ratio is often lower for the former. Savings banks and other institutions that accept only time deposits, however, are often exempt. Presumably the exemption reflects the belief that the activities of these institutions are less likely to be inflationary than those of commercial banks.

Variable reserve requirements are hardly ever applied to savings and loan associations or building societies, hire-purchase companies, or similar nonbank financial intermediaries. From the point of view of savers, a claim on one of the nonbank financial intermediaries is safer and more liquid than a direct claim on the investor who borrows from the intermediary; the borrower, for his part, finds it more convenient and cheaper to obtain funds from the intermediary than to go directly to the savers. Although there are important differences in detail, it will be seen that the activities of the nonbank financial intermediaries closely resemble those of banks in certain significant respects. The comments made above about the specialized banks apply also to the nonbank financial intermediaries. Furthermore, experience in the Union of South Africa, the United Kingdom, the United States, and other countries in recent years indicates that shifts of deposits between banks and nonbank intermediaries and changes in relative rates of growth of deposits can complicate the carrying out of a restrictive credit policy. It is for this reason that there has been discussion of the advisability of applying reserve requirements to certain nonbank financial intermediaries. The implications of proposals for the extension of reserve requirements to nonbank financial institutions, however, will not be further considered in the present paper.11

Bases for computing reserve requirements

There are three general bases for determining bank reserve requirements, namely, deposit liabilities, turnover of deposits, and assets. One of these bases is used in every method for computing reserves.

By far the most widespread method is to relate the required reserve to the deposit liabilities of the individual bank. The theory underlying this method is that deposits form the basis for the expansion of credit and therefore, by requiring a portion of these deposits to be immobilized as legal reserves, the expansion of credit will be controlled at its source. The advantage of this method is its simplicity in the computation of the reserve and the ease with which it can be enforced. A disadvantage is that the method does not distinguish among deposits on the basis of frequency of turnover or the use to which the bank puts the funds at its disposal. A requirement based on deposits, moreover, is less easily adaptable to the objective of selective credit control than a requirement based on earning assets. As shown below, however, it is possible to adapt the conventional type of requirement to the purposes of selective control. Variants of the liability base relate reserve requirements to total liabilities, to the ratio of current liabilities to total liabilities, and to demand deposits and time deposits separately.

A second method of determining the reserve requirement is to relate it to the turnover of deposits; this is sometimes called the velocity reserve.12 This method was formerly employed in Mexico and has recently been under discussion in the United States. The purpose is to apply automatically a brake to an expansion of monetary demand associated with an increasing velocity of circulation of deposits during a period of boom and high interest rates; the brake is automatically eased as the turnover rate falls with a slowing down of business activity and a reduction of interest rates. The method has the further advantage of preventing the avoidance of higher reserve requirements that is possible under a system that differentiates on the basis of type of deposit or bank. Under such a system, required reserves may be minimized by shifting deposits from categories with high required ratios to categories with low ratios—e.g., from demand deposits to time deposits or from deposits in large city banks to country banks—without changing the real character of the deposit. This has been a practical problem of some importance in a few countries. The turnover system allows a reduction in required reserves only when the shifts are actually associated with a change in frequency of use of the deposit.

Several objections to the velocity reserve plan have been raised. The principal theoretical objection arises from the consideration that the rate of deposit turnover is not necessarily related in any simple way to the income velocity of money or to transactions that the monetary authorities wish to control. Differences in deposit turnover between banks reflect partly differences in the nature of the business in which banks specialize: a bank in a financial or commercial center is likely to have a high deposit turnover, but this does not necessarily mean that its operations contribute more to final demand for goods and services than those of an average bank. The average deposit turnover rate may change as a result of changes in payments technique, such as the introduction of clearing houses that materially reduce the amount of payments made by check. To meet such objections, it has been suggested that reserve requirements be related to cyclical changes in the deposit turnover of individual banks or in the average turnover rate of the banking system. This approach, however, would add considerably to the administrative difficulties that are involved in any velocity reserve requirement.

After experimenting with a velocity reserve in the 1930’s, Mexico abandoned it in 1941 largely because of its administrative complexity. The principle underlying the turnover method, however, provides a partial rationalization for several other types of differentiation now employed in certain countries for computing reserve liabilities.

The third method of computing the legal reserve requirement is to relate it to assets. The principal advantage of this base is that it permits different reserve requirements to be imposed on different assets and thus facilitates selective credit control. If the interest return, the risk, and other relevant features are the same for two types of loan or investment, banks will prefer the type carrying a lower reserve requirement. Credit will therefore tend to be diverted from one use to another, and a difference between the interest rates on the two types of loan may appear. A difficulty is that the banks can control the total amount of reserves required by changing the composition of their assets and may be able to bring about an excessive expansion of total credit by concentrating on loans and investments subject to low reserve ratios. The task of the monetary authorities is complicated because the amount of credit expansion that can occur on the basis of a change in reserve money is indeterminate when substantially different reserve ratios apply to different kinds of loans and investments. The system, like all methods of selective credit allocation, is also subject to the difficulty that the ultimate destination of credit may be quite different from the immediate or ostensible purpose of the loan. At the present time, there seems to be no country that bases its reserve requirements solely on bank assets. Mexico obtains somewhat the same selective effect by allowing required reserves based on deposit liabilities to be satisfied in part by loans of specified types. Permission to substitute certain securities for cash reserve requirements is more common.

Another variant of the asset-based reserve ratio is a requirement that paid-in capital (and sometimes surplus) equal some minimum percentage of deposit liabilities. Requirements of this kind are rather widespread and are usually intended to protect depositors. Capital requirements in a few countries have at times prevented expansion of bank deposits and it is possible that on some occasions the authorities have withheld permission to increase the capitalization of a bank in order to prevent the growth of deposits.

Definition of legal reserves

For reserve requirements to be effective in limiting credit, legal reserves must be in some sense scarce. The definition of the legal reserves of banks has an important bearing on their scarcity.

Primary reserves

Primary reserves consist of cash, but the actual items of currency and deposits included in this category vary from country to country. In some countries only deposits with the central bank may be counted as legal reserves. In other countries part of the primary reserve may be kept in the form of currency in the banks’ own vaults. An argument in favor of allowing vault cash to be included in the legal reserve is the fact that banks located far from a branch of the central bank must generally keep a higher proportion of their assets in the form of vault cash because of the difficulty of replenishing it on short notice. If vault cash cannot be included in legal reserves and if reserve requirements are geographically uniform, the banks in more remote areas may have to maintain a higher ratio of cash to deposits than banks in financial centers.13 The practice of including vault cash in reserves, however, has sometimes led to abuses by banks which have failed to maintain the portion of legal reserves that they were allowed to hold in their own vaults.

The treatment of interbank deposits varies from country to country. Smaller banks and banks outside financial centers often find it expedient to maintain deposits with correspondent banks. If these deposits are included in legal reserves, the banks maintaining them are subject to the same limitation on their lending power as they would be if the deposits were held with the central bank, but the capacity of the banking system as a whole to expand credit is not reduced unless the correspondent banks are required to hold higher reserves against interbank deposits than against other deposits. If 100 per cent reserves were required against interbank deposits, inclusion of these deposits in legal reserves would not increase the capacity of the banking system to expand credit; indeed, failure to include the interbank deposits in reserves would mean that required reserves of the system could change fortuitously as interbank deposits were shifted from a correspondent bank to the home bank. Extending this line of reasoning, banks could be permitted to include as reserves that portion of deposits with other banks which those banks are required to hold as reserves against their interbank deposits. For example, if banks were required to hold a 20 per cent reserve (in the central bank) against deposits of other banks, the latter might be allowed to include in their legal reserves 20 per cent of their deposits with correspondent banks. Under this arrangement, no change in the total reserves of the banking system and its lending capacity would occur as a result of the shifting of bankers’ deposits between the central bank and commercial banks that act as correspondents for other banks.14

A special problem exists where branches of foreign banks constitute an important part of the banking system. These branch banks usually hold funds abroad with their head offices and may meet increases in reserve requirements by repatriating these balances or by borrowing from the head office. A similar situation exists with domestic banks which hold large balances abroad, and also with domestic banks which may be able to borrow abroad. There are, of course, limits to the extent to which foreign balances may be drawn down and the amount of loans that may be raised abroad. Nevertheless, such shifts of funds introduce a lag in the response of the banking system to changes in reserve requirements. This phenomenon has been of some importance in New Zealand, where frequent changes in the reserve ratio have been made.

Secondary reserve requirements

Secondary reserves, as the term is used here, consist of assets other than cash or deposits with the central bank which can satisfy reserve requirements. Secondary reserve requirements have been used for three general purposes: (1) as a means of preventing the private banks from disposing of government securities to expand other loans and investments; (2) as an instrument of selective credit control to influence the allocation of credit among private borrowers; and (3) as a means to ensure good banking practices. They also have the effect of permitting banks to earn a return on required reserves and thus of divorcing to some extent the question of bank profits from reserve policy.

It is important to emphasize that a secondary reserve requirement does not limit the amount of loans and investments that an individual commercial bank can acquire; it merely influences the form that the loans and investments can take. Whether a secondary reserve requirement influences the lending capacity of the banking system as a whole depends on how the banks acquire the assets to satisfy the requirement. In the special case in which these assets consist of government securities or other paper that is available only from the portfolio of the central bank, and the central bank does not automatically replenish its portfolio, the secondary reserve has exactly the same effect on the banking system’s lending capacity as a reserve in the form of a deposit with the central bank. At the other extreme, if the commercial banks acquire assets to satisfy secondary reserve requirements direct from the borrowers, then the requirement has no immediate effect on the lending capacity of the banking system. A secondary reserve requirement may nevertheless make it possible to hold total credit to a lower level than would otherwise be feasible, because the government and other borrowers whose obligations are acceptable in satisfaction of the requirement may be so influential that their demands for credit would be met in any case. In the absence of the secondary reserve requirement, the commercial banks would continue to lend to regular customers, and the central bank would have to allow the preferred borrowers to be accommodated also.

The use of secondary reserve requirements to prevent banks from reducing their holdings of government debt in order to expand private loans and investments came into prominence after World War II. At the end of the war most of the banking systems of the world held large amounts of government securities, usually bearing low coupon rates of interest. Because of the large amount of debt outstanding and the necessity, in many cases, of incurring more government debt, many countries chose to support the price of government securities to hold down the interest burden. The support policy made it possible for the private banks and other holders to convert their government securities into cash with, at most, small losses. The central bank acted as the ultimate buyer and thus supplied more cash reserves to the banking system. The banks were naturally inclined to liquidate a large part of their government securities and to acquire assets with higher yields, and at the same time to restore the traditional distribution of their assets. To prevent this from happening, Belgium in 1946, Italy in 1947, and France in 1948 established reserve requirements compelling the banks to hold certain amounts of specified government securities. For similar purposes Austria, the Netherlands, and Sweden introduced similar requirements following the outbreak of the Korean war. A proposal along these lines was extensively debated in the United States but was not adopted. In general, it appears that the secondary reserve requirements had some success in reducing pressures on the government bond market, but certain undesirable consequences also appeared in the course of time.

Most secondary reserve requirements took the form of relating minimum holdings of government securities to the level of deposits. As deposits grew the banks were required to purchase additional securities, thus providing the government with a cheap means of financing budget deficits. There was, moreover, a certain perversity in the requirement, in that in a time of boom, when deposits were growing, the financing of government deficits was made easy, whereas during a recession the government had to enter the capital market or money market in competition with private investors at a time when this might not be desirable.

In 1957 Belgium introduced several reforms to correct the difficulty arising from holding government securities as required reserves. The principal feature of this action was to make future issues of securities for reserve requirement purposes independent of the Government’s budgetary position. The banks were allowed to exchange part of their treasury certificates held as secondary reserves for new securities bearing a more realistic rate of interest. In future, as their deposits increase, the banks must invest, as required reserves, a prescribed amount in certificates issued by the Fonds des Rentes, a government bond equalization account. The banks may also invest, as partial fulfillment of their reserve requirements, in short-term treasury bills which, in contrast to their former holdings of such securities, will bear realistic interest rates and may be traded in among the banks and other financial institutions. Besides halting the automatic financing of the Treasury, it was also hoped that the reform would lead to the development of a more active government securities market, would increase the effectiveness of discount rate policy, and would enable the monetary authorities to engage in open market operations.15

As the authorities of more and more countries have abandoned the policy of supporting government bond prices and interest rates have risen from artificially low levels, reliance on secondary reserve requirements to freeze bank holdings of government securities has diminished.

Secondary reserve requirements may also be used as a means of selective control of private credit. Several different methods of introducing selectivity have been employed. One method is to allow certain earning assets to be substituted for part of a cash reserve requirement. A second method is to prescribe a lower set of cash reserve requirements if certain amounts of particular assets are held; this method has been employed in Brazil. Sometimes, however, the banks are explicitly required to hold certain assets equal to a stated proportion of their deposit liabilities.

The form of the law or regulation has some relation to the purpose for which the reserve requirement is employed. If the law or regulation states that a certain type of asset must be held as part of the requirement and does not permit the substitution of cash or other assets, the system may provide a quantitative restraint on the amount of credit that the banking system may extend if the specified asset is scarce. If the specified asset is easily available, the effect is merely to influence the allocation of credit.

Mexico has made the most extensive use of differentiated secondary reserve requirements to channel credit into “productive” purposes, maintaining an elaborate system for this purpose.

The original purpose of secondary reserve requirements was to ensure good banking practices by requiring the banks to maintain a minimum degree of liquidity. Legislation of this type was introduced in Denmark, Norway, Sweden, and Switzerland, mostly in the 1930’s, and is still in force. The original minimum ratios, however, were generally low and had little influence in the postwar situation, when banks were in a highly liquid position. New ratios have been introduced in several countries for purposes of monetary control. In Switzerland, for example, the secondary reserve requirement was used as an instrument of monetary policy but retained the form of the original law.16

A less formal prescription of secondary reserves has been made in some countries by central banks “making suggestions” to the private banks concerning desirable asset structures. This approach was followed in the Federal Republic of Germany to provide a guide for the private banking system to what the central bank regarded as proper banking practices, in view of the fact that the traditional relationships among the assets of the banks had been disrupted by war and postwar events. Permission to exceed rediscount ceilings was based in part on the banks’ adherence to these standards. Other central banks at various times have suggested desirable asset ratios for particular assets, with mixed results. The success of these suggestions has often depended on whether the central bank had alternative powers which it could employ if it did not obtain compliance.

Differentiation among deposits

In an attempt to remove some of the shortcomings of the deposit liability method of computing reserve requirements and to make reserve requirements a more precise and specialized means of credit control, different reserve ratios are often required for different categories of deposit.

Demand and time deposits

The most widespread distinction between deposits is that between demand deposits and time deposits, with lower requirements being applied to the latter. This form of differentiation originated at a time when reserve requirements were regarded as primarily a safeguard of the banks’ solvency. It was based on the general banking principle that more liquid assets should be held against more liquid liabilities. This principle was the underlying basis for the original form of the reserve laws in Denmark, Norway, Sweden, and Switzerland. However, there are also monetary reasons for a distinction between demand and time deposits. It is a convenient method of applying the principle of relating reserves to the turnover of deposits, since demand deposits usually have a higher rate of turnover than time deposits. In addition, institutions that have a large percentage of time deposits, e.g., savings banks, traditionally have invested in less liquid assets and extended longer term credits, e.g., mortgages. These institutions have usually held a lower liquidity ratio than commercial banks, so that the application of lower reserve requirements to time deposits disrupted customary banking practices less than the enforcement of a uniform requirement.

It has been suggested, moreover, that the application of a lower reserve ratio to time deposits may have the effect of bringing about an automatic countercylical variation in total required reserves. This suggestion is based on the assumption that demand deposits typically grow at a faster rate than time deposits during periods of business expansion. On this assumption, the ratio of required reserves to total deposits will, under a differential system, increase during a boom and decrease during a recession.17 The assumption that demand deposits increase faster than time deposits during boom periods is, however, open to question; during the last few years the reverse has been true in several large countries (e.g., Canada, the Netherlands, the United Kingdom, and the United States). This is not surprising because, with interest rates tending to rise and fall with general business activity, higher rates on time deposits provide an incentive to business firms and households to shift funds from demand deposits to time deposits. Such behavior may well be encouraged by the banks if there is a large difference between the reserve requirements on the two kinds of deposits and if the authorities are pursuing a policy of credit restraint.

The shifting of deposits from demand to time accounts has in practice created difficulties in several countries, e.g., Colombia and Peru. The shift has to be initiated by depositors, but banks can often find means of encouraging it. Therefore, in order to combat the practice it has sometimes been considered necessary to narrow or eliminate the difference between the required reserve ratios for time deposits and demand deposits. But when the change is brought about by increasing requirements for time deposits, institutions that customarily hold large amounts of time deposits face a difficult adjustment problem. One method of ameliorating their position would be to place a high reserve requirement on increases in time deposits rather than on total time deposits.

While the differentiation between time and demand deposits is widespread, it is by no means universal. Many countries have the same requirement for both types of deposit; others have reserve requirements only against demand deposits. There are some unusual cases, such as Paraguay, which has higher requirements for time deposits, and Iceland, which has requirements only for time deposits; presumably these requirements are intended to provide a greater margin of safety for savings depositors.

Interbank deposits

Interbank deposits are often subject to higher reserve requirements (up to 100 per cent) on the ground that these deposits are part of the reserves of other banks and are subject to withdrawal in times of financial crisis, and therefore it is unwise to allow a credit structure to be built upon them. The geographical differentiation of reserve requirements in the United States was originally based principally on differences between the ratios of interbank deposits to total deposits in New York and Chicago, regional financial centers, and other places.

Foreign currency deposits

Higher reserve requirements (often up to 100 per cent) may be imposed on foreign currency deposits. The usual motive is probably to assure conservative banking practices with respect to a volatile type of deposit, but in some cases it appears that the objective has been to discourage the banks from accepting such deposits. By imposing a high reserve requirement on foreign currency deposits, moreover, the authorities may hope to prevent a capital outflow that would otherwise occur because the banks would acquire foreign assets to cover their foreign currency liabilities in order to avoid exchange risk. In Mexico, for example, following the devaluation of 1954 there was a rapid growth of dollar deposits in Mexican banks, and the authorities required the banks to hold a 100 per cent secondary reserve against these deposits in the form of dollar securities issued by Mexican institutions. The banks thus acquired the cover they demanded for dollar liabilities, without the country experiencing a loss in its foreign exchange reserves. The opposite approach is followed in Switzerland, which imposes no reserve requirement on foreign currency deposits, in order to encourage the banks to favor the deposit of these funds in Swiss banks.

Nonresident accounts

Deposit accounts of nonresidents were made subject to higher reserve requirements in the Federal Republic of Germany in 1957, to prevent credit being extended against these deposits—which in large part represented short-term speculative movements of capital—there being at the time rumors of a devaluation of sterling and an appreciation of the deutsche mark.

Marginal reserve requirements

Special requirements, sometimes called marginal reserve requirements, are applied in a number of countries to increases in deposits after some specified date or to increases above the average of a past period, rather than to the absolute level of deposits.

This technique may be especially helpful in minimizing the expansionary effect of a large increase in foreign exchange holdings due to an export boom, a capital inflow, or a large government deficit. The banks may be unequally affected in the first instance, and an increase in average reserve requirements great enough to deal with the situation would work a severe hardship on some of them. Marginal reserve requirements at high rates—100 per cent in some cases—can be applied without disrupting the normal operations of the banks. An example of the successful use of high marginal reserve requirements to counteract the inflationary effect of a balance of payments surplus is provided by Mexico. In the devaluation of 1954 the growth of private credit was limited by high marginal reserve requirements, which facilitated the rebuilding of the country’s foreign exchange reserves.

Another example of the use of the technique is provided by Australia, which in 1940 instituted a 100 per cent reserve against increases in deposits and was successful in limiting the inflationary impact of government deficits arising from defense spending.

An objection to high marginal reserve requirements is that they discriminate against banks whose deposits were abnormally low at the base date or which had not shared fully in an expansion taking place prior to the base date. If the marginal requirements are continued for an extended period of time, they discriminate harshly against new and growing banks. Several methods have been adopted to minimize these objections. In India the law authorizing marginal reserve requirements specifies that the average ratio to deposits shall not exceed a certain point. Peru at one time varied the marginal requirement according to the age of the bank, in order to reduce the calls on new banks. Other countries have followed the practice of consolidating marginal reserve requirements into higher average requirements from time to time.

Differential requirements for increases in deposits are to be regarded primarily as a supplement to average requirements which may be especially suitable for short-period, emergency use. Australia seems to be the only country that has relied solely on marginal reserves.

Differentiation among commercial banks

In the Federal Republic of Germany, Mexico, and the United States, reserve requirements are differentiated on the basis of the location of the bank. In all three countries, banks located in cities in which there is a branch of the central bank are required to hold higher reserve ratios than banks in other places. One reason for this distinction may be that banks located near a branch of the central bank need to maintain less vault cash than banks in more remote locations, because they can obtain currency more quickly to meet withdrawals by their customers. (Another way of meeting the problems of banks in remote locations, as noted above, is to allow the inclusion of vault cash in required reserves; this is not done in Germany, Mexico, or the United States.) Higher reserve requirements for banks in financial centers may also be considered necessary to protect their solvency, since some of them hold substantial amounts of deposits of other banks, which are especially likely to be drawn down rapidly when there is a contraction of total deposits. In the United States, the higher requirements for banks in New York and Chicago, which constitute a separate class, are a carry-over from the period before the establishment of the Federal Reserve System, when interbank deposits were more important than they are today and when reserve requirements were intended to assure solvency rather than to be used as an instrument of monetary policy. The Federal Reserve Board has recently recommended amendments to the banking law that would reduce the geographic differentiation of U.S. reserve requirements.18

Geographic differentiation can be supported from the point of view of monetary policy on the ground that banks located in financial centers tend to have a higher rate of deposit turnover than other banks. Expansion of deposits at banks in financial centers may lead initially to a stimulus to demand for shares and other financial assets and thus promote speculative activity to an undesirable extent. Furthermore, the proportion of liquid assets that banks in financial centers choose to hold is usually higher than the proportion held by other banks; therefore, it is less inconvenient for them to comply with a high reserve requirement than it is for banks in smaller cities and towns.

Differential reserve requirements can also be based on the size of the banks, large banks being required to hold higher ratios than small banks, as in Belgium, Germany,19 Norway, and Sweden. It has been suggested that the characteristics often attributed to location—such as the importance of interbank deposits and the rate of deposit turnover—are in fact associated with the size of the bank or individual banking office. It is clear that most of the banks with the special characteristics mentioned are large institutions; it is less clear whether large banks in provincial capitals differ from those in financial centers. One statistical study found that in the United States there is a positive correlation between the size of the individual banking office and the rate of deposit turnover, and the author suggested that differences in reserve ratios should be based on the size rather than the location of the bank.20

The ultimate in differentiation among banks is the Australian special accounts system, under which a separate reserve ratio can be imposed on each bank. This approach would obviously give rise to administrative difficulties in a country having a large number of banks, and its application may create friction between the central bank and the commercial banks.

Treatment of reserve deficiencies

In order for reserve requirements (or any quantitative controls short of explicit credit ceilings) to be effective as an instrument of credit restraint, it is essential that the banks be reluctant to maintain their reserves below the required minimum.

In countries with highly developed financial systems it appears that moral suasion and the fear of adverse publicity are the most powerful forces preventing reserve deficiencies. The private banks are expected to discount with the central bank, borrow from other banks, or liquidate assets, whenever necessary to maintain the required level of reserves.

A penalty rate of interest is usually imposed for any bank reserve deficiencies. The penalty is ordinarily set at a rate somewhat above the discount rate of the central bank. Sometimes it is a fixed amount above the discount rate; at other times the level is left to the discretion of the central bank, perhaps with a maximum differential specified. Frequently, the penalty rate of interest is progressive and increases with the length of time that the reserve deficiency persists.

In order to be effective in discouraging reserve deficiencies, the penalty rate of interest may have to be a good deal higher than the discount rate of the central bank and even higher than the official or legal maximum rates of interest that the private banks may charge on loans. In many countries, particularly the underdeveloped countries, the curb rate of interest may be considerably higher than the bank rate of interest, and the private banks through various devices (such as requiring a proportion of any loans to be left on deposit) may increase the effective rate of interest to borrowers.

In view of the widely varying circumstances and degree of cooperation that may be encountered, it is desirable that the monetary authorities have discretionary power to vary the penalty rate, and that if any maximum rate is set it be high enough to deal with wide fluctuations in the effective rate of interest. In many cases a penalty rate that is only slightly over the discount rate may be sufficient, while in other cases even a rate of 7 or 8 per cent above the discount rate may not be adequate to discourage reserve deficiencies.

Aside from penalty rates of interest, some monetary authorities are empowered to take control of a bank if a reserve deficiency persists for more than a specified period of time and may, in extreme cases, liquidate the offending bank and fine or imprison its officers.

The computation of legal reserve deficiencies is usually based on an average of deposit liabilities and reserve holdings over a period of time, but the determination is sometimes made with reference to the position on a particular statement day or call date. The method of averaging, or the time interval of reporting, may allow deficiencies to exist for a short period. When deficiencies are increasing rapidly, reserve statistics may be misleading, especially if reporting dates are widely separated. It is convenient to have all banks report on the same day to prevent the shifting of reserves from bank to bank as a means of circumventing reserve requirements.


I. Effect of Cash Reserve Requirements on Expansion of Money Supply and Allocation of Credit21


article image

Expansion factor


Substituting from equation (2) in equation (6):


and further substituting from equations (3) and (5):


Canceling ΔM:


The change in deposits in relation to a primary change in reserve money is (1—c)E; the change in currency is cE.

Allocation of credit

The change in central bank loans and investments equals the change in reserve money, which equals 1E times the change in total money supply, or c + r (1 - c).

The change in commercial bank loans and investments in relation to the change in reserve money is


In relation to change in total money supply, the change in commercial bank loans and investments is


II. Summary of Reserve Requirements in Various Countries

This summary is based on the latest information available in Washington on April 30, 1958. It is an abstract of a detailed compilation (97 pages) by the authors, which gives information on the variable reserve requirements in individual countries and a chronology of all known changes in legal reserve requirements.

Mimeographed copies of the detailed compilation may be obtained from The Secretary, International Monetary Fund, Washington 25, D.C.

A liberal interpretation of variable reserve requirements has been employed, and the survey includes several countries with legally fixed requirements which have been altered at various times. Countries with variable reserve requirements of either cash or other assets have been included.

Countries with variable reserve requirements

In the following countries, reserve requirements have been varied from time to time (the figures in parentheses indicate the number of times):

Australia 22

Belgium (1)

Bolivia (4)

Brazil (6)

Burma (1)

Ceylon (1)

China (Taiwan) (2)

Colombia (30)

Cuba (2)

Ecuador (19)

Egypt (4)

Finland (1) 23

France (1)

Germany (9)

Greece (3)

Haiti (1)

Honduras (2)

Iran (1)

Israel 22

Italy (1)

Korea (10)

Mexico 22

Netherlands (13)

New Zealand (25)

Nicaragua (1)

Paraguay (1)

Peru (14)

Sweden 22

Switzerland (5) 23

Syria (2)

Thailand (1)

United States (26)

Viet-Nam (2)

In the following countries, the reserve requirements have never been varied:



Cambodia 24



Costa Rica

Dominican Republic

Federation of Nigeria 24

Federation of Rhodesia and Nyasaland

Ghana 24





Ireland 24

Japan 24




Spain 24

Union of South Africa



Countries with fixed reserve requirements


El Salvador






Countries without reserve requirements








Saudi Arabia

Sudan 25


United Kingdom 26

III. Selected Bibliography

  • Burgess, W. R., The Reserve Banks and the Money Market (New York, Harper, 3rd ed., 1946).

  • Fousek, P. G., Foreign Central Banking: The Instruments of Monetary Policy (New York, Federal Reserve Bank of New York, 1957).

  • France, Ministère des Finances, “Le contrôle des banques et du crédit en France et à l’étranger,” Statistiques et Etudes Financières: Supplément (Paris), No. 90, June 1956, pp. 631719.

    • Search Google Scholar
    • Export Citation
  • Goldenweiser, E. A., American Monetary Policy (New York, McGraw-Hill, 1951).

  • Grove, D. L., Commercial Bank Reserves and Reserve Requirements (Washington, Board of Governors of the Federal Reserve System, 1952, mimeographed).

    • Search Google Scholar
    • Export Citation
  • [King, W. T. C.], “Should Liquidity Ratios Be Prescribed?” The Banker (London), Vol. 106 (1956), pp. 18697.

  • Phillips, C. A., Bank Credit (New York, Macmillan, 1920).

  • Sen, S. N., Central Banking in Undeveloped Money Markets (Calcutta, Bookland, 1952).

  • U. S. Congress, Joint Committee on the Economic Report, Monetary Policy and the Management of the Public Debt: Replies to Questions and Other Materials for the Use of the Subcommittee on General Credit Control and Debt Management (82nd Cong., 2nd Sess., Washington, Government Printing Office, 1952, 2 vols.).

    • Search Google Scholar
    • Export Citation
  • Zwoll, J. H. van, Mindestreserven als Mittel der Geld-und Kreditpolitik (Berlin, Duncker & Humblot, 1954). Contains bibliography of works available in German dealing with reserve requirements.

    • Search Google Scholar
    • Export Citation

Mr. Goode, Assistant Director of the Asian Department, was formerly Chief of the Finance Division. Before joining the Fund staff he was assistant professor of economics at the University of Chicago and economist at the U.S. Bureau of the Budget and the U.S. Treasury Department.

Mr. Thorn, economist in the Finance Division, is a graduate of Columbia College, the University of Maryland, and Yale University. He is temporarily assigned to the European Office of the International Monetary Fund in Paris.


The banks can also extend loans and investments that are financed out of bank capital and surplus and borrowings, but in most countries these items are small compared with deposit liabilities.


A primary increase (or decrease) in deposits is an increase (or decrease) in deposits of the commercial banking system that is matched by an equal change in the central bank’s liabilities to the commercial banks.


See J. J. Polak, “Monetary Analysis of Income Formation and Payments Problems,” Staff Papers, Vol. VI (1957–58), pp. 1–50.


No allowance is made for the external drain, banks’ till money, or the possibility that banks will hold excess reserves. Hence the actual expansion in a particular case might fall far short of the maximum potential expansion.


See Appendix II. It appears that up to April 30, 1958 there had been no actual variation in 23 of the 56 countries having variable reserve requirements.


In the United States, the extension of reserve requirements to all banks accepting demand deposits, including nonmember banks, was recommended by the Douglas Committee in 1950 (Report of the Subcommittee on Monetary, Credit, and Fiscal Policies, 81st Cong., 2nd Sess., Washington, 1950, pp. 2–3) and by the Patman Committee in 1952 (Report of the Subcommittee on General Credit Control and Debt Management, 82nd Cong., 2nd Sess., Washington, 1952, p. 45).


R. S. Sayers, Central Banking After Bagehot (Oxford, 1957), pp. 89–90.


Table 2 shows that, given a ratio of currency to money of 25 per cent, the proportion of a monetary expansion reflected in an increase in central bank liabilities will be 32.5 per cent with a 10 per cent reserve ratio and 62.5 per cent with a 50 per cent reserve ratio: .325 × .05 = .01625; .625 × .05 = .03125. Another way of looking at the matter is to note that, regardless of the reserve requirement, the central bank will have available one fourth of the increase in money represented by currency–1.25 percentage points out of a 5 per cent expansion of money. With a marginal reserve ratio of 10 per cent, the total available to the central bank will be 1.25 + .10 (5.00–1.25) = 1.625 percentage points. Raising the marginal reserve ratio to 50 per cent will increase the total to 1.25 + .50 (5.00–1.25) = 3.125 percentage points.


In July 1958 the U.K. Chancellor of the Exchequer announced a scheme for “special deposits” at the Bank of England by the commercial banks, which would be employed, when necessary, “in support of other monetary measures, to restrict the liquidity of the banking system and thus the ability of the banks to extend credit.…” Calls would be made on each group of banks separately and would be related to total gross deposits. The use of special deposits had an antecedent in the Treasury Deposit Receipts (TDR’s) introduced as a wartime measure. TDR’s bore a fixed, low rate of interest and were “levied” upon the banks by a bankers’ committee after the over–all amount was determined by the Treasury. See The Times (London), July 4, 1958, p. 6, and The Economist (London), July 12, 1958, pp. 141–43.


For example, after the creation of the Reserve Bank of India, the cash reserve ratio of the private banks in India fell from 17.4 per cent in 1935–36 to 9.5 per cent in 1938–39; and in the United States the ratio (including required reserves and interbank deposits) fell from 34 per cent in 1913 to 20 per cent in 1926. See S.N. Sen, Central Banking in Undeveloped Money Markets (Calcutta, 1952), p. 88.


For detailed studies of these problems, see Richard S. Thorn, “Nonbank Financial Intermediaries, Credit Expansion, and Monetary Policy,” and Eugene A. Birnbaum, “The Growth of Financial Intermediaries as a Factor in the Effectiveness of Monetary Policy,” Staff Papers, Vol. VI (1957–58), pp. 369–426.


This method was originally proposed by W. W. Riefler, of the Board of Governors of the Federal Reserve System, in 1931; see E. A. Goldenweiser, American Monetary Policy (New York, 1951), pp. 51–52.


In the United States, the Board of Governors of the Federal Reserve System recently proposed that the law be amended to authorize the Board to permit member banks to include all or part of their vault cash in required reserves. See Federal Reserve Bulletin, April 1958, pp. 427–29.


E. A. Goldenweiser, op. cit., pp. 49–50.


“La réforme du marché monétaire,” Banque Nationale de Belgique, Bulletin d’Information et de Documentation, November 1957, pp. 377–90.


The gentlemen’s agreement requiring deposits in blocked “M” accounts at the Swiss National Bank was canceled in March 1958.


J. M. Keynes, A Treatise on Money (London, 1930), Vol. II, Chap. 23.


The Board recommended that the range of permissible variations in reserve ratios for New York and Chicago banks (central reserve city banks) be reduced to that prevailing for banks in other large cities (reserve city banks) and proposed that the Board be given more authority to permit individual banks in central reserve cities and reserve cities to carry reserves lower than those now specified for banks in these cities. See Federal Reserve Bulletin, April 1958, pp. 427–29.


In Germany there are 6 size classes and 2 geographic classes—thus 12 classes in all.


Neil H. Jacoby, “Flexible Bank Reserves and Sustained Prosperity,” Commercial and Financial Chronicle (New York), November 21, 1957, p. 44.


No allowance is made for till money of banks, for the possibility that banks will hold excess reserves, or for any outflow of money due to an “external drain”; see text, p. 11.


Reserve requirements have been effectively varied many times.


Variable legal reserve requirements have been terminated.


Reserve requirements have not yet been implemented.


New central bank draft law provides for variable reserve requirements.


Customary reserve ratios are in force and are regularly maintained. See also footnote 9.

IMF Staff papers: Volume 7 No. 1
Author: International Monetary Fund. Research Dept.