Abstract
Financial programs supporting stand-by arrangements in the upper credit tranches typically include ceilings either on the domestic credit extended by the central bank or on the domestic credit extended by the banking system as a whole. 1 Domestic credit ceilings are applied whether a country has a fixed or a flexible exchange rate. The emphasis placed by the Fund on domestic credit has recently been subjected to criticism. One criticism arises from the fact that, whereas domestic credit expansion is the important determinant of international reserve developments, it is the rate of monetary expansion that is the important determinant of price level and income level developments. A recent assessment of the economic outlook in the United Kingdom concluded:
Financial programs supporting stand-by arrangements in the upper credit tranches typically include ceilings either on the domestic credit extended by the central bank or on the domestic credit extended by the banking system as a whole. 1 Domestic credit ceilings are applied whether a country has a fixed or a flexible exchange rate. The emphasis placed by the Fund on domestic credit has recently been subjected to criticism. One criticism arises from the fact that, whereas domestic credit expansion is the important determinant of international reserve developments, it is the rate of monetary expansion that is the important determinant of price level and income level developments. A recent assessment of the economic outlook in the United Kingdom concluded:
As a concept, DCE [domestic credit expansion] is less appropriate than sterling M3 for monitoring the overhang of purchasing power which is a potential risk of an inflationary avalanche. DCE is primarily an operational device for enforcing a squeeze on the domestic money supply of a country whose balance of payments is leading to a loss of reserves…. [Sterling M3] should be reinstated as “the focus of our monetary policy”….2
In the context of flexible exchange rates, domestic credit ceilings have been criticized as being inappropriate; the overall balance of payments is seen as the result of official intervention policy and not the result of credit expansion. In 1977, a number of distinguished economists reported that: “The IMF recommendations for control of domestic credit expansion rather than money growth have no rationale in a world of floating exchange rates. Their proposals are a poor substitute for effective policies that can reduce inflation and increase growth.” 3 The purpose of this paper is to examine the appropriateness of domestic credit ceilings under alternative exchange rate regimes and to assess whether it would be beneficial to replace or supplement domestic credit ceilings with money supply ceilings.
The structure of this paper is as follows. Section I discusses the concept of domestic credit in relation to the more familiar concept of the money supply, and examines a number of definitional and measurement problems. Section II compares domestic credit and money ceilings under exchange rate regimes where there is a known predetermined parity or target for the exchange rate. Section III compares domestic credit and money ceilings under flexible exchange rates.
I. The Concept of Domestic Credit
In general, domestic credit may be viewed as a measure of the total purchasing power generated by financial institutions in a country. This purchasing power extends over goods, services, and financial assets produced in both the home country and overseas. In a sense, the stock of money is an ex post measure of the purchasing power remaining after the leakage of residents’ purchases of foreign goods, services, and financial assets and the injection of nonresidents’ purchases of domestic goods, services, and financial assets. Equivalently, domestic credit expansion is a measure of the increase in the money supply caused by domestic financial institutions—that is, what is obtained after adjusting the realized increase in the money supply to take account of the change caused by the external payments surplus or deficit. Only if there is external payments balance will domestic credit expansion and realized monetary expansion be equivalent.
A more rigorous examination of the concept of domestic credit is presented in the remainder of this section. However, since the discussion occasionally becomes intricate and technical and is not crucial to an understanding of the subsequent sections, it could be omitted.
Domestic credit refers to a particular subset of assets minus liabilities in the balance sheet of either a single financial institution or a group of financial institutions. This paper considers only the domestic credit extended by the central bank and the domestic credit extended by the banking system as a whole 4(i.e., by the central bank and commercial banks combined). Other measures of domestic credit that might be considered of economic significance are the domestic credit extended by commercial banks alone, and the domestic credit extended by both the banking system and those nonbank financial intermediaries whose liabilities are close substitutes for money.
Since the total of assets minus liabilities in any balance sheet is identically equal to zero, any subset of assets minus liabilities must add up to the remaining sum of liabilities minus assets in the balance sheet. Hence, from the balance sheet of the banking system, there are two ways of measuring domestic credit that yield identical results. Furthermore, since any asset in the balance sheet of the banking system has as a counterpart a liability in the balance sheet of a nonbank, and since this liability is itself equal to the sum of all other assets minus liabilities in the non-bank’s balance sheet, it transpires that there are innumerable ways of measuring domestic credit; all measures are identically equal but, in a sense, incorporate different levels of aggregation. Publication of alternative ways of obtaining the same total may be useful in assessing the economic significance of the total; in the maze of numbers contained in a full flow-of-funds matrix, the alternative measures of domestic credit expansion that are published may be seen as presenting at least some of the important interrelationships that exist within the flow-of-funds accounts. 5 Furthermore, owing to delays in obtaining certain data and to the limited ability of the authorities to forecast with reasonable accuracy the future values of certain other data, alternative measures may be useful in obtaining data on past values faster and in obtaining better estimates of future values.
Three ways of measuring domestic credit expansion are popularly used. In very general terms, domestic credit expansion of the banking system equals
(1) the increase in the sum of banking system claims on the public sector, the nonbank private sector, and the overseas sector; 6
(2) the increase in the money supply less the increase in the net foreign assets of the banking system;
(3) the public sector borrowing requirement less the increase in public sector debt held by the nonbank private sector and the overseas sector, plus the increase in banking system claims on the nonbank private sector and the overseas sector.
The first measure of domestic credit expansion looks primarily at the asset side of the banking system’s balance sheet. As the name suggests, domestic credit expansion is simply the sum of the credit extended by the domestic banking system to other sectors of the economy, where the other sectors are conveniently split into the public sector, the nonbank private sector, and the overseas sector. Credit to the overseas sector should be included, since it is likely to be as important in determining the balance of payments 7 as credit extended to other sectors; if it were to be excluded, a ceiling on domestic credit would place no limit on the possible adverse effect on the balance of payments that could arise from the overseas sector borrowing domestic currency and converting it into foreign currency.8
The second measure of domestic credit expansion is simply the sum of the changes in those balance sheet items not included in the first measure. It focuses primarily on the liabilities side of the banking system’s balance sheet, being equal to a measure of the increase in the money supply less a measure of the surplus on the overall balance of payments. The money supply definition that is used is typically a broad money definition including interest-bearing deposits as well as checking deposits. Although countries customarily publish data on more than one definition of money, they do not generally publish corresponding data on the various aggregates of domestic credit. The measure of the overall balance of payments that is used is the increase in the net foreign assets of the banking system. As such, capital flows undertaken by both the commercial banks and the central bank are below the line. In this regard, it is significant that a freely flexible exchange rate does not ensure that domestic credit expansion and monetary expansion are equal. A combination of a domestic credit ceiling and nonintervention by the authorities in the exchange market does not in itself prevent the possibility of money supply changes resulting from the importation or exportation of funds by commercial banks. 9
The third measure of domestic credit expansion is simply the first measure with credit extended to the public sector by the banking system replaced by the public sector borrowing requirement less the credit that is received by the public sector from sectors other than the banking system. This measure shows clearly that a change in fiscal policy will have an inevitable effect on the flow of funds between sectors and is suggestive of a danger in examining fiscal policy independently of monetary considerations.
Domestic credit expansion by the central bank may be similarly measured in three ways:
(1) the increase in the sum of central bank claims on the commercial banks, the public sector, the nonbank private sector, and the overseas sector;
(2) the increase in the monetary liabilities of the central bank (i.e., reserve money, high-powered money, or the monetary base),less the increase in the net foreign assets of the central bank; and
(3) the public sector borrowing requirement, less the increase in public-sector debt held by the commercial banks, the nonbank private sector, and the overseas sector, plus the increase in central bank claims on the commercial banks, the nonbank private sector, and the overseas sector.
There are some important differences between the domestic credit extended by the central bank and the domestic credit extended by the banking system. First, the former includes the central bank claims on the commercial banks that are netted out when the central bank and the commercial banks are aggregated as the banking system. Second, being a component of the monetary base, central bank domestic credit is “high-powered” credit. Third, the measure of the balance of payments that is relevant is the change in the net foreign assets of the central bank, which is closely related to the magnitude of official intervention in the exchange market and which places capital flows by the commercial banks above, rather than below, the line.
Although domestic credit, money, and the overall balance of payments need to be defined in a consistent way from a flow-of-funds point of view, the actual definitions that are used by countries exhibit many technical differences. This section concludes by discussing briefly a number of these technical factors.
(1) Definitions of money differ between countries according to whether they include or exclude public sector deposits, nonresident deposits, and resident foreign currency denominated deposits. If public sector and nonresident deposits are included, domestic credit will be defined to include claims on, but to exclude liabilities to, the public and overseas sectors. If the definition of money excludes these deposits, domestic credit will be defined to include net claims (i.e., claims minus liabilities) on the public and overseas sectors. In the event that resident (or nonresident) foreign currency deposits are included in the definition of money, they will obviously be excluded from the category of banks’ foreign liabilities. Definitions of reserve money typically exclude frozen deposits, such as import deposits and special deposits. However, a resulting problem is that, when such deposits are included as a negative item in central bank domestic credit, an increase or calling in of such deposits increases the amount of potential credit remaining under a given domestic credit ceiling.
(2) One problem in defining the overall balance of payments is whether public sector foreign currency borrowing should be interpreted as autonomous or accommodating. In many countries, public sector industries are encouraged to raise funds abroad in order to help the balance of payments. Such capital flows are accommodating rather than autonomous and should be included below the line. When public sector foreign borrowing is included below the line, it affects the definitions of both net foreign assets and domestic credit; net foreign assets must then be defined to exclude any effects from such foreign borrowing (i.e., foreign liabilities obtained by the public sector will be included as a negative item in net foreign assets and will exactly offset the accompanying foreign assets acquired by the central bank); domestic credit will be defined to include this component of foreign credit and will therefore ensure that, for as long as this definition of domestic credit is being controlled, public sector foreign borrowing will not accommodate any additional increase in the money supply. Yet another problem in measuring the overall balance of payments concerns whether the overall balance of payments should represent a pure flow balance or a change in stocks of net foreign assets—that is, whether revaluation profits should be placed above or below the line. The flow change in international reserves measures the success of trade and capital transactions and is the measure that determines the initial impact on domestic liquidity; the change in the stock of international reserves also reflects the success of official reserve holding policy and determines the ability of the country to meet future flow deficits. From a flow-of-funds point of view, revaluation profits need to be excluded from the data; an adjustment below the line should be made to the change in net foreign assets to offset the revaluation effect. However, if the country is unable to identify independently the revaluation effect and has to measure domestic credit as a residual from the money supply and net foreign asset data, there is no choice but to include the revaluation profit above the line in “errors and omissions” and, implicitly, as a negative component of domestic credit.
For countries whose currencies are used as reserve currencies, there exists the problem of whether domestic currency (i.e., reserve currency) acquisitions by foreign authorities should be placed above or below the line. If such capital flows are regarded as autonomous and placed above the line with all other nonresident acquisitions of domestic currency, then the relationship between domestic credit expansion and monetary expansion will depend only upon the magnitude of official intervention in the exchange market undertaken by the domestic authorities. When liabilities to foreign authorities are treated in this way, as domestic rather than foreign liabilities, controlling the resulting measure of domestic credit will, by definition, insulate the money supply from interventions by foreign authorities. In contrast, if such acquisitions are regarded as accommodating rather than autonomous and are included below the line as foreign liabilities, controlling the resulting measure of domestic credit will mean that foreign interventions will give rise to money supply effects in the same way as domestic interventions.
Owing to institutional differences between countries, there is no presumption that domestic credit should be defined identically for each country. Although an attempt is made in the Fund’s International Financial Statistics to produce money and credit data of a common definition across countries, the definitions of money and credit used by the Fund in stand-by arrangements often preserve the definitions of money and credit that have evolved in the particular country.
II. Domestic Credit and Money Ceilings Under Fixed Exchange Rates
In the context of fixed exchange rates, it has been well established in the literature, particularly by Polak and Argy, 10 that controlling domestic credit generally provides greater stability for the overall balance of payments than is achieved through controlling money; 11 the changes in the money supply that accompany payments imbalances when domestic credit is controlled have a stabilizing effect that tends to move the balance of payments back to equilibrium. More precisely, controlling the domestic credit extended by the central bank provides greater stability for the net foreign assets of the central bank than is achieved through controlling reserve money, while controlling the domestic credit extended by the banking system as a whole provides greater stability for the net foreign assets of the banking system than is achieved through controlling the money supply.
Obviously, if it were possible to forecast correctly the future levels of all exogenous variables and all structural parameters, a given balance of payments target could be achieved by setting either the appropriate domestic credit target or the corresponding money target. However, since there will always be errors in anticipating such future magnitudes, it becomes important to consider whether there will be greater disruption to the balance of payments resulting from an unanticipated shock to the system when money is controlled or when domestic credit is controlled. To illustrate how greater stability is achieved through controlling domestic credit, consider a situation in which there is an unanticipated decrease in foreign interest rates. The resulting capital inflow increases the money supply when domestic credit is controlled and thus leads to an equilibrating downward movement in the domestic rate of interest. However, when money is controlled, the monetary impact of the capital inflow is sterilized with the result that there is no tendency for the domestic interest rate to fall in line with foreign interest rates. Sterilization effectively insulates the domestic interest rate from foreign rates and perpetuates any incentive for capital to move from one country to another. In general, since an overall balance of payments deficit decreases the money supply when domestic credit is controlled, it will tend to increase the domestic rate of interest and to decrease expenditure. In this way, both the trade account and the capital account 12 will improve. This automatic stabilizing effect will continue until balance of payments equilibrium is achieved. 13
A qualification to the above result is that, whereas controlling the domestic credit extended by the banking system produces greater stability for the net foreign assets of the banking system than is achieved through controlling money, it does not necessarily produce greater stability for the net foreign assets of the central bank. If commercial banks convert domestic assets into foreign assets at the central bank, the net foreign assets of the banking system will remain unchanged with a decline in the net foreign assets of the central bank corresponding to the increase in the net foreign assets of commercial banks. In order to maintain an unchanged level of banking system domestic credit, the central bank must offset the decline in the commercial banks’ domestic assets by an increase in its own domestic assets. In effect, the central bank must sterilize the impact on the money supply of international capital movements undertaken by commercial banks. Since sterilization preserves the incentive for such capital movements, controlling banking system domestic credit may successfully stabilize the overall level of net foreign assets of the banking system but may be accompanied by large changes in the division of net foreign assets between central bank and commercial bank holdings. Similarly, whereas controlling the domestic credit extended by the central bank is an appropriate policy for stabilizing the net foreign assets of the central bank, it is uncertain whether it will have more success in stabilizing the net foreign assets of the banking system than will be achieved by controlling reserve money. 14
Although controlling domestic credit normally brings greater stability to the overall balance of payments than is achieved through controlling money, it may bring less stability for the level of income and the trade balance, especially if international capital flows are highly responsive to interest rate changes (i.e., in countries with open, well-developed financial markets). For example, the capital inflow that accompanies a decrease in foreign interest rates will increase the money supply when domestic credit is controlled and thus increase expenditure and worsen the trade balance. When money rather than domestic credit is controlled, no such adverse reactions occur. However, it should be emphasized that, although controlling the supply of money may stabilize the level of income and the trade balance, it will hinder the achievement of equilibrium in the exchange market and the money market. Changes in the money supply resulting from disequilibrium in the overall balance of payments are an important channel through which the supply of money and the demand for money are brought into equilibrium under fixed exchange rates. If this channel is closed off by the authorities and if the demand for money is prevented from adjusting to supply through interest rate changes owing to international interest rate arbitrage, the authorities may face an unsustainable situation in which the money market cannot achieve equilibrium. 15
The single most important problem associated with controlling domestic credit under fixed rates is that, in itself, it does not ensure that realized rates of monetary expansion will be consistent with the long-term maintenance of fixed rates. The danger exists that money will flow into a country through the balance of payments to accommodate increases in the demand for money caused by potentially inflationary factors, such as excessive government spending and excessive wage increases. For as long as the exchange rate is expected to remain unchanged, price level and expenditure increases are likely to improve the overall balance of payments if international capital movements are highly responsive to interest rate changes. Of course, exchange rate expectations will, in time, reflect the accumulated level of divergence in rates of monetary expansion; money will then begin being exported rather than imported. However, owing to the speed with which money can leave a country whose exchange parity is suspect, the monetary adjustment mechanism may not work smoothly or may not be permitted to work by the authorities and an exchange rate adjustment rather than a money supply adjustment may be required.
In order to remove the danger that monetary developments may not bring smooth trade balance adjustment when domestic credit is controlled, an accompanying policy of “one-way sterilization” has been recommended 16—that is, countries with excessive underlying deficits should sterilize net international capital inflows but should not sterilize net international capital outflows (and vice versa for countries with excessive underlying trade surpluses). 17 The sterilization of capital inflows would ensure that trade balance adjustment was not hindered through unwanted increases in the money supply. Furthermore, sterilization would preserve the incentive for capital inflows and thus encourage a greater capital inflow to finance the trade deficit than would otherwise occur. The nonsterilization of capital outflows would mean that the decrease in the money supply would be at least as great as the trade deficit; nonsterilization of capital outflows would also allow the rate of interest to be affected by capital outflows, thus deterring large-scale unwanted capital outflows. In general, one-way sterilization encourages capital to move from surplus countries to deficit countries but prevents any harmful effects of such capital flows on expenditure adjustment.
A simple operational method of achieving some of the properties of one-way sterilization for deficit countries is to accompany a domestic credit ceiling with a supplementary money supply ceiling. If there were no problems in accurately forecasting trade flows, a supplementary money supply ceiling could be placed at a level such that it would be precisely equivalent to one-way sterilization; the permitted increase in the money supply would be made less than the permitted increase in domestic credit by the size of the trade deficit. If ever the money supply tended to exceed its ceiling while domestic credit was below its ceiling, it would mean that capital was flowing into the country. The sterilization of such capital inflows would prevent undesirable effects on expenditure 18 and would also encourage the continued flow of capital into the deficit country. A supplementary money supply ceiling would have an objective of preventing excessive increases in overall expenditure. For example, it would ensure that excessive government spending would tend to crowd out private spending, and that excessive wage increases would tend to reduce employment.
To the extent that trade flows are not forecast accurately, a supplementary money supply ceiling will not achieve all the advantages of one-way sterilization. The money supply ceiling would not prevent the possibility of the monetary impact of an unanticipated worsening in the trade balance being offset by accompanying capital inflows. It would also limit the monetary impact of an unanticipated improvement in the trade balance. The main problem with supplementary money supply ceilings is the possibility that they could give rise to excessive inflows of capital to deficit countries resulting from greater difficulty in the money market achieving equilibrium when the supply of money is prevented from adjusting to changes in demand. However, this problem seems rather chimerical and certainly more desirable than the alternative of excessive monetary expansion in deficit countries. Apart from this problem, there seems little to be lost and much to be gained by accompanying domestic credit ceilings with appropriate supplementary money supply ceilings. Provided money supply ceilings allowed for a reasonable unanticipated improvement in the trade balance, they could not hamper, but could only foster, a stable adjustment in the trade balance.
It should be emphasized that an appropriate supplementary money supply ceiling is not in itself a deflationary policy that threatens employment and growth in the economy. The money supply ceiling would only transcend the domestic credit ceiling in the event of potentially inflationary large-scale capital inflows. It is the domestic credit ceiling that makes it possible that large-scale outflows of money through the balance of payments may result initially in reductions in employment and output.
The conclusions of this section are pertinent to any country whose currency is pegged, either to a single major currency (regardless of its identity) or to a basket of currencies (regardless of its composition). For such countries, the exchange rate is an exogenous variable, and the supply of money should be allowed to adjust to the demand for money through the nonsterilization of the monetary impact of international reserve changes, except when the monetary impact would have a significant, perverse effect on the trade balance.
The arguments for controlling domestic credit subject to a supplementary money ceiling remain as relevant when the exchange rate is following a predetermined rate of crawl as when the exchange rate is fixed in nominal terms. However, it could be argued that there is some steady rate of crawl of the exchange rate that is consistent with the maintenance of approximate equilibrium in the overall balance of payments and that therefore would not constrain the ability of the country to control money rather than domestic credit. Obviously, the more successful is a country in forecasting the rate of crawl that is consistent with payments equilibrium, the less constraint will be imposed by such a predetermined rate of crawl on the achievement of a money supply target. However, it was explained at the beginning of this section that the distinction between whether money or domestic credit should be controlled is not meaningful in a world in which exogenous variables and structural parameters can be forecast accurately. When there are unanticipated permanent shocks to the system that incorporate a shift in the demand for money, either the money supply must be allowed to change or the rate of crawl must change.
The actual constraint that would be imposed on the achievement of a money supply target by the pursuance of a predetermined rate of crawl would depend upon how wide a margin of variation around the exchange rate path was allowed and how often the rate of crawl was changed. It would obviously be feasible to announce simultaneously predetermined target rates of monetary expansion and exchange rate appreciation, provided there was, by definition, a sufficient allowed margin of variation in one or both. However, this simply recognizes that there is a trade-off between the levels of exogeneity and endogeneity in the money supply and the exchange rate—a topic better discussed in the context of a managed floating exchange rate system.
The analysis and conclusions of this section are applicable both to the determination of whether the domestic credit extended by the central bank or reserve money should be controlled, and to the determination of whether the domestic credit extended by the banking system or the money supply should be controlled. In general, institutional factors are likely to have an important bearing on whether controls are placed on central bank or banking system aggregates. However, from a purely theoretical point of view, there seems to be a presumption in favor of controlling central bank aggregates. This section concludes by briefly presenting some of the theoretical arguments.
(1) Unforeseen changes in the demand for money will have a smaller impact on the balance of payments if the domestic credit extended by the central bank, rather than by the banking system, is controlled. For example, if there is an increase in the demand for money and the domestic credit extended by the banking system is held constant, all of the increase in the demand for money will need to be accommodated by an inflow of money through the balance of payments. However, if the domestic credit extended by the central bank is controlled, the inflow will only need to be a fraction (equal to the reciprocal of the money multiplier) of the increase in the demand for money; most of the increase in the demand for money will be met by credit creation by the commercial banks that is accommodated by the relatively small increase in the foreign asset component of reserve money.
(2) A corollary to the above is that the monetary adjustment mechanism will work faster 19 when the domestic credit extended by the central bank, rather than by the banking system, is controlled. A given balance of payments deficit will result in a decrease in high-powered money when the domestic credit extended by the central bank is held constant and will necessitate a multiple contraction in credit by the banking system. The same balance of payments deficit will result in a reduction in “low-powered money” when the domestic credit extended by the banking system is held constant and, by definition, will not be accompanied by any contraction in bank credit.
(3) International capital movements undertaken by the commercial banks are, by definition, sterilized when the domestic credit extended by the banking system is controlled but are not sterilized when the domestic credit extended by the central bank is controlled. Depending upon the particular circumstances, sterilization or nonsterilization may be preferable. By perpetuating the incentive for such capital flows, sterilization poses the danger of large-scale switches of foreign assets between the central bank and the commercial bank. On the other hand, nonsterilization poses the danger that the commercial banks may, in effect, circumvent the limit on the credit they may obtain from the central bank by obtaining credit overseas. In general, it would be desirable for countries with underlying deficits to sterilize net inflows of capital by the commercial banks but not to sterilize net outflows.
(4) An important determinant of the desirability of controlling the domestic credit extended by the banking system or the central bank is the relative stability and predictability of the velocity of money against the velocity of reserve money. A priori, it is uncertain which is more stable. Since currency has a greater weight in reserve money and is likely to be demanded for transaction purposes rather than asset purposes, this in itself will give greater stability to the velocity of reserve money. However, to the extent that holdings of excess reserves by the commercial banks show large fluctuations, the stability of the velocity of reserve money will be reduced; consequently, this velocity may become less stable than the velocity of money. The question of the relative stability of the velocity of money versus the velocity of reserve money is empirical, and the answer may be expected to vary between countries. 20
III. Domestic Credit and Money Ceilings Under Flexible Exchange Rates
Exchange rates are generally referred to as being “flexible” when there is an absence of a predetermined exchange rate parity or path. At one extreme, a country may give overriding importance to controlling the nominal level of its exchange rate, with the particular rate that it sets being determined in light of actual economic and political developments. At the other extreme, a country may refrain from any large-scale intervention in the exchange market and may set other financial policies without regard for their impact on the exchange rate.
When overriding importance is being given to controlling the exchange rate, the exchange rate is an exogenous policy variable, even though it is “flexible.” When the exchange rate is exogenous, whether fixed or flexible, this presents an argument for controlling domestic credit. Controlling domestic credit endog-enizes the money supply and, at least in the long term, “validates” whatever exchange rate the country chooses to maintain. For example, if a country is maintaining an overvalued 21 exchange rate and has developed a weak reserve position, a ceiling on domestic credit will at least ensure that the money supply will decrease, and will thus help validate the exchange rate in the event that further resources are used to hold up the exchange rate.
Although an endogenous money supply will eventually validate an exogenous exchange rate, there may, in the short term, be important transitional costs. If an overvalued rate is being maintained, a domestic credit ceiling will force a contraction in the money supply that may have a significant, though temporary, effect on output and employment, As a result of such costs, the existence of a domestic credit ceiling may militate against the maintenance of an overvalued rate. To the extent that domestic credit ceilings actually call forth adjustments in overvalued exchange rates, they will facilitate the external adjustment mechanism.
Although the domestic costs of validating an undervalued rate may be acceptable to the country concerned, there are important costs from an international point of view. The money supply increases necessary to validate an undervalued rate represent balance of payments deficits in the rest of the world. It is possible (though with an important caveat) to alleviate this problem by imposing a supplementary money supply ceiling. Institutional considerations in a country may be such that sterilizing the monetary impact of balance of payments surpluses may not be feasible above a certain level, and the country may therefore have to abandon its quest for an undervalued rate in order to remain within its money supply ceiling. If a money supply ceiling were in effect and the actual money supply were at its ceiling, a country could only acquire foreign assets in the exchange market if it were simultaneously to reduce domestic credit. However, large reductions in domestic credit may simply not be possible without short-run increases in interest rates of a size that are unacceptable to the country. Central banks typically set the price of credit they extend but then allow more or less automatic access by the commercial banks (and occasionally by the government) to credit at that price. The financial requirement of the government may well be totally insensitive to the cost of credit, while commercial banks may be willing to accept sharp variations in the cost of funds in the short run in order to retain business. Furthermore, there may be little that can be done to prevent the nonbank private sector from taking advantage of unutilized borrowing facilities. Only if the nonbank private sector and the overseas sector can be induced to use the total proceeds of the balance of payments surplus to acquire public sector debt and/or to reduce their recourse to bank credit will the needed reduction in domestic credit be possible. However, since the imposition of an undervalued exchange rate is likely to increase the price level and may also increase real income, this may in itself result in a tendency for the balance of payments surplus to be invested in the form of money balances and may even increase the demand for bank credit. The greater is the difficulty with sterilization, the greater will be the constraint imposed by a money supply ceiling on the maintenance of an undervalued exchange rate. 22
An important caveat to the above argument is that a money ceiling would have a perverse result in the event that the country incurred no problem in sterilizing and would not have sterilized in the absence of a money ceiling. The lower is the money supply, the more undervalued a given nominal exchange rate will become. Successful sterilization preserves the incentive for money to flow into a country through the balance of payments and, in itself, increases the balance of payments deficit incurred by the rest of the world.
At the opposite extreme from countries that give paramount importance to controlling the exchange rate are countries that give paramount importance to controlling the money supply. In recent years, a growing number of countries have begun publicly announcing predetermined target rates of monetary expansion. If there were certainty that such countries would refrain from future official intervention in the exchange market or would intervene only to achieve a known predetermined target level of net foreign assets, it would make no difference whether a ceiling were placed on money or domestic credit; there would be a predetermined one-to-one relationship between domestic credit and money. 23 However, to the extent that there is uncertainty concerning future intervention, controlling the money supply subject to an overriding ceiling on domestic credit is preferable for two reasons. First, if the money supply is being held on target and domestic credit is subject to a ceiling, there is an implicit floor on net foreign assets. 24 An appropriate domestic credit ceiling could therefore ensure that the pursuit of a money supply target was not accompanied by the maintenance of a significantly overvalued exchange rate. 25
Second, a domestic credit ceiling would be appropriate owing to the possibility that the country would wish to switch at a later date from giving paramount importance to controlling the money supply to giving paramount importance to controlling the exchange rate. For example, the introduction of an aggressive policy of holding up the exchange rate would only be possible if the money supply were allowed to grow below its target rate—that is, if the money supply were allowed to grow at a rate that was consistent with the exchange rate that was being imposed.
When there is reasonable certainty concerning the future magnitude of official intervention, there is a qualification to the statement that it makes no difference whether there is a money supply ceiling or a domestic credit ceiling. If the domestic credit extended by the banking system rather than by the central bank is controlled and the authorities refrain from official intervention in the exchange market, it transpires that an importation of capital by commercial banks will have the paradoxical result of reducing the money supply. Since such an inflow of capital will increase the foreign liabilities of the banking system (i.e., will reduce the net foreign assets of the banking system), it must decrease the money supply if the domestic credit extended by the banking system is to remain unchanged. This money supply decrease is a reflection of the fact that the central bank must decrease its own domestic assets in line with any switches of foreign assets into domestic assets by commercial banks in order for the domestic credit of the whole banking system to remain unchanged. In general, it does not seem necessarily undesirable for capital inflows by commercial banks to reduce the money supply in this way. Under a free float, the exchange rate always adjusts to a level that induces net short-term capital inflows that are sufficient to clear the exchange market. To the extent that an inflow of capital by commercial banks is induced by an unanticipated worsening in the balance of payments, a reduction in the money supply will reinforce the adjustment mechanism that leads to balance of payments equilibrium.
The analysis in this paper has implicitly assumed that, under both fixed and flexible exchange rates, exchange market interventions in the domestic currency are undertaken by the domestic authorities. However, for reserve currency countries, interventions in the domestic currency are undertaken by foreign authorities as well as by the domestic authorities. While it is clear that the money supply should reflect large-scale or protracted interventions by the domestic authorities, it is less clear whether domestic credit should be controlled in such a way that the money supply reflects interventions by foreign authorities. 26
In order that an exchange rate imposed by foreign authorities be validated and external equilibrium achieved, interventions in the domestic currency by foreign authorities must be allowed to change the money supply in the foreign country and/or the reserve currency country. If the money supplies in both countries are allowed to be affected, the attainment of external equilibrium will be faster; any internal costs of adjusting to the exchange rate imposed by the foreign authorities are shared by both countries. If the domestic money supply is insulated from foreign interventions while the foreign money supply is allowed to change, any internal costs of adjustment will be borne primarily by the foreign country.
It seems desirable for the reserve currency country to share in the process of adjustment if it is cyclically out of phase with its main trading partners and has unstable internal conditions that are “calling forth” the foreign interventions. However, if it has stable internal conditions, it does not seem desirable for it to risk its internal stability by sharing in the process of adjustment necessitated by foreign interventions that are the result of unstable internal conditions in foreign countries.
Obviously, it is not possible to define domestic credit in such a way that its control gives rise to money supply changes resulting only from interventions by some foreign authorities at some times. However, there is a theoretically simple way of achieving a suitable division of external adjustment between the reserve currency country and its trading partners. It is for the reserve currency country to control domestic credit defined such that the money supply is insulated from foreign interventions but also to set and adjust the target level of domestic credit with the objective of maintaining or achieving stable internal conditions. If the reserve currency authorities maintain a policy of refraining from active intervention in the exchange market, this policy is equivalent to setting money supply targets with a view to achieving stable internal conditions. In foreign countries, the monetary consequences of foreign interventions will lead toward an equilibrium in which stable internal conditions are shared by the reserve currency country and the foreign countries.
IV. Conclusion
Under fixed exchange rates or any other exchange rate regime where there is a known predetermined exchange rate, the distinction between whether domestic credit or money should be controlled is only meaningful in the context of uncertainty concerning the future levels of exogenous variables and structural parameters. If there were no uncertainty, a given balance of payments target could be achieved by setting either the appropriate domestic credit target or the corresponding money supply target. The arguments for controlling domestic credit rather than money under fixed rates rest essentially on the well-established fact that unanticipated shocks to the system will have a less disruptive short-term effect on the overall balance of payments if domestic credit is controlled.
Under flexible exchange rates, the distinction between whether domestic credit or money should be controlled is only meaningful in the context of uncertainty concerning the future magnitude of official intervention in the exchange market. If there were no uncertainty, there would be a one-to-one relationship between domestic credit and money, and it would make no difference whether a ceiling was placed on domestic credit or money. However, in the inevitable context of uncertainty, the ceiling should be placed on domestic credit. An exchange rate may equally well be used as an exogenous policy variable when the exchange rate is classified as flexible as when it is classified as fixed. The arguments for controlling domestic credit when the exchange rate is exogenous are the same irrespective of the classification of the exchange rate. For example, if a country uses resources to hold up the exchange rate, an appropriate domestic credit ceiling will ensure that the money supply will decrease and will help validate the exchange rate.
Under both fixed and flexible exchange rates, there may be arguments for accompanying a domestic credit ceiling with a supplementary money supply ceiling. Under fixed rates, a supplementary money supply ceiling would limit the danger that potentially inflationary developments would be accommodated by inflows of money through the balance of payments. Under flexible rates, a supplementary money supply ceiling may place a constraint on the maintenance of a significantly undervalued exchange rate. If a country is already pursuing a previously announced target path for the money supply, an appropriate domestic credit ceiling will place a constraint on the maintenance of a significantly overvalued exchange rate.
The conclusions of this paper support the practice of the Fund in using domestic credit ceilings as performance criteria in standby arrangements whatever exchange rate regime is in effect. Furthermore, the conclusions suggest that domestic credit and money ceilings can be determined not only with a view to achieving balance of payments, price level, and income level objectives but also with a view to deterring exchange rate manipulation that may be economically harmful to the individual country or the rest of the world.
Mr. Day was an economist in the External Adjustment Division of the Research Department when this paper was prepared. He is an external graduate of the University of London and received his doctorate from the University of Birmingham.
See G.G. Johnson and Thomas M. Reichmann, “Experience with Stabilization Programs Supported by Stand-By Arrangements in the Upper Credit Tranches, 1973-75” (unpublished, International Monetary Fund, February 28, 1978); and Richard C. Williams, “Evolution of Alternative Forms of Credit Ceilings” (unpublished, International Monetary Fund, May 26., 1971).
“Economic Outlook,” Midland Bank Review (Summer 1978), p. 9.
See the statement of the Shadow European Economic Policy Committee (made jointly by Messrs. Karl Brunner, Allan H. Meltzer, Peter Bernholz, André Fourcans, Michéle Fratianni, Brian Griffiths, Pieter Korteweg, Manfred J.M. Neumann, Michael Parkin, and Jean-Jacques Rosa), reprinted in the U. S Congressional Record, Vol. 123 (June 21, 1977), p. S 10344.
Although the term “domestic credit expansion” is normally used to refer to the banking system as a whole, the majority of stand-by arrangements have placed ceilings on the domestic credit extended by the central bank (or monetary authorities) only. The terms “domestic credit” and “net domestic assets” are synonymous.
Aside from problems concerning revaluations, interest earnings, and retained profits, the flow variable—domestic credit expansion—is simply the change in the stock variable—domestic credit—over the relevant time period.
Lending to the overseas sector is most important in countries, such as the United States and the United Kingdom, that have major international financial centers. In other countries, there is a tendency to include lending to nonresidents as a foreign asset rather than as a domestic asset. From a theoretical point of view, such a practice is not to be recommended, for reasons given later on.
For simplicity only, this section uses primarily “fixed exchange rate language” where credit is assumed to affect the balance of payments rather than the exchange rate.
There is a sense in which credit extended to the overseas sector may be seen as beneficial for the balance of payments. This is when credit is extended for the purpose of financing domestic exports. However, exports financed in this way give rise to no inflow of funds through the exchange market. In the absence of significant imperfections in the domestic and international credit markets, credit extended ostensibly for this purpose is more appropriately viewed as credit extended for the purpose of accommodating an outflow of capital or precluding the need for an inflow of capital—a purpose detrimental to the balance of payments.
It is shown in Section III that, in this context, net capital imports (exports) by commercial banks lead paradoxically to a reduction (an increase) in the money supply.
See, for example, J.J. Polak and Victor Argy, “Credit Policy and the Balance of Payments,” Staff Papers, Vol. 18 (March 1971), pp. 1–24 Manuel Guitian, “Credit Versus Money as an Instrument of Control,” Staff Papers, Vol. 20 (November 1973), pp. 785–800; and William H.L. Day, “A Comparison of Alternative Domestic Credit and Money Aggregates as Instruments of Control” (unpublished, International Monetary Fund, March 16, 1977). The simplest possible monetary model of balance of payments determination treats the demand for money as exogenous. This immediately leads to the result that shifts in the demand for money must be accommodated by either movements in domestic credit or by balance of payments flows. Under such a model, it is futile for the authorities to attempt to control the supply of money. Unless supply is allowed to adjust, shifts in the demand for money will result in persistent money market disequilibrium, persistent balance of payments disequilibrium, and limitless cumulative changes in net foreign assets. Controlling domestic credit is then the only policy option open to the authorities.
A variable is said, in this paper, to be controlled rather than subject to a ceiling when the analysis implicitly assumes that the variable is maintained at a predetermined level. Obviously, a ceiling on a variable per se has no immediate constraining effect if the actual level of the variable is well within the ceiling or if there is an incipient reduction in the variable from its ceiling. In particular, a domestic credit ceiling only precludes sterilization when domestic credit is already at its ceiling and there is a payments deficit.
For simplicity, this paper distinguishes only the trade account and the capital account in the balance of payments. This is to avoid ambiguities that arise when one is considering the current account. For example, it is uncertain whether an incipient decrease in the money supply will improve or worsen the current account; while the visible trade balance will improve, the invisible trade balance may worsen owing to increased interest payments on capital pulled into the country.
Holding domestic credit constant and having an endogenous money supply will not necessarily lead to a steady state in which the trade account is in balance. For example, if new opportunities for profitable investment within a country are continually opening up, or if the public sector deficit is excessive in relation to the private sector’s propensity to save, a steady state could well be reached in which the money supply is constant and the overall balance of payments is in equilibrium, but in which there is a trade deficit that is offset by a continuous inflow of capital.
For a fuller discussion, see Day,op.cit., pp. 17–20.
This problem is likely to be more severe for deficit countries sterilizing outflows rather than inflows (and surplus countries sterilizing inflows rather than outflows). Owing to the asymmetrical nature of the risk of parity changes, it is possible that an increasing stock of capital could flow out of a deficit country without a growing interest rate discrepancy against the country, while a growing interest rate discrepancy in favor of the country would be needed to induce an increasing stock of capital to flow into the country.
See Day, op. cit.
For deficit countries, actual sterilization of capital inflows would mean that domestic credit would be forced to remain below its ceiling or target level; for surplus countries, sterilization of capital outflows would mean that domestic credit would be expanded above its target level.
For countries that could allow more expansion without increasing the rate of inflation if they could only stand the balance of payments effects, money supply increases resulting from capital flows would be desirable up to a certain level and should be allowed for in a money supply ceiling.
It is recognized that “faster” does not necessarily mean more efficiently.
In an empirical study of 40 countries, the velocity of reserve money was found to be more stable than the velocity of money; see J.R. Lothian, “The Demand for High-Powered Money,” American Economic Review, Vol. 66 (March 1976), pp. 56–68.
Overvalued and undervalued exchange rates are assumed in this section to be characterized by a significant and persistent excess demand for, and supply of, the domestic currency. It is recognized that there may be occasions when an exchange rate appears, say, overvalued on this basis but undervalued on the basis of relative price and cost considerations.
A money supply ceiling would not preclude a country from maintaining an undervalued exchange rate through the technique of large-scale intervention in the forward market. The immediate impact of spot purchases of foreign currency on the money supply and reserves can be sterilized by swaps in which foreign currency is simultaneously sold spot and purchased forward. Equiv-alently, an undervalued exchange rate can be imposed without any immediate consequences for the money supply by simply undertaking outright forward purchases of foreign currency.
This statement applies strictly only to a comparison of a central bank domestic credit ceiling and a reserve money ceiling.
Financial programs supported by the use of Fund resources have, in a number of instances, included a “balance of payments test” comprising an explicit floor on net foreign assets. Such floors have been designed to ensure that exchange rate movements would be used to counter excessive balance of payments pressures on the reserves.
In the event that it were deemed appropriate to resist market forces by abnormally large purchases of domestic currency in the exchange market, it would clearly be necessary to increase the domestic credit ceiling. In order to avoid the possibility that forward intervention could be used to maintain an overvalued exchange rate, domestic credit could be defined to equal the money supply less the sum of net foreign assets and net forward foreign currency claims.
Section I described how controlling domestic credit leads to the money supply being affected by foreign interventions when domestic credit is defined in such a way that foreign interventions appear below the line in the balance of payments; the money supply is, by definition, insulated from foreign interventions when they are placed above the line.