7. Bank-by-Bank Credit Ceilings: Issues and Experiences

Tomás Baliño, and Lorena Zamalloa
Published Date:
September 1997
  • ShareShare
Show Summary Details

A number of countries used direct methods of monetary control and, in particular, bank-specific credit ceilings to control monetary and credit aggregates by allocating credit directly among financial intermediaries. These ceilings are usually cast as instructions informing individual banks of the maximum amount of loans allowed in their portfolios. Countries use these ceilings either as a temporary instrument of monetary policy to contain demand shocks in the short run or as a permanent tool to control aggregate credit.

This paper discusses the experience of a sample of these countries by focusing attention on the effectiveness of bank-specific credit ceilings in achieving the assigned objectives (see Table 1). The analysis is conducted against the background of a historical shift from direct methods of monetary control to indirect methods.1 The sample comprises industrial and developing countries, as well as economies in transition.2 Countries were chosen according to the availability of data and the diversity of experiences they encountered, particularly in the design of the credit ceilings.

Table 1.Bank-Specific Credit Ceilings in Selected OECD Countries
Country (period in use)Rules of AllocationInstitutions Subject to Credit CeilingsPenalty in Case of NoncomplianceTrading of Unused MarginsOther Monetary instruments in Use
  • 1964–65
  • 1966–67
  • 1969–71
  • 1974–75
  • 1978
Ceilings established on maximum rates of expansion with allowance made for certain priority categoriesBanks, savings institutions, public credit institutions, and life insurance companiesReduction of banks’ rediscount quota at the central bank by an amount at least equal to the excess over the limitNoMinimum required monetary and security reserves, maximum lending rate of interest
  • 1970–80
Limits on banks’ gross credit, using deposit and credit sharesCommercial, savings, and cooperative banksN.A.NoReserve requirements, rediscount operations
  • 1969–87
Limits set on short-term credit not matched by increase in long-term liabilitiesDeposit banksProgressive special reserve requirementsYesRediscount policy, cash reserve ratio
  • 1973–75
  • 1976–83
  • 1985–87
Separate ceilings on growth rate of short- and long-term creditsDeposit banksN.A.NoInterest rate ceilings
  • 1961–67
  • 1969–72
  • 1977–81
  • 1989–93
Net credit ceilings




Monetary cash reserve arrangement
Deposit banksProgressive penalty that consisted of the imputed cost of excess lendingAs of 1990Open market operations, reserve requirements, rediscount window
  • 1955–58
  • 1961–65
Applied in terms of a maximum percentage increase in credit allowed relative to the same month in the previous year; ceilings also applied to foreign credits extended in foreign currencyCommercial banks and life insurance companiesN.A.NoOpen market operations
  • 1977–89
Applied initially to domestic credit but was later extended to external credit and borrowing in foreign currencyAll banksN.A.NoRediscount facilities, reserve requirements
Calculated based on the structure of bank’s liabilities, with capital and long-term liabilities having higher weights; over time, preferential credit, sight deposits, and time deposits with maturities of less than six months were excluded; also, purchases of T-bills were given relatively high weight so as to favor the development of this instruments
United Kingdom
  • 1964–71
Limits set on growth of credit to low-priority sectors; many exemptions appliedAll banks; gradually broadened to include finance companiesDeposits placed at the central bank equal to the excess of lendingNoOpen market operations, reserve requirements, ceilings on short-term interest rates
  • 1973–79
Limits set on growth of deposit liabilities

The second section of the paper discusses countries’ experiences with credit ceilings. It describes the reasons for adopting credit ceilings, their various design features such as allocation rules, administration of the system, and penalties. The third section points to a number of problems that were associated with the use of credit ceilings, including significant distortions and loss of effectiveness over time. In the long run, with credit rationed through quantitative controls and interest rates often administratively set below market-clearing levels, ceilings discouraged mobilization of financial savings and distorted the flow of funds to the most profitable lending. Such distortions resulted in circumvention of the ceilings and an increase in the importance of informal financial intermediaries.3

By the early 1980s, most member countries of the Organization for Economic Cooperation and Development (OECD) sought to abandon credit ceilings because of these significant distortionary effects. A similar process took place in a number of developing countries.4 However, under conditions of financial market liberalization, initial macroeconomic imbalances, worldwide market integration, or more region-specific structural changes in financial systems, the shift to indirect instruments was often accompanied by large increases in interest rates. At times, such swings in the rates led a number of central banks to reinstate temporarily direct controls. Other countries reinstated controls when their balance of payments deteriorated owing to large liquidity overhangs. Many central banks retained credit ceilings as a backstop during the transition to indirect instruments, while they developed the techniques of monetary control through indirect instruments.5 By the end of 1994, 16 countries were using credit ceilings as an instrument of monetary control.6

In certain circumstances, such as when monetary relationships are highly unstable (e.g., during the transition to a framework of indirect control), credit ceilings can be a practical way to directly target net domestic assets of the banking system. The practicality of the instrument may of course be offset by administrative costs, disintermediation, and inefficiency. This paper identifies a series of principles that countries may follow in designing a system that can minimize the drawbacks associated with credit ceilings, including, among others, trading of unused credit margins.

Experiences in the Use of Credit Ceilings

Country experiences with the use of credit ceilings varied greatly in terms of the authorities’ objectives and the specific design of the instrument.

Reasons for Adopting Credit Ceilings

The sample countries adopted credit ceilings because this instrument (1) appeared easy to implement; (2) was perceived to deliver monetary and credit targets with accuracy; (3) was thought to be the most effective monetary instrument, given the underdeveloped financial markets and/or limited technical capabilities of the monetary authorities; and (4) was seen as the best tool through which resources could be diverted to the most “favored” sectors, such as the agricultural sector.

Notwithstanding the varying degrees of development and sophistication of financial markets, a closer look at the countries that used credit ceilings reveals some common features, including highly regulated financial markets (except in the United Kingdom), noncompetitive banking structures, underdeveloped markets for government securities, low propensity of the private sector to hold financial assets owing to interest rate controls, high transactions costs, and restrictions on capital outflows.

Often, the noncompetitive banking systems were characterized by high intermediation costs and large interest rate spreads, which hampered the efficient transmittal of the monetary policy to the real economy. They also impeded the efficient functioning of open market operations. Moreover, in countries with a small number of commercial banks, the oligopolistic nature of the financial system prevented the running of efficient auctions of government or central bank bills by limiting the number of market participants.

In addition to the above reasons, a number of OECD countries employed credit ceilings on other grounds (see Table 2). Some countries used credit ceilings essentially as an emergency instrument during periods of overheating and suspended them immediately afterward (Belgium, Denmark, Netherlands).

Table 2.Bank-Specific Credit Ceilings in Selected Developing Countries
Country (period in use)Rules of AllocationInstitutions Subject to Credit CeilingsPenalty in Case of NoncomplianceTrading of Unused MarginsOther Monetary Instruments in Use

A ceiling of 3 percent a quarter on the growth of credit in Egyptian pounds to the private sector for commercial purposes; a ceiling of 2.5 percent a quarter on credit in Egyptian pounds to householdsCommercial bankN.A.NoneLoan-to-deposit ratio amounting to 65 percent of Egyptian pound deposits, interest rate ceilings, reserve requirements, selective credit ceilings
1987–88Applied to both domestic and foreign currency lendingAll financial institutions, including investment and specialized banksIncluded holding of excess reserves with central bank of an equivalent amount and period

May 1988–

Sept. 1988
Banks’ outstanding credit at the end of previous period plus allowed percentage growthTwo deposit banksNoNoneInterest rate ceilings, reserve requirements

Later changed to four indigenous banks

Banks’ outstanding credit at the end of previous quarter plus allowed percentage growthCommercial banksNoNoneCash and liquid asset reserve requirements, refinance windows, selective credit ceilings, savings deposit floor rate, open-market-type operations (after 1985)
1989–91Required increase in banks’ credit was kept at no more than 5 percent of outstanding amount in each 90 daysAll financial institutions; many exemptions appliedPenalty of ⅙ of 1 percent a day on balances in excess of limits

Bank’s outstanding credit at the end of previous quarter plus allowed percentage growthCommercial banksAs of 1990, penalty included placement of a non-interest-bearing deposit at the central bank equivalent to 20 percent of excess credit extendedNoneMaximum lending rate and minimum savings deposit rate


Dec. 1995
Up to 1990, ceilings were set quarterly based on banks’ performance in terms of sources and uses of funds over the preceding three years and projections for the coming year; as of 1990, ceilings were based on banks’ increase in deposit share; in 1994, allocation rule was changed so that base would include volume of banks’ capital, reserves, and increases in depositsCommercial banksSince 1987As of March 1993Credit authorization, reserve requirements, rediscounting

Other countries, such as France (1969–87), Italy (1970s and early 1980s), Portugal (1977–89), and the United Kingdom (1960s and early 1970s), adopted credit ceilings as one of their main instruments of monetary policy for longer periods of time. In France (1969–87), interest rate changes were viewed as an ineffective tool for managing monetary conditions because of the perceived inelasticity of credit demand to short-term interest rates. Italy and Portugal used credit ceilings either as a quasi-fiscal means of channeling resources to priority sectors or to finance the external current account deficit. Italy placed ceilings on credit extended in domestic currency to encourage foreign borrowing. This was especially important in the early 1980s when the balance of payments had deteriorated considerably. By the early 1960s, the authorities in the United Kingdom had introduced credit ceilings in response to a deteriorating external balance. Ceilings were also used to channel credit to export sectors, state enterprises, and government debt.

A number of developing countries also used credit ceilings to allocate credit to “favored” sectors at subsidized interest rates (Egypt during the early 1980s). Others used this instrument to meet targets set on the net domestic assets of the banking system in the context of programs supported through the use of IMF resources. In these cases, controls were placed to address particularly severe monetary disturbances and adverse macroeconomic developments (Jamaica, during the 1980s). Some countries with rudimentary financial systems opted for credit ceilings as their only feasible instrument until their legal and institutional framework for market-oriented instruments was further developed (Kenya, 1987–90; Guyana, 1988–91; and Madagascar, 1985–95).

In the 1990s, a few economies in transition, mostly in Eastern Europe, used credit ceilings to control monetary and credit aggregates during the transitional stage from centrally planned to market-based systems (Albania, Bulgaria, Czechoslovakia, Poland, Romania, and the former Yugoslav Republic of Macedonia). In these economies, ceilings were often seen as an alternative to the immediate shift to more market-oriented instruments of monetary policy and reliance on reserve money management. Making such a shift at that time could have posed a high risk of loss of control because of the unknown transmission mechanisms. In particular, money multipliers were showing signs of erratic behavior because of structural breakdowns. Also, frequent changes in reserve requirements (one of the few immediately available instruments) would have meant undesirable large adjustments in bank reserves. Finally, large liquidity overhangs limited the effectiveness of monetary control based on reserve money management. Credit ceilings allowed credit and monetary expansion to maintain a pace consistent with macroeconomic stabilization, until they could be replaced with market-based instruments of monetary control.

Most of these countries adopted credit ceilings in combination with a number of other indirect instruments. For example, Albania employed reserve and liquidity requirement ratios. The authorities in the Czech Republic relied on auctions of refinance credit along with reserve requirements while using credit ceilings as a backstop. The National Bank of Poland used credit refinancing facilities, auctions of central bank bills, and treasury bills. In Bulgaria, the system of monetary control consisted of credit ceilings, nonremunerated reserve requirements, central bank refinance, and limited open market operations.

Rules of Allocation

Notwithstanding countries’ initial reasons for employing credit ceilings, the rules of credit allocation varied according to the specific priorities and the institutional environment of each country. In general, the central banks set periodic ceilings on net domestic assets of the banking system—with a subceiling on credit to government, consistent with ultimate policy objectives for balance of payments, inflation, and economic growth. A global ceiling on credits to the economy was then derived as a residual. The base used for credit ceilings and the weights allocated to each covered institution were, however, more country specific.

Typically, in calculating the weights allocated to individual institutions, each period, the authorities permitted a maximum percentage increase of bank advances in relation to a certain base period. Initially, in some cases, banks that granted preferential credit to certain sectors were given a higher share of the global credit ceilings (Portugal), but over time, most countries treated all banks and types of credit equally to minimize distortions that were arising owing to preferential treatment of some institutions.

Tables 13 provide an overview of the various allocation rules in different countries. While some countries based credit ceilings on gross credit, others used only the credit financed with monetary liabilities (the so-called net credit ceilings) as the base for credit ceilings (France and Netherlands).7 Portugal introduced modifications to the base as new financial instruments, including treasury bills and certificates of deposits, were introduced in the market. It also included in the base loans extended in foreign currency and borrowing from abroad. France and the United Kingdom exempted loans to sectors that the authorities wished to support, such as the export sector.

Table 3.Bank-Specific Credit Ceilings in Selected Eastern European Countries
Country (period in use)Rules of AllocationInstitutions Subject to Credit CeilingsPenalty in Case

of Noncompliance
Trading of Unused MarginsOther Monetary Instruments in Use
  • Since 1992
Ceilings are set based on the share of credit outstanding, new lending and repayment, nonperforming loans, and increase in depositsCommercial banksBanks required to pay fine on excess lending equal to one period’s interest on spread between maximum lending rate and minimum deposit rate, adjusted for reserve requirements, plus 0.5 percent of the excessYesReserve and liquidity requirement, refinancing window
  • 1990–July 1994
As of 1993, ceilings set monthly based on changes in relative share of deposits, size of bank’s capital, relative liquidity of bank, and riskiness of portfolioCommercial banksBanks required to deposit amount equivalent to excess credit in a non-interest-bearing account with central bank50 percent of unused margins are allowed to be tradedReserve requirements, open-market-type operations, refinance credit facilities
Czech Republic
  • 1990–92
Ceilings set on quarterly basisCommercial banksNoNoAuction of refinance credit, reserve requirements, open market operations
Macedonia, former Yugoslav Republic of
  • Since 1991
Ceilings are set monthly based on the relative growth in bank’s deposit, paid-in capital and general reserves, share of credits outstanding. and riskiness of portfolioCommercial banksBased on discount rateNoReserve requirements, refinancing facilities, special liquidity credits, auction of deposits and central bank bills
  • 1989–92
Ceilings were set based on bank’s outstanding share of credit at the end of previous periodCommercial banks NoNoReserve requirements, weekly auctions of T-bills. credit refinancing facilities
  • 1990–91
Ceilings were set quarterly based on bank’s outstanding share of credit at the end of previous periodCommercial banks NoNoMaximum spread of 3 percent allowed between deposits and lending rates, refinance policy

Between 1973 and 1984, the Bank of France excluded long-term resources for bank lending from credit growth. Thus, banks could expand their lending activities beyond the ceilings, provided that these credits were financed with new capital funds or through the issuance of long-term bonds. The rationale for such an arrangement was that at times when the ceilings were binding because of excess demand for credit, banks could fund additional lending by issuing bonds. This would raise interest rates, leading to a reduction in demand for credit and restoration of the equilibrium (Quintyn, 1991).

Between 1973 and 1984, the Bank of France set an annual growth target on the stock of credit. In 1973 and 1974, the annual growth rate was based on the credit portfolio of the previous year, using a moving base. Monthly growth rates were then derived from this annual growth rate. As of 1975, the targets were calculated with respect to an index of 100, corresponding to the credit stock at the end of 1994 (Quintyn, 1991).

By the mid-1980s, however, these direct credit controls were no longer binding in the aggregate because some banks had accumulated a large amount of unused credit that could be carried over. Eventually, difficulties in forecasting these accumulated credit margins contributed to the abandonment of the system. By the end of 1986, the authorities had completely removed quantitative credit controls and begun managing monetary policy through the indirect control of short-term market interest rates.

In the 1960s, the Netherlands also adopted a system of “net credit ceilings.” Under this system, no limits were set on the banks’ total lending as long as the proportion of lending not matched by long-term funding did not exceed an envisaged growth rate. Accordingly, only credits financed with short-term bank deposits were subject to direct restrictions. As in France, the advantage was that banks could finance their credit by issuing long-term paper or mobilizing long-term savings deposits. Such a system aimed at encouraging the development and deepening of the market for long-term financial instruments. Except for brief interruptions, direct credit restrictions were in force from June 1961 until March 1972, and from May 1977 until early 1982. In early 1986, the Netherlands again imposed the ceilings through an informal agreement, under which the banking system was to limit the two-year growth of its net money-creating operations to 11–12 percent.

Faced with an increase in net lending by banks and pressures on the exchange rate peg, in 1989, the Dutch authorities introduced a new instrument, the “monetary cash reserve arrangement.” This instrument superseded the credit ceilings and sought to remedy the latter’s shortcomings, which stemmed from its non-market-oriented nature. Like credit ceilings, the monetary cash reserve arrangement sought to curb the banks’ net money-creating operations. Where credit growth exceeded a certain permitted target, banks had to pay an interest rate equal to the interest that would have been charged if compulsory reserve requirements had been in place (the so-called cash reserve percentage). This way, a price was paid for lending in excess of permitted volume.

Such a system had several advantages over the previous system of credit ceilings. First, unlike the latter, it was market oriented, since it acted not on the volume of lending financed from short-term funds, but on its price. It raised the price of excessive lending financed from short-term funds. Second, it did not curtail interbank competition. Each bank could expand its lending to any desired level, provided that a price was paid. Third, it encouraged the development of the interbank market, since banks could trade unused margins of credit ceilings with one another (see section Trading of Credit Ceilings).8

In a number of developing countries, the allocation of credit ceilings among banks was based on each bank’s outstanding credit at the end of the previous period (Egypt, Guyana, Jamaica, and Kenya). However, at times, such an allocation rule gave banks an incentive to breach the ceilings even at the cost of severe penalties in order to obtain higher credits in the next period. Also, such an allocation rule often limited bank competition by discouraging them from mobilizing deposits. This prompted a number of countries to change the base of the allocation rule as they gained experience with the implementation of ceilings. For example, in Madagascar, the authorities used to allocate quarterly global credit ceilings to individual banks based on banks’ credit and deposit market shares over the preceding three-year period and their projections for the forthcoming year. In November 1990, the authorities changed the base to one where credit ceilings were based on the increase in deposit share of each bank (and for new banks on their expected deposit share). In addition, they adjusted ceilings to take into account the evaluation of each bank’s needs. Further, in 1994, they began to base the ceilings on the volume of each bank’s capital, reserves, and increases in deposits. Albania changed the original allocation rule from one where ceilings were based on a share of a bank’s deposits to one that combined a bank’s share of credits outstanding, new lending and repayment, and increases in a bank’s deposits. Bulgaria initially set quarterly credit ceilings based on a bank’s share of outstanding credit. However, significant lags in reporting and slippage on the part of the commercial banks led to changes in the allocation rule. The new rules took into account variations in banks’ share of deposits, size of banks’ capital, and a qualitative assessment of relative liquidity and riskiness of banks’ portfolios. The new ceilings were set monthly.

Administration of Credit Ceilings

Monitoring and Circumvention

Central banks had to measure compliance frequently for signs of circumvention of the ceilings. A number of countries shifted from end-of-quarter to monthly monitoring to better observe deviations from the targets (Albania and Egypt). At times, however, monitoring of the ceilings was complicated because of the long lags in reporting from the commercial banks to the central banks (Albania).

In some countries, the authorities determined compliance with ceilings by comparing the ceiling with the outcome on a particular date (Portugal). However, that allowed banks to engage in “window dressing” to comply with the ceilings.9 At times, banks hid credit extended in off-balance-sheet accounts just before the reporting date and used various accounting ploys to conceal the true volume of credit outstanding on that particular date.

In almost all cases, prolonged use of credit ceilings led to evasion and circumvention. In those countries where borrowing in foreign currency was exempted from the base, banks circumvented ceilings by increasing foreign currency credits in total loans. In Jamaica, banks circumvented credit ceilings by developing new ways to extend credit, which included classifying certain credit items as contingent liabilities to avoid an expansion in the assets shown on their balance sheets. Often, having reached their credit limits, banks charged brokerage fees to match depositors with borrowers directly. Also, they shifted loans and deposits to nonbank affiliates, often stimulated in addition by differential reserve and capital requirements on nonbanks (see Marston, 1995). In Bulgaria, banks that sold government securities to nonbanks were allowed to increase credits by an identical amount. Often, however, banks circumvented the ceilings by selling government securities just before the test date and buying them back later.

In many cases, circumvention of ceilings and/or overall loss of competitive advantage by banks forced the authorities to extend the ceilings to instruments or institutions that previously were exempted (Egypt, Portugal, United Kingdom). For example, in the United Kingdom, larger borrowers who were denied credit at their commercial banks were able to discount their commercial bills through brokered channels or turn to the capital markets or finance houses instead. As a result, finance houses grew rapidly at the expense of deposit banks, which were subject to these restrictions. To deal with this problem, the Bank of England extended the ceilings to include all financial intermediaries as well as the capital market (by imposing controls on the issuance of securities) to prevent further deterioration of banks’ market shares. Eventually, all important channels for lending and borrowing activities of credit institutions were put under the supervision and control of the authorities. In Belgium, exchange controls, a two-tier foreign exchange market, and regulations imposed on the net foreign position of commercial banks helped reduce evasion of credit ceilings through foreign capital markets. In Italy, the authorities brought the credit supply of nonbank financial institutions under control by imposing an upper limit on the securities’ yields issued by these institutions.


In most countries, credit ceilings were effective only if the authorities established and preannounced stiff penalties for noncompliance. Typically, these penalties depended on the amount of excess credit extended and on market interest rates to reduce the profitability of exceeding individual credit ceilings for the noncompliant bank.

Penalties varied from one country to another, ranging from purely monetary fines equivalent to the excess of ceilings (Egypt), to a less biting form that included a reduction in the bank’s rediscount quota at the central bank by an amount at least equal to the excess credit growth (Belgium and Bulgaria).10

Generally, penalties took two forms, depending on the degree of monetary control that was to be achieved through credit ceilings. They were prohibitive, discouraging overshooting altogether with a view to adhering as closely as possible to the overall monetary targets. High penalties were particularly useful if the central bank wished to contain bank lending to residents during episodes of capital inflows and to nonresidents during episodes of speculative attacks.

Alternatively, penalties could be progressive, which allowed greater flexibility in the volume of credit because some loans could be profitable enough to offset the cost of the penalty. Because progressive penalties did not aim at discouraging overshooting altogether, they also introduced flexibility in the enforcement of credit ceilings on particular banks. Since 1973, France has applied a progressive reserve requirements ratio to the nonobservant bank’s total outstanding amount of lending to the private sector.11 Noncompliant banks were required to deposit these non-interest-bearing “supplementary reserves” in a special account with the central bank. The Dutch monetary cash reserve arrangement was also a system of progressive penalties whereby fines were levied that consisted of the imputed cost of the reserve requirements, that is, lending in excess of ceiling times a penalty interest rate. The penalty was calculated each month, with undershooting months allowed to offset overshooting months over a 12-month period. At the end of the 12-month period, the penalty was levied if it exceeded a pre-announced small amount. The preannounced amount was designed to shelter the smaller banks from the penalties, and thus encouraged competition in the banking sector.

In Madagascar, to ensure that the overall limit was not broken, the central bank allocated less than the overall ceilings. At times, however, such a strategy led to unnecessary credit constraints without, at other times, preventing the overall limit from being exceeded. Penalties for exceeding the ceilings were related to the frequency of violations. First-time offenders were given a warning for the first violation, whereas repeat offenders were required to make a non-interest-bearing deposit with the central bank equivalent to 50 percent of the excess credit for the second infraction until the next test date.

Alternatively, some countries used moral suasion to enforce credit ceilings. The sole use of moral suasion, however, was unsuccessful in containing aggregate demand because some banks increased their credit above the targeted limits (Kenya). Thus, a number of countries combined moral suasion with penalties. In France, for example, moral suasion was backed by the supplementary reserve requirements.

Trading of Credit Ceilings

A number of countries allowed banks to trade credit rights among themselves to minimize distortions created by the administrative allocation of credits. Banks that wanted more credit than permitted under the ceilings acquired credit rights from elsewhere in the industry. The trading of credit rights allowed for the outcome of banks’ market shares at the end of the period to be closer to the one that would have prevailed under pure market conditions.

In the Netherlands, for example, under the monetary cash reserve system, a fast-growing bank could buy an unused margin from a bank whose net lending growth remained below the allowed ceiling. The development of a market for unused quotas helped reduce the circumvention of ceilings. Moreover, the incentive to extend lending to the economy was affected by the price of unused margin in the market for credit. This mitigated the constraints that the initial allocation put on banks that could profitably expand their share of the credit market faster than the ceiling. However, it did not eliminate the windfall profit to banks whose ceiling implied a market share higher than they would have achieved under pure market conditions.

In July 1992, Bulgaria began to allow interbank trading of up to 50 percent of the unused portion of a bank’s credit limit. However, over time, some banks appeared to be always sellers and others, buyers. Also, because ceilings were monitored monthly, the selling banks waited until the end of the month to carry out trades, leaving the buyers uncertain about the status of their portfolio.

In France, banks with unused room under the ceiling began trading credits with other banks that were in danger of exceeding their credits. Banks could also trade up to six months’ worth of unused past credit.

Effectiveness of Credit Ceilings

In all countries reviewed, monetary control through the imposition of ceilings—often combined with other administrative controls—left the banking industry highly uncompetitive, inhibited growth of bank financial intermediaries, and led to a large buildup of excess reserves, often caused by the monetary financing of the budget deficit. The problem was further complicated in countries where ceilings had to be applied to a large number of financial institutions. Such a task often led to significant administrative difficulties.

In general, while credit ceilings were effective in containing credit aggregates in the short run, they became less effective in the longer run as borrowers substituted away to unregulated channels of credit financing. In most cases, since banks had to limit their credit expansion according to some arbitrary allocation rule, the most expansive banks were progressively put at a disadvantage as changes in the bank’s market share for credit were in effect frozen. In fact, in almost all countries where credit ceilings were put in place for prolonged periods, the demand for credit tended to develop alternative channels of financing. The main advantages and disadvantages of credit ceilings are summarized in Table 4.

Table 4.Credit Ceilings: Advantages and Disadvantages
Perceived to be reliable in controlling credit aggregatesCannot be used for fine-tuning short-term liquidity management and are effective only as long as they are binding
May be the only feasible monetary instrument available in countries with severe market failures or with very thin financial marketsTend to perpetuate credit market shares, independent of a bank’s competitiveness
Relatively easy to implement when the number of banks is limitedCause disintermediation from regulated financial sector to unregulated domestic markets or foreign markets (transfer of savings overseas)
Relatively easy to link to a monetary programming formalLimit competition among banks and hamper the development of a money market
Unaffected by the exchange rate regimeMay lead to overhang of liquidity if monetary policy is loose
Cause micromanagement of credit allocation by the central bank
Difficult to monitor since credit may be extended in forms not covered by ceilings
May lead to multiplication of direct controls and financial repression
Difficult to enforce and create administrative difficulties in countries with a large banking system

In the United Kingdom, for example, under binding credit ceilings, banks invested their excess liquidity in government papers or lent mainly to those sectors excluded from the ceilings. As a result, in some cases, the quality of projects financed progressively weakened owing to the dissociation between lending rates and the quality of borrowers, to the point where real returns became negative on some investments. By the early 1970s, dissatisfaction with the deadweight efficiency losses resulting from a directly controlled financial system grew, and disintermediation rendered the direct controls ineffective. Thus, in 1971, the authorities dismantled all credit controls.12

In most countries, the number of exemptions introduced in the system complicated administrative control. Several methods were generally used in this respect: partial or total exemptions from the ceilings, the fixing of higher ceilings for specific credit categories, and the upward adjustment of the individual ceiling to include credits already granted. In Bulgaria, the number of loopholes and exemptions led to the abandonment of the credit ceilings altogether in 1994. The primary difficulty was that credits extended in foreign currency were not included. With the growth in foreign currency deposits, there was a parallel growth in foreign exchange credits. Also, agricultural credit and credit extended by newly established banks were exempted from credit ceilings.

In France, insufficient coverage of financial institutions, combined with a significant number of exemptions and complexities based on the categories and types of lenders and borrowers, seriously jeopardized the effectiveness of credit ceilings. In addition to the growing number of exemptions, smaller and newly established banks were allocated looser ceilings, while, in general, ceilings for larger banks were tightened. The proportion of the amount of credit exempted from the ceilings rose from 1.7 percent of total bank lending in 1973 to 7 percent in 1978 and nearly 11 percent in 1983. These measures increasingly led individual banks to request special waivers when they experienced difficulties, introducing additional arbitrariness into the system. As mentioned above, the carryover of unused credits also undermined the credit ceilings’ effectiveness. Indeed, to add flexibility to the system, banks were allowed to carry forward, up to a maximum of six months, unused credit ceilings. Eventually, however, difficulties in forecasting the use of accumulated credit margins contributed to the abandonment of the system altogether.

Exemptions also complicated monetary programming, because it was necessary to project the path of the exemptions in order to derive the credit subjected to ceilings as a residual. The result was a reduction in the usefulness of credit ceilings as an operational target for monetary policy (Quintyn, 1991).

Design Features of a More Efficient System of Credit Ceilings

Although credit ceilings tend to be cumbersome and a fairly blunt monetary instrument, a number of central banks still use them. To prevent loss of monetary control, some countries use the ceilings during their transition from a centrally planned to a market economy until the financial markets are sufficiently developed for full reliance on indirect monetary instruments. Others use credit ceilings as a backstop during the transition from direct to indirect instruments, while the authorities develop the techniques of monetary control through indirect means. Still, other countries keep the ceilings in reserve, in case of difficulties.

In certain circumstances, such as when monetary relationships are highly unstable, which makes the demand for money too volatile for monetary management through indirect instruments, the ceilings can temporarily be used as a pragmatic way to target net domestic assets of the banking system directly. However, credit ceilings need to be carefully designed to minimize the many distortions discussed above.

Appropriately tight reserve money is a prerequisite for the effectiveness of credit ceilings if and when they are used: the looser the central bank’s reserve money, the greater the liquidity of the banking system, and the greater the incentives for all banks to circumvent credit ceilings, no matter how the instrument is designed. This being said, there may nevertheless be incentives for particular banks to circumvent the system, if the methods of apportioning credit lead over time to allocations that are at odds with the competitive evolution of banks’ market shares. In these cases, the imposition of an overall credit target will tend to affect certain banks more than others. Therefore, the former will be inclined to circumvent their ceilings, whereas the latter will enjoy rents.

To make credit ceilings more effective, monetary control must be market oriented. Thus, the ceilings should incorporate features that enhance the flexibility of credit allocation. Allocation methods can be designed to minimize distortions of the credit ceilings. To this end, the Netherlands’ approach to setting rules of allocation may be useful; the base can be a net domestic credit aggregate; the weighing rule can be a flow method, which would allocate the change in net domestic credit of the banking system. The flow method reflects as closely as possible the banks’ competitive evolution of market shares. Competition is also enhanced through uniformity of treatment and transparency of rules.13 The flow approach introduces some flexibility by encouraging banks to compete for deposits and accumulate capital.14 Finally, to preserve the integrity of the credit ceiling as an instrument of monetary control, financial support for “priority” sectors that the authorities wish to support should explicitly flow through the budget.

The design of penalties for overshooting a credit ceiling is a crucial element. In particular, progressive penalties introduce a useful element of flexibility in enforcing the ceilings. In effect, such penalties allow for a range of outcomes; the upper limit of the range corresponds to the desired outcome, at which point prohibitive penalties would be assessed. The central bank can also modulate the progress of the penalties from period to period to signal monetary policy intentions.

A progressive penalty may consist of a reserve requirement proportional to credit extended above the ceiling. However, such a constantly changing reserve requirement may complicate monetary management, especially as it interferes with the money market. A simpler system of progressive penalties would levy fines that consist of the imputed cost of the reserve requirement, that is, lending that exceeds the ceiling times a penalty interest rate. In monetary control frameworks that rely on indirect instruments, if those instruments coexist with (nonbinding) credit ceilings, penalties should become prohibitive to block credit expansion in those episodes where the indirect instruments fail.

Alternatively, in certain cases where a bank repeatedly overshoots its credit ceiling by large amounts, penalties may be imposed on management or, in an extreme case, the bank license may be withdrawn. These actions would tend to ensure compliance with the central bank’s guidelines.

In the simplest possible framework, compliance with ceilings is determined by comparing the ceilings with the outcome on the prescribed test date. However, verification of compliance on a particular day may sometimes cause certain banks to manipulate the outcome, that is, to reduce their declared outstanding credit through accounting gimmicks.

Credit extended just before the test date may be temporarily hidden in a suspense account, or banks may ask clients to “repay” credits with deposits on the test date, the operation being reversed thereafter. If these practices tend to be widespread and cannot be readily deterred through the threat of penalties, some form of averaging of outstanding credit subject to ceilings may be needed to ensure better implementation of the instrument. In periods of high inflation, it may be necessary to take the average of the series, for example, during the last month of the quarter, if quarterly test dates are used.

The averaging method that was used in the Netherlands’ monetary cash reserve arrangement goes one step further in that it assesses compliance against a moving average of outcomes. A series of averages of net lending is computed for the most recent six months; the year-on-year growth rates of these averages are compared with the desired growth rate of net lending—the ceiling. An excess in a particular month is assessed a penalty; otherwise, a credit is given. Finally, these monthly penalties are cumulated over 12 months, giving the net penalty. The advantage of this system is to provide flexibility to banks in the month-to-month management of their credit activity, and therefore to minimize the impact of credit ceilings on the banks’ business. However, the cost of this flexibility is relatively looser control by the central bank over credit conditions.

The treatment of credit rights is an important dimension of market-based monetary control under a system of credit ceilings. Banks that want to grant more credit than permitted under the ceilings may acquire credit rights from elsewhere in the industry. This mitigates the constraints that the initial allocation puts on banks that can profitably expand their share of the credit market faster than the ceiling. As explained above, however, trading does not eliminate the windfall profit to banks that obtain a ceiling higher than the one they would have achieved under pure market conditions. Banks would need to register the transfer of unused credit quotas with the central bank. This is particularly important for the flow method, as the treatment of base drift needs to take the trading of ceilings into account (see Appendix I).15

It is possible to allow banks with unused credit margins to increase their ceilings by the amount of the margin during a given number of periods, after which the margins still unused are lost. However, a system that allows unused credit rights to be accumulated or traded for use outside the current period would require projection of these unused margins. Therefore, for administrative simplicity, it is preferable to allow for the use and trading of credit ceilings within the current program period only.

Complete market determination of credit ceilings can in principle be achieved if, in addition to post-allocation trading of ceilings, the initial rule-based allocation is replaced by an auction of credit rights. Such an auction is conceptually identical to auctions of pollution rights or telecommunications frequencies. Banks would bid cash amounts for the right to expand credit by tranches of, say, 100 during the period (month or quarter). The more efficient banks would have the lowest costs and would tend to make the highest bids. Clearly, prudential safeguards need to be put in place to minimize distress bidding, but such safeguards are in principle similar to the one used in the operation of other monetary instruments, such as central bank credit auctions.16 Thus, flexibility can in principle be achieved if the credit ceilings are allocated through a periodic auction of credit rights, supplemented by post-auction trading of unused margins, thereby resulting in a system of credit ceilings that is fully market determined. The auction system would theoretically eliminate all rents arising from administrative allocation methods.


The country experiences provided an overview of the range of practices and various designs that have characterized credit ceilings. While the instrument is inherently a crude and inflexible method of monetary control, its specific design and enforcement have, in many cases, led to its ineffectiveness. Suboptimal specification of rules of allocation tended to ossify the distribution of credit by perpetuating market shares and limiting the incentives for banks whose ceilings were binding to finance profitable projects. Additionally, inadequate monitoring and enforcement of penalties by the authorities often led to circumvention and ineffectiveness of the instrument. Insofar as credit ceilings were effective, they led to financial repression and disintermediation.

Credit ceilings may serve a useful purpose in certain circumstances, such as when monetary relationships are highly unstable or during the transitional period until the financial markets are sufficiently developed for full reliance on indirect monetary instruments.

In those circumstances, the proper design of credit ceilings tends to reduce the distortions they usually entail, albeit without eliminating them completely. Over time, as the market orientation of monetary control is established, the country can rely increasingly on indirect instruments (such as credit or deposit auctions, operations with government securities, standing facilities) to achieve its monetary and inflation targets. At such a stage of development of market-based monetary control, the role of credit ceilings should be substantially diminished. However, the instrument could conceivably be maintained in the central bank’s arsenal as a fallback tool, to be used only when indirect methods falter because of uncertainties in monetary relations or when indirect methods need to be temporarily supplemented in order to contain the impact on credit of surges in capital flows.

Appendix I.
Illustrative Accounting Framework and Choice of System Parameters

As discussed in the main body of the paper, when monetary relationships are highly unstable (e.g., during the transition to a framework of indirect control over broad money or episodes of banking distress), credit ceilings can be a practical instrument to directly target net domestic assets of the banking system on a temporary basis. In those cases, certain design features of the instrument can minimize the distortions associated with its use. This appendix describes an accounting method and administrative procedures that would make credit ceilings more effective.

Choice of the Base

For effective targeting, the base used for credit ceilings should be an aggregate as close as possible to the operational target, which is assumed to be net domestic assets of the banking system. This principle implies that a net credit aggregate should be used, such as total credit (including nonperforming loans) minus capital and provisions. The use of this net aggregate allows banks to extend more credit than under a gross ceiling if such credit is matched by shareholder funds.

Choice of the Weighting Rule

In addition to the choice of a base in a reference period, a weighting rule (weights for each covered institution) needs to be determined. It needs to be designed carefully because of the sensitivity, and arbitrariness, of individual bank credit allocations associated with particular weighting rules. The stock and flow weighting rules are defined below and their outcomes compared. It turns out that the stock method results in less stable outcomes when underlying parameters shift. The stock method also imposes constraints on the choice of system parameters so that certain banks would not have to recall loans. These issues are illustrated in Appendix II, which also shows how to recalculate the base when credit ceilings can be traded and how to calculate the penalties.

Stock Method

Let the base be measured in period t as

PCt+1 is the projected level in period t+1 of overall net domestic credit of the banking system as defined above.17 Then, g t is the targeted growth rate of credit from t to t+1; Ci, t is outstanding net credit of bank i at the end of the base period; and PC i, t+1 is the next period’s ceiling on net credits of bank i.A stock method would allocate the stock PCt+1 as follows:

This is equivalent to allocating PCt+1 according to each bank’s share of the credit market:

A shortcoming often found with this method is that it does not reward banks that perform comparatively better in the deposit market. C excludes capital and provisions. Let us therefore consider allocating PCt+1according to a mix of the banks’ gross credit (CC), deposit, and capital shares (stock method):

Denmark used such a mixed system of credit ceilings in the early 1980s. More precisely,

Under this scheme, credit weights were supplemented by deposit weights; however, the National Bank of Denmark scaled down the banks-specific deposit growth to counteract what it assessed as being too fierce competition for deposits.18

The case of Denmark is in fact illustrative of a general shortcoming of the stock method or variations thereof in that the choice of the αi may lead to a decrease of allowed credit for particular banks, even though overall PC t+1 increases. To see this, compute the individual credit increase:

It is possible to develop examples of banking systems where this expression can be negative under normal circumstances if α1gt is too small. One such example is presented in Appendix II. It implies that α1 must be set rather high and the other coefficients correspondingly lower to avoid situations in which a particular bank would have to recall loans.19 The trading of credit ceilings would alleviate the problem, but, even so, particular banks would be penalized simply because the stock method was chosen as the weighting rule.

Flow Method

One may consider a weighting rule in which, instead of allocating PCt+1, one allocates ΔPCt+1—the flow envisaged in the monetary program during period t. The expression for the ceiling of bank i under the flow method becomes

Here, individual ceilings automatically increase as long as overall credit increases. Comparing PCi, t+1 - Ci, t under the stock method to the same expression under the flow method, one sees that the stock method almost always imposes larger adjustments than the flow method and may therefore appear to be more arbitrary. The limiting case when the two methods converge is for α1 to equal 1.

If applied mechanically, the stock and flow methods have different implications for the treatment of unused credit margins, because of base drift. The stock method automatically allows some proportion of unused margins to be used in the periods ahead. The proportion depends on the definition and evolution of the weights and on the extent to which the overall credit target will be revised in the next period. But all else being equal, the entire unused credit of period t is available in period t+1, because the target for overall credit in the next period, PC t+2, is in principle not affected by the nonutilization of a particular PCi, t+1. On the contrary, the flow allocation automatically penalizes those banks that undershoot their ceilings (all else being equal, it rewards banks that overshoot), because the nonutilization of PCi, t+1 lowers the base C i, t+1 to which the flow ΔPCi, t+2 is applied. These problems are somewhat circumscribed if credit ceilings are allowed to be traded, because banks may then directly determine the outcome, Ci, t+1. Given the above, in implementing the flow method, one needs to take base drift into account, as shown in Appendix II. Appendix II also shows how to modify the treatment of base drift when the trading of credit ceilings is allowed under the flow method.

It is easy to see from equation (7) that the implementation of the flow method is not constrained by particular values of the parameters α1, whereas in the stock method, α1cannot be set too low, in order to avoid loan recalls for certain banks. This property therefore makes it possible to emphasize the role of capital in the mix of weights; that is, it is possible to set α1 = α2 = 0 and α3 = 1 (flow method 2 in Appendix II), or even to introduce capital adequacy in the computation of weights (flow method 3). Under those circumstances, banks would have an additional incentive to increase their shareholder funds if they wanted to develop their credit business. However, Appendix II shows that the effect of this property is marginal and that the choice of the base as credit net of shareholder funds already provides a strong incentive to increase those funds.

Appendix II. Numerical Example of Bank-by-Bank Credit Ceilings

The following example illustrates the points made in Appendix I. It assumes a banking industry composed of three banks: Bank L, whose main activity is lending to the economy, funded by interbank funds and deposits; Bank S, a savings bank; and Bank N, a new bank. The notations of Appendix I are being used. Section I presents the initial balance sheet of the banks at time T. Section II calculates individual bank weights based on these data and various examples of weighting schemes. Section III calculates bank credit ceilings for the period T to T + 1 under the stock and flow allocation methods by applying the weights of Section II to the aggregate banking system credit target and the base of individual credit volumes of Section I. Section IV is the tabulation of the outcome and the calculation of the new base of individual credit volumes as a function of the deviations of the realized credit volumes from the credit ceilings and of the trading of credit ceilings during T to T+1.

I. Initial balance sheet(time T)
Bank LBank SBank NTotal
In domestic currency units
Credit to nonbanks850600301,480
Other assets25045070770
Liabilities and equity1,1001,0501002,250
Own funds and provisions80737160
Other liabilities18073190
II. Calculation of individual bank weights
Shares at T

for i = L, S, N
Bank LBank SBank NTotalProposed

In percent
Credits (Ci, T)57412100α1 = 0.6
Deposits (Dj, T)44515100α2 = 0.3
Own funds and provisions (Ki, T)50464100
Memorandum item:
Capital adequacy (CAi, T)18.28.810.8

These are assumptions Capital adequacy needs to be calculated separately using the usual standards

These are assumptions Capital adequacy needs to be calculated separately using the usual standards

Using the proposed coefficients given in the far right column above, the individual bank shares of aggregate credit are computed using equation (4) of Appendix I (for i = L, S, N):

Bank LBank SBank NTotal
Weights 10.5270.4420.0311.0

Another possibility is to use the same formula as above, except that α1 = α2 = 0 (the absolute level of capital is given all the weight). The result is

Bank LBank SBank NTotal
Weights 20.5000.4630.0371.0

Still another option is to define weights as Ki, t/Kt*CAi, t normalized to one. In this option, weights 2 are scaled by the banks’ relative capital-asset ratios in order to reward banks with higher capital adequacy. The result is

Bank LBank SBank NTotal
Weights 30.4770.4680.0551.0
III. Aggregate banking system target and Individual bank ceilings
Bank LBank SBank NTotal
Net credit base at T1770527231,320
Overall banking system target for T +12...............................1,400
Overall change in target for T, T+1>......................................80
It is now possible to compute the banks’ credit ceilings under various methods.
Bank LBank SBank NTotal
Stock method (using weights 1)3738619431,400
Flow method 1 (using weights 1)812562251,400
Flow method 2 (using weights 2)810564271,400
Flow method 3 (using weights 3)808565271,400

Defined as gross credit minus own funds and provisions.

Data given by the monetary program.

Using the coefficients (0.6. 0.3, 0.1), the outcome of the stock method would require Bank L to reduce its lending, possibly by recalling loans. To avoid this, α1 would need to be greater than 0.8, with corresponding reductions in α2 and α3 thereby rewarding artificially those banks with the largest shares of the credit market. The flow method does not have this limitation.

Defined as gross credit minus own funds and provisions.

Data given by the monetary program.

Using the coefficients (0.6. 0.3, 0.1), the outcome of the stock method would require Bank L to reduce its lending, possibly by recalling loans. To avoid this, α1 would need to be greater than 0.8, with corresponding reductions in α2 and α3 thereby rewarding artificially those banks with the largest shares of the credit market. The flow method does not have this limitation.

IV. Outcome at T+1 and recalibration of base and weightsAs an example, we use the credit ceilings computed above under flow method 3.
Bank LBank SBank NTotal
Net credits at T + 1(B)820550381,408
Deviation from allocation(B-A)12-15118
Of which, traded(C)7-1140
Base for assessing the penalty(B-A-C)58
New credit base(A+C)816554301,400

Compute new weights Ki, T+1/KT+1*CAT, t+1, etc.


    AbramsRichardK.1988“The Financial Reform in Finland,”IMF Working Paper 88/89 (Washington: International Monetary Fund).

    • Search Google Scholar

    AghevliBijanEduardoBorensztein and Tessavan der Willigen1992Stabilization and Structural Reform in the Czech and Slovak Federal Republic: First StageIMF Occasional Paper No. 92 (Washington: International Monetary Fund).

    • Search Google Scholar

    AlexanderWilliamTomásJ.T. Baliño and CharlesEnoch1995Adoption of Indirect Instruments of Monetary PolicyIMF Occasional Paper No. 126 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    BinghamT.R.G.1985Banking and Monetary Policy (Paris: Organization for Economic Cooperation and Development).

    • Search Google Scholar

    BisatAmerR.Barry Johnston and V.Sundararajan1992“Issues in Managing and Sequencing Financial Sector Reforms: Lessons from Experiences in Five Developing Countries,”IMF Working Paper 92/82 (Washington: International Monetary Fund).

    • Search Google Scholar

    Blundell-WignallAdrian and FrankBrowne1991“Macroeconomic Consequences of Financial Liberalisation: A Summary Report,”Department of Economics and Statistics Working Paper No. 98 (Paris: Organization for Economic Cooperation and Development).

    • Search Google Scholar

    Board of Governors of the Federal Reserve System1980Federal Reserve Bulletin (April).

    • Search Google Scholar

    CaprioGerardJr. and PatrickHonohan1990“Monetary Policy Instruments for Developing Countries,”Policy, Research, and External Affairs Working Paper WPS 528 (Washington: World Bank).

    • Search Google Scholar

    CottarelliCarlo1993Limiting Central Bank Credit to the Government: Theory and PracticeIMF Occasional Paper No. 110 (Washington: International Monetary Fund).

    • Search Google Scholar

    CottarelliCarlo and GiampaoloGalli1986“Monetary Policy Through Ceilings on Bank Lending,”Economic Policy: A European ForumVol. 1 (October) pp. 673710.

    • Search Google Scholar

    DavisRichardG.“An Analysis of Quantitative Credit Controls and Related Devices,”Brookings Papers on Economic Activity: 1 (1971) pp. 65104.

    • Search Google Scholar

    EideLeif1985“Changes in Credit Policy—New Instruments,”Economic Bulletin Norges BankVol. 56 (March) pp. 4448.

    • Search Google Scholar

    GalbisVicente1993“High Real Interest Rates Under Financial Liberalization: Is There a Problem?”IMF Working Paper 93/7 (Washington: International Monetary Fund).

    • Search Google Scholar

    GuitiánManuel1993“The Role of Monetary Policy in IMF Programs” (unpublished; Washington: International Monetary Fund).

    • Search Google Scholar

    HilbersPaul1991“Monetary Cash Reserve Arrangement,”Quarterly Bulletin Nederlandsche Bank pp. 2739.

    • Search Google Scholar

    HilbersPaul1993“Monetary Instruments and Their Use During the Transition from a Centrally Planned to a Market Economy,”IMF Working Paper 93/87 (Washington: International Monetary Fund).

    • Search Google Scholar

    JohnstonR.Barry1990“Issues in the Design and Implementation of Monetary Control and the Transition to Indirect Monetary Instruments” (unpublished; Washington: International Monetary Fund).

    • Search Google Scholar

    JohnstonR.Barry1993“Aspects of the Design of Financial Programs with the Adoption of Indirect Monetary Controls,”IMF Paper on Policy Analysis and Assessment 93/16 (Washington: International Monetary Fund).

    • Search Google Scholar

    LindgrenCarl-Johan1992“The Transition from Direct to Indirect Instruments of Monetary Policy,” in The Evolving Role of Central Banksed. by P.Downes and R.Vaez-Zadeh (Washington: International Monetary Fund).

    • Search Google Scholar

    MagnussenEinar1984“Use of Credit Policy Instruments in the Years Ahead,”Economic Bulletin Norges BankVol. 55 (June) pp. 14754.

    • Search Google Scholar

    MarstonDavid1995“Financial Sector Reform in Jamaica During 1985–1992: Possible Lessons for the Caribbean,”IMF Working Paper 95/90 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    MolhoLazaros and V.Sundararajan1988“Financial Reform and Monetary Control in Indonesia,”IMF Working Paper 88/4 (Washington: International Monetary Fund).

    • Search Google Scholar

    National Bank of Denmark1984Annual Report (Copenhagen).

    QuintynMarc1991“From Direct to Indirect Monetary Policy Instruments: The French Experience Reconsidered,”IMF Working Paper 91/33 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

    SaalMatthew and LorenaZamalloa1995“Use of Central Bank Credit Auctions in Economies in Transition,”Staff PapersInternational Monetary FundVol. 42 (March) pp. 20224.

    • Search Google Scholar
    • Export Citation

    vander WerffA.A. and Th.J.Sluijter1989“Informal Restrictions on Banks’ Money Creation (1986 and 1987),”Quarterly BulletinNederlandsche Bank (No. 235) pp. 3347.

    • Search Google Scholar

    van RoodenR.P.A.1990“Monetary Cash Reserve Arrangement,”Quarterly BulletinNederlandsche BankNo. 3 (December) pp. 2132.

    • Search Google Scholar

Note: The authors wish to thank Tomás Baliño, Charles Enoch, and Lorena Zamalloa for their valuable comments. Information provided by de Nederlandsche Bank is gratefully acknowledged. The authors are responsible for any errors.


In the broadest sense, indirect methods seek to achieve monetary control (of either a quantity or a price target) through full use of market prices. See Alexander. Baliño, and Enoch (1995)


Countries surveyed include Albania, Belgium, Bulgaria, the Czech Republic, Denmark, Egypt, France, Guyana, Italy, Jamaica, Kenya, the former Yugoslav Republic of Macedonia, Madagascar, the Netherlands, Norway, Poland, Portugal, Romania, and the United Kingdom.


Empirical discussion on the effectiveness of credit ceilings is difficult because, in most cases, ceilings were used in combination with other forms of direct monetary control, including, among other instruments, administratively set interest rates and liquidity ratios. This instrument mix made it difficult to isolate the effects of the credit ceilings on the economy from those of other instruments of direct control. Also, in a majority of cases, the authorities exempted many categories of credit from the ceilings, typically to combine controls of aggregate credit ceilings with attempts at directing credit to priority sectors. Such an approach further complicates the task of identifying and assessing the effectiveness and efficiency of credit ceilings on their own. Therefore, the problems associated with the use of credit ceilings may be broadly viewed as representative of problems associated with repressed financial systems.


For instance, nearly all Asian countries, including Indonesia, the Republic of Korea, Malaysia. Nepal, the Philippines, and Sri Lanka, reduced or eliminated direct controls on bank lending. In June 1995, however, Thailand introduced controls on credits extended in foreign currency. The experience of African countries in removing credit ceilings has been more mixed.


This is referred to as a “belts and braces” approach in Alexander, Baliño, and Enoch (1995).


Albania, China, Ethiopia, Guinea-Bissau, Jordan, Lesotho, Madagascar, Mauritania. Mongolia. Mozambique, Pakistan, São é and Principe, Slovak Republic. Tanzania, Vietnam, and former Yugoslav Republic of Macedonia. Furthermore, Cambodia and the Lao People’s Democratic Republic introduced credit ceilings in 1995. The majority of these countries implemented these ceilings while carrying out programs supported through the use of IMF resources.


Net credit aggregate includes total credit, including nonperforming loans, less capital, provisions, and nonmonetary liabilities.


An unused margin was the difference between the ceiling and a bank’s actual growth of net lending. The monetary cash reserve arrangement was approved in July 1989, for a period of one year. In June 1990, the authorities renewed the arrangement in a slightly modified form by allowing the cash reserve percentage to remain zero. This was motivated by the decline in the growth of banks’ money creation. This arrangement has not been renewed since then, but can be reactivated at any time.


This term refers to banks’ adjusting their balance sheets to meet the credit ceilings only on the test dates.


As of November 1989, the authorities in Egypt changed the penalty for noncompliance to include ⅙ of 1 percent a day on balances in excess of limits.


The system was highly progressive, since the ratio was a quadratic function of the increase in excess of the set standard. See Quintyn (1991).


Faced with a rapid growth of monetary aggregates in the following periods, in 1973, the authorities were compelled to reintroduce the credit controls in the form of a Supplementary Special Deposit Scheme, or “corset.” Under this system, the Bank of England established limits for the growth of interest-bearing liabilities of banks for a specific period of time relative to a defined base period. Any bank that exceeded the specified rate of expansion in its interest-bearing liabilities was required to place a certain amount of noninterest deposits with the Bank of England. The volume of such penalty deposits required of a bank rose progressively as the rate of expansion of the banks’ interest-bearing liability exceeded the stipulated limit. By adjusting this penalty rate, the Bank of England could increase the effective cost of new funds to deter banks from aggressive competition for funds and to encourage banks to increase their lending rates. By the end of 1980, the Bank of England had removed all quantitative restrictions, including the corset. While the option to call special deposits has remained in place, it has not been exercised since then.


However, all allocation rules will involve an element of arbitrariness as the weights are not chosen through a market mechanism.


See appendices for a technical discussion of the stock and flow methods.


In the flow method the base against which compliance is assessed needs to be recalculated with the same frequency as the one used for testing compliance.


For a discussion of ways to design credit auctions under adverse selection, see Saal and Zamalloa(1995).


PCM is consistent with an overall broad money program that typically puts a separate ceiling on net credit to government and a floor on net foreign assets. Using standard notations, NC = M + OIN- NFA - NCG, where OIN is defined to exclude capital and provisions. M, NFA, NCG, and OIN are projected; NC is the residual.


Moreover, it was found that banks whose loan-to-deposit ratios were lower than the industry average were often too constrained in expanding credit; for these banks, the credit ceiling was increased toward the average. These higher-than-normal credit ceilings were then offset by allocating lower-than-normal ceilings to banks with relatively low increases in deposits, irrespective of the banks’ actual deposit mobilization. See National Bank of Denmark (1984).


Some central banks (e.g., Italy and Jordan) used to compute credit ceilings according to a variant of the stock method and then adjust outcomes based on a review of the particular circumstances of the banks. However, this approach introduces the type of discretion that may ultimately undermine the integrity of the instrument by granting too many exemptions.

    Other Resources Citing This Publication