a. General considerations

The main objectives of financial reform are usually to develop a financial system that promotes savings and a more efficient allocation of resources while also providing a framework that allows for the implementation of effective monetary control. These characteristics are normally associated with well-functioning markets and a competitive financial structure. However, if the underlying structure of prices is distorted or if enterprises face “soft” budget constraints and do not respond to price signals, well-functioning markets may not provide an optimum allocation of resources. ^3/ Hence, financial sector reform has to be seen as part of a broader economic restructuring. ^4/

There is no guarantee that a market-based system would be competitive. Lack of competition between financial institutions could result in wide transaction margins and inhibit the development of new instruments and markets. There is a risk that the inefficiencies associated with direct controls would be replaced by the inefficiencies of a monopolistic structure. While a monopolistic structure may not necessarily interfere with monetary control, it could result in a sluggish response of deposit and lending rates to monetary policy initiatives. ^5/ Where banking cartels exist which include state-owned banks, these banks may have to be instructed to compete to weaken the cartels. Monopolistic agreements on interest rates and transaction margins may have to be challenged by the central bank if they run contrary to the law. The removal of direct controls can itself reduce incentives for monopolistic pricing arrangements by increasing the areas where competition is possible.

For interest rates to have the appropriate impact on resource allocation, borrowers have to be responsive to market prices. A high incidence of credit indiscipline and loan delinquency, or arrangements where borrowers or banks expect to have their loans covered by the government, may render interest rates meaningless in the allocation of credit. In some countries, direct credit controls have encouraged loan delinquency and a lack of attention to the quality of bank balance sheets. The strengthening of bankruptcy laws, the legal procedures for loan recovery, and policies to promote the profit motivation of managers may have to accompany the financial reforms.

In a liberal and competitive system, interest rates would be differentiated to reflect the riskiness of borrowers and the maturities of instruments, and credit would flow into those activities which yield the highest expected rates of return. When the system is competitive, transaction margins in the financial sector usually fall and new financial instruments are created to reflect the preferences of investors and borrowers, including active secondary markets which provide increased liquidity to financial instruments.

b. Monetary control framework

The framework for monetary control in a liberal financial system varies between countries. The distinction between operational variables, intermediate targets, and ultimate objectives is one that is in common use. The intermediate targets are variables, such as monetary or credit aggregates, which are related in a stable manner to ultimate policy objectives—real GDP, inflation, the balance of payments. Variables, such as money market interest rates or the level of money market liquidity, which are directly under the control of the monetary authorities, and which are related to the path of the intermediate target variables in a predictable manner, are the operational variables for monetary policy. ^6/ The authorities manage the operational variables and leave market mechanisms to determine the detailed structure of deposit and lending rates and the allocation of credit. In a liberal system, the main instrument of monetary control would be the central bank’s control over interest rates in the money market through its ability to manage the stock of reserve money (cash and balances with the central bank). Money market rates, in turn, affect other lending and deposit rates. Such a system does not rely on high or even compulsory reserve or liquid asset requirements, but only on the central bank’s ability to manage its own balance sheet and to control the terms at which it is willing to provide assistance to cover reserve shortages.

Financial reform may raise basic questions about the selection of appropriate target variables. Relationships between money and credit and final objectives that existed during the pre-reform phase are likely to be less reliable following financial reform. The introduction of new instruments may shift the interest sensitivity of money demand. For example, a liberalization of interest rates on bank deposits may result in broad money becoming less sensitive to changes in the general level of interest rates. The removal of credit ceilings that constrain portfolio allocation can result in portfolio shifts, including a demand to hold larger money balances at any given level of interest rates and income. ^7/ Money multipliers may also change with reform, for example, when there is reintermediation to the banking system leading to reduced cash holdings by the nonbank public. Not to allow for the portfolio adjustments in policy formulation could lead to serious policy errors and loss of monetary control.

Not only can the relationships shift, but the controllability of different aggregates may change following financial reform. As higher interest rates may lead initially to distressed borrowing, credit growth may thus be difficult to control in the short run through changes in interest rates. Once credit ceilings are removed, the impact of a change in interest rates on broad money and credit may become uncertain, since in a liberal environment it will depend on the response of financial institutions in adjusting their rates to a change in indicative market rates. This may suggest reformulating target variables in terms of non-interest-bearing monetary aggregates, which have a more predictable response to changes in the general level of interest rates.

Following financial reform, a good deal of judgment may be needed for a time in formulating and implementing monetary policy. Judgment could be based on the examination of a range of indicators of financial conditions including the exchange rate, inflation, asset prices, as well as various monetary and credit aggregates. It obviously becomes more difficult to formulate precise monetary and credit targets that can act as benchmarks in macroeconomic adjustment programs when structural changes are occurring in the financial system. Nevertheless, quantitative monetary or credit targets often remain a central element in macroeconomic policy formulation. ^8/

c. Phasing of the reform

Important questions for financial reform are raised by the appropriate phasing of the reform process. ^9/ The move to more flexible interest rate arrangements based on indirect management of interest rates and money and credit aggregates usually requires the adaptation or development of the instruments for monetary control; the promotion of financial markets and competition in the financial sector; the buildup of expertise with regard to operating in a market environment; and a change in attitudes toward the market determination of interest rates and their use for macroeconomic policy. This process of market orientation takes time as experience with and confidence in the new system has to be gained by the central bank and the private sector, and new institutional arrangements may have to be established to ensure effective monetary control and competitive market mechanisms. A phased approach to reform is therefore often necessary. However, if the initial financial system is highly regulated, with each regulation supporting the other, it may be difficult to implement a gradual, effective reform. In such cases, a major one-step change in the operational system may be unavoidable.

Some of the distortions due to direct controls could be minimized by freeing up commercial bank lending and deposit rates from official interference, by promoting competition and by removing credit subceilings. However, overall credit ceilings on individual banks could still inhibit competition in the financial system and would still impose an implicit tax on regulated financial institutions and thus create incentives for avoidance. Hence, partial liberalizations may not be possible or desirable. Still, during the early phases of reform the maintenance of credit ceilings is often essential as a temporary measure.

Under a phased approach, interest rates could be set free in stages, for example, by gradually widening the permissible ranges for deposit and lending rates, or by steadily adjusting minimum deposit rates and maximum lending rates. The traditional instruments—reserve requirements and rediscount windows—could be deployed initially to manage bank reserves and thereby the credit-creating capacity of the banking system, while phasing out credit ceilings on banks.

The eventual elimination of credit ceilings on banks requires that macromonetary management would be exercised through the indirect management of the aggregate supply of money and credit which calls for adaptation in the instruments of monetary policy. Thus, market-based instruments of monetary policy have to be introduced as the reforms proceed to manage interest rates and credit. The objectives of efficient resource mobilization and allocation would also be helped by promoting flexibility in the determination of interest rates and the allocation of credit.

The next phase would involve the development of money markets and the promotion of a more competitive financial system in order to strengthen the monetary control and interest rate mechanisms. The development of market-based instruments of monetary control can be regarded as the initial step in this process. This phase of the reform may require institution and market building.

The particular phasing of reforms would depend upon the institutional circumstances. For example, the use of uniform reserve requirements to constrain credit expansion in the early phases of reform may be difficult when the distribution of excess bank reserves is uneven, reflecting the structure of the banking markets. Moreover, it would be undesirable to raise reserve requirements if they are already at high levels and are nonremunerative since this may encourage disintermediation. Therefore, it may be necessary to develop market-based instruments of greater flexibility at an early stage in the reform process.

a. The role of primary bill markets as monetary control instruments

The starting point for developing market-based monetary control in developing countries is the existing institutional structure. Because commercial banks are often already required to hold treasury bills to meet liquid asset requirements or have built up excess reserves with the central bank, a first step in developing market-based instruments has often been primary issues of treasury bills and/or central bank bills, supported by a restructuring of central bank rediscount windows to allow for a greater role for market forces. There is no necessary reason why treasury or central bank bill issues have to be the first step, and in some countries the catalyst has been markets in longer-term government securities. ^13/ Issues of central bank paper have been considered useful because of the potentially greater freedom for the central bank to tailor issues of its own papers to achieve monetary objectives without direct interference from government budgetary considerations. However, the net budgetary and monetary impact from using central bank rather than treasury paper would be similar—for example, government budget finance would be affected pari passu with the net cost of issues of central bank bills through reduced central bank profits available for appropriation by the government. Hence, the difference may be largely presentational.

The role of security issues as instruments of monetary control can be illustrated using basic financial flow and balance sheet equations. This section illustrates the relationship between the treasury bill issue and control of the reserves of the banking system; a similar relationship can be derived for central bank bill issues or other debt issues by the government. The derivation is based on sectoral financial flows and institutional balance sheets during a planning period.

The basic financial flow and balance sheet equations are:

the government’s current financing requirement, dG,

the external current account surplus, CA,

the flow balance sheet of the central bank

the flow balance sheet of commercial banks

and the change in net financial wealth of the nonbank private sector, dW,

where the subscripts G, CB, NB and S refer, respectively, to the government, central bank, nonbank private sector and commercial banks; dT is the treasury bill issue; dS is other government domestic borrowing; dF is the change in net holdings of foreign assets; dC is the change in holdings of cash and reserves with the central bank; dL is the change in net loans by the central bank; dA is the change in loans by commercial banks to the nonbank private sector; and dD are deposits of the nonbank private sector with commercial banks. In addition, the total issues of treasury bills and other government domestic debt are purchased by nonbanks, banks and the central bank.

and

Furthermore, the financial flows identity requires that private sector net wealth changes through surpluses with other sectors—the government and the foreign sector

Equation (3) shows that the private sector’s (bank and nonbank) holding of reserves is influenced by the central bank’s net lending to the private sector, plus its uptake of government debt (treasury bills and other debt) plus its net purchase of foreign assets through intervention in the foreign exchange market. The change in reserve holdings of commercial banks, dC_B,, can be derived in terms of the domestic financing of the government’s current deficit, central bank net intervention in the foreign exchange market, sales of treasury bills and other government debt outside the central bank, central bank net lending to banks and nonbanks, and the change in cash holdings of the nonbank private sector

This equation can be derived either by substituting equations (1) and (2) into (8) and using equations (4) and (5) or by substituting equations (6) and (7) into (3).

Equation (9) provides the framework for using the treasury bill issue (or other central bank or government security issues) to control the reserves of the banking system. The procedure would be to develop forecasts for the relevant reserve holding period of the government’s net domestic financing requirement (E_G - R_G + dF_G), net foreign exchange market intervention by the central bank, dF_CB, other domestic borrowing by the government from banks and nonbanks (dB_NB + dB_B), net lending by the central bank to the banks and nonbanks (dL_NB + dL_B), and the change in currency holdings by the nonbank private sector, dC_NB. Using equation (9) a forecast can be made of the change in the supply of bank reserves before a new net issue/net redemption of treasury bills to the private sector (dT_B + dT_NB = 0)

where ^* indicates a forecast and ${\mathrm{d}\mathrm{C}}_{\mathrm{B}}^{\mathrm{S}}$ is the forecast supply of new bank reserves before a new net issue of treasury bills.

The change in the desired demand for reserves by the banking system is given by

where α is the compulsory cash reserve ratio on bank deposits, β is the estimated precautionary reserve holding ratio of banks and dD* is the change in deposits for the relevant reserve holding period. ^14/ With lagged reserve accounting, required reserves would be based on deposits in a previous time period and may be known before deciding on the amount of treasury bills to be issued. With contemporaneous or leading reserve accounting dD* would be a forecast.

The demand for precautionary reserves may depend on the rules and procedures for reserve holding and interbank clearing, the efficiency of the interbank market, and the volatility in the reserves of banks due to seasonal and other factors, such as the extent to which the central bank stands ready to smooth the reserves of the banking system. For example, automated access by banks to a loan or rediscount window could reduce precautionary reserve holdings to a low level. However, even if it is known that the central bank will always remove reserve shortages, uncertainty may remain about the level of interest rates charged by the central bank, which might still lead to precautionary reserve holdings by banks.

Equations (10) and (11) in combination provide a forecast of the banking system’s reserve deficiency (if $\mathrm{d}{\mathrm{C}}_{\mathrm{B}}^{\text{D}}\mathrm{\text{>}}{\mathrm{\text{dC}}}_{\mathrm{B}}^{\mathrm{S}}$) or excess (if $\mathrm{d}{\mathrm{C}}_{\mathrm{B}}^{\text{D}}\mathrm{\text{<}}{\mathrm{\text{dC}}}_{\mathrm{B}}^{\mathrm{S}}$) compared with desired holdings before net new issue of treasury bills. ^15/

Given these forecasts, the net new issue of treasury bills can be set to achieve monetary objectives. The transmission mechanism for monetary policy could be described by the following relationships:

where r_m is the money market interest rate, r_d and r_L are average deposit and loan rates, and M is the intermediate target variable (e.g., the stock of money or credit) which is responsive to interest rates.

A neutral monetary policy stance—i.e., one which would leave money market rates unchanged—would involve removing the excess reserves of the banking system by selling a volume of treasury bills to the private sector equal to the difference between the forecast of the supply of reserves to the banking system and the estimate of the banking system’s demand for reserves (or relieving a forecast reserve shortage by not rolling over maturing bills). A restrictive policy stance designed to raise interest rates would involve the creation of a reserve shortage by selling a volume of bills in excess of the forecast reserve surplus. In this case, and assuming the forecasts are accurate, banks would find themselves short of reserves and interbank rates would tend to rise. If the sale of treasury bills reduced the aggregate supply of reserves below reserve requirements, banks would have to seek marginal accommodation from the central bank and the interest rate at which the central bank provides the assistance would become the key determinant of the overall level of interest rates. ^16/ An expansionary stance would involve selling fewer bills than the forecast cash surplus. In this case banks would have excess cash reserves which would push interest rates down. Interbank rates would fall, and there would be repayments of the central bank credit. In this framework, the design of rediscount windows is an integral element of reserve and interest rate management (see section 5.d). ^17/

b. Implications of liquid asset requirements

The preceding analysis assumes that banks have to meet only reserve ratios. However, in practice many developing countries also impose statutory liquid asset requirements and treasury/central bank bills and reserves are generally the main instruments available to satisfy such requirements. ^18/ Liquid asset requirements are generally used to allocate a proportion of bank credit to the government in the form of purchases of treasury bills, and can also support the introduction of more market-oriented control techniques. However, under a system of bank reserve management as described above, liquid asset requirements can also complicate monetary management during periods of reserve shortages.

The liquid asset requirement could be written as

where it is assumed that excess reserves are counted as liquid assets and δ is the liquid asset ratio. If banks are initially meeting their liquid asset requirements, i.e., equation (16) is satisfied, then the requirement will continue to be satisfied when there is a reserve surplus and the central bank sells treasury bills to meet targets for reserves of the banking system. Such sales alter the composition of liquid assets but not the overall total. However, during periods of reserve shortages, when commercial banks hold limited liquid assets in excess of the requirement, a purchase of treasury bills by the central bank to relieve the reserve shortage could create a shortage of liquid assets: the liquid asset requirement could become a binding constraint.

To avoid this potential complication, the monetary authorities could reduce the liquid asset requirement. As liquid assets held to meet requirements are by definition not liquid, there are advantages for liquidity management from replacing the ratio with a more flexible procedure such as a “liquidity norm,” determined by prudential considerations, that should be met on average over a sufficiently long time period, but which should not form a binding constraint at any given point in time. ^19/ In general, a reduction in the ratio is also often desirable as part of the policy toward liberalizing interest rates and removing the captive market for treasury bills. However, monetary authorities are often reluctant to reduce liquid asset ratios during the early phases of financial reform since they may view the ratios as a way of ensuring the demand for bills. In this case, the complications could be avoided by creating sufficient new liquid assets through the conversion of outstanding credit lines to government into treasury bills or by allowing short-term prime commercial bills to be created and counted as liquid assets. An alternative procedure could be to relieve reserve shortages through reversed transactions in treasury bills, rather than outright purchases, and allow the treasury bills to continue to be counted toward meeting liquid asset requirements. ^20/ The latter approach would not, however, solve the problem of a fundamental shortage of discountable paper which has occurred in some countries.

c. Selling techniques for bills

Selling techniques for primary bill issues range from “free” auctions to fixed price tenders. Under “free” auctions the central bank announces the volume of bills to be sold and asks for competitive bids. Bills are allocated starting with the highest bid price and moving to the next highest, etc. until the tender is fully sold. The auction produces a range of interest rates, with the highest rate determined by the marginal bid that exhausts the tender. This marginal rate is sometimes known as the “cut-off” or “striking” rate. Under a fixed price tender, the central bank sets the interest rate on bills and accepts the volume demanded at that rate. With a “free” auction, interest rates fully adjust to achieve the quantitative objectives for the sale of bills and hence for the targeted stock of reserves, but this may risk volatility in interest rates from auction to auction. On the other hand, with a fixed price tender, the interest rate is determined in the short run and quantities allowed to vary. The tender price can be varied from one tender to the next to exercise the desired quantitative control, but this procedure may risk a “bias to delay” in adjusting interest rates.

Central banks have operated a number of selling procedures between these two extremes to avoid possible excessive interest rate volatility, and at the same time allow for flexibility in interest rate adjustment. In some cases, bills are auctioned, but no tender volume is announced in advance. The central bank exercises its discretion in accepting or rejecting bids, depending on its quantitative or interest rate objectives. The decision not to announce the tender amount can thus give the central bank added flexibility, although at the cost of reducing information to market participants. In the initial stages, the benefits to the central bank of increased flexibility can help overcome concerns about implementing new monetary control procedures. In some cases, bills are auctioned and tender volumes announced in advance, and a permissible range is also announced for bids that would be accepted. The range is usually based on the interest rate set at the previous auction. In some countries the central bank, or other financial institutions, underwrite the bill issue at interest rates indicated/agreed with the central bank. This limits the potential upward adjustment in rates. Another procedure is for the central bank to preannounce an initiative auction amount, but to vary it once bids are received in order to achieve interest rate objectives.

d. Intra-bill issue operations

Issues of primary bills could in principle be held frequently. ^21/ However, sales of bills to the general public involve administrative costs and delays associated with advertising the sale, receiving and filling bids, etc. Primary bill issues have, as a result, tended to occur no more frequently than once a week. Between the issues, there may therefore be a need for supplementary central bank operations to manage money market liquidity. For example, the actual outcome for reserves may deviate significantly from the outcome projected by the authorities, or the issue of bills may be such as to create a reserve shortage and force commercial banks to borrow from the central bank. If not offset by the central bank, the resulting reserve shortages or surpluses could result in excessive volatility in interest rates and could complicate banks’ compliance with legal reserve requirements. Concerns about day-to-day reserve shortages and the potential volatility of money market interest rates can be reduced by requiring that reserve requirements are met on a period-average basis rather than continuously or at a specific point in time. ^22/ Hence, the adaptation of reserve averaging procedures is usually recommended.

It is conceivable that under a strict monetary base control regime, behavioral changes, such as increased precautionary bank reserve holdings, would remove the need for frequent money market intervention by the central bank. In some countries where the central bank has been inactive in the money markets, commercial banks have built up large precautionary reserve balances, apparently to meet short-term fluctuations in their liquidity, although this build-up has also often been accompanied by official concerns about “excess liquidity” and the underutilization of commercial bank resources. ^23/

Even if banks could be assured of meeting their legal reserve ratios, a high degree of money market interest rate volatility may be undesirable because of implications for exchange rates, or because it may obscure the underlying trend in interest rates. Also, interest rate volatility may threaten the solvency of financial intermediaries that have traditionally operated in an environment of stable interest rates. Nevertheless, some day-to-day flexibility in money market rates is necessary to provide profitable speculative and arbitrage opportunities that encourage secondary markets and market makers.

The aim of achieving interest rate flexibility while avoiding excessive volatility often calls for day-to-day intervention by the central bank. This can take a number of forms, with different implications for financial market development. A common technique is for the central bank to preannounce its willingness to buy or sell bills at interest rates close to those determined on the last primary issue. However, providing banks with unlimited and automatic access to central bank facilities at market rates—be they sales and purchases of treasury bills, repurchase agreements, foreign currency swaps or rediscounts—may have drawbacks for financial market development. As such facilities allow banks to employ (relieve) temporary surpluses (shortages) of funds at little cost, there are limited incentives for the development of secondary bill or interbank markets.

To encourage secondary market activity, some central banks have widened the margins between their buy and sell rates for bills, and/or set penal interest rates (discount or Bank rate) at which they are willing to lend to commercial banks, usually against collateral. ^{24/ 25/} The wide margins/penal loan rate provide incentives for banks that are short of funds to seek them initially from other banks, and banks with surplus funds to seek outlets for these funds in the interbank market, and therefore encourages interbank market development.

A loan window has the drawback that it operates asymmetrically, providing a facility only to eliminate reserve shortages, and could thus still permit substantial interest rate volatility, ^26/ although there is also a risk that the loan window rate would be viewed as the indicative rate for all interest rates, inhibiting their market determination. These concerns can be mitigated by providing graduated access to the loan window at progressively higher penal rates, and special interest-bearing accounts with the central bank for excess reserves, or by paying interest on all cash reserves deposited with the central bank. ^27/ Graduated access could be provided in a number of ways including the use of quotas and credit tranches for accommodation from the central bank.

The above procedures are based on preannounced interest rates, and leave the initiative to obtain or dispose of bank reserves entirely to the commercial banks. The alternative is for the central bank to take the initiative as regards the quantity of intervention and leave the interest rates to be market determined. For example, the central bank could conduct auctions of its money market intervention instruments, including outright sales and purchases of specified securities, repurchase agreements, foreign exchange swaps, and rediscounts. ^28/ When shortages are indicated, the central bank would announce its willingness to buy bills up to a certain amount, and ask for competitive offers; when surpluses are indicated the central bank could announce its willingness to sell a certain volume of bills, and ask for competitive bids. By restricting these auctions to a few key money market participants, they could be more flexibly implemented than the general auctions of primary bills.

The identification of the need for such auctions and the daily shortages or surpluses in the money market could be based on: (1) a monitoring of interbank rates throughout the day; (2) the amount of assistance provided through the loan window (if such a facility existed); (3) the ex post monitoring of developments in the central bank balance sheet components, relative to the forecast used to determine the size of the primary bill issue; and (4) a daily monitoring of commercial banks’ progress in meeting their cash reserve requirements.

e. Supporting policy and institutional reform

The development of market-based instruments of monetary control needs to be supported by changes in the overall policy and institutional framework.

The extent to which the interest rates determined through primary bill issues lead to an appropriate interest rate structure depends on institutional circumstances. As noted, the liberalization of interest rates on government securities could be inhibited by the captive markets created by high liquid asset requirements or other portfolio restrictions that may therefore need to be relaxed. When substantial resources are mobilized through public sector financial institutions (savings banks and provident funds) which traditionally buy government securities, these institutions may need to have their securities allotted outside the auctions, so that auction rates can reflect the preferences of noncaptive investors. Institutional reforms may be necessary to ensure that auction rates are reflected in banks’ deposit and lending rates. For example, differential cash requirements influence the competitiveness of financial institutions in raising deposits and granting loans, and may have to be unified. ^29/

As a result of interest rate liberalization, interest rates on government debt may rise, with budgetary implications. Similarly, parastatals and other borrowers, which had access to bank loans at preferential rates, may experience an increase in borrowing costs. As a result, certain activities may no longer be viable and productive sector restructuring has to take place.

For the central bank to implement the changes in the monetary control system described, it may be necessary to establish a new monetary policy operating framework. A monetary programming procedure would have to be established in order to carry out projections of the main elements in equations (10) and (11) (page 14), and thus determine the volume of bills to be sold to achieve the monetary objectives. The operational framework that may need to be developed for the purposes of determining the regular auction amounts, and the volume of money market intervention to be undertaken by the central bank between auctions, are illustrated in Table 2.

Table 2.

Framework for Regular Forecasting and Daily Monitoring of Cash Positions

Table 2.

Framework for Regular Forecasting and Daily Monitoring of Cash Positions

		Regular Forecast for Auction	Daily Monitoring for Money-Market Intervention	Action Necessary
1.	Commercial banks required cash balances	Based on projected growth of monetary aggregates developed by central bank as part of macroeconomic policy framework and analysis of seasonal pattern in deposits of commercial banks.	Based on framework for monitoring progress under the cash ratio requirements.	Central bank to prepare quarterly forecast of monetary aggregates and analysis of seasonal pattern in deposit data. Monitoring unit to prepare assessment of progress under the cash reserve ratios and discuss with banks the level of their desired precautionary cash balances.
2.	Foreign exchange	Based on quarterly projections of the balance of payments developed by the central bank and analysis of seasonal pattern in payment flows and intervention by the central bank.	Actual intervention by the central bank to be communicated daily to the monitoring unit. Forecast based on 2-day delay in valuing international transactions.	Central bank to prepare quarterly projection of the balance of payments and analysis of the seasonal pattern of payment flows.
3.	Currency outside banks	As in 1. with reference to currency holdings	The currency issue to be notified daily to the monitoring unit.	Central bank to prepare quarterly forecast of currency holdings and an analysis of the seasonal pattern in data.
4.	Government accounts	Forecasts of regular receipts and payments, for example, customs duties, wage and social security payments, etc., to be prepared by the relevant government departments. Accountant General to notify large irregular payments expected during forecast period. List of securities outstanding with amounts and maturity dates and notification of new issues with dates and amounts to be sold to the public.	Actual receipts and payments through government accounts to be notified to the monitoring unit.	Request forecasts and information from government departments.
5.	Maturing central bank bonds and assistance	Lists of bonds and assistance outstanding with amounts and maturity dates.	Actual repayments to be notified to the monitoring unit.	—
6.	Central bank current transactions	Central bank staff to prepare weekly forecasts.	Actual payments and receipts to be communicated to the monitoring unit.	—
7.	Other central bank current accounts	—	Actual balances in accounts to be communicated to the monitoring unit.	—

View Table

Table 2.

Framework for Regular Forecasting and Daily Monitoring of Cash Positions

		Regular Forecast for Auction	Daily Monitoring for Money-Market Intervention	Action Necessary
1.	Commercial banks required cash balances	Based on projected growth of monetary aggregates developed by central bank as part of macroeconomic policy framework and analysis of seasonal pattern in deposits of commercial banks.	Based on framework for monitoring progress under the cash ratio requirements.	Central bank to prepare quarterly forecast of monetary aggregates and analysis of seasonal pattern in deposit data. Monitoring unit to prepare assessment of progress under the cash reserve ratios and discuss with banks the level of their desired precautionary cash balances.
2.	Foreign exchange	Based on quarterly projections of the balance of payments developed by the central bank and analysis of seasonal pattern in payment flows and intervention by the central bank.	Actual intervention by the central bank to be communicated daily to the monitoring unit. Forecast based on 2-day delay in valuing international transactions.	Central bank to prepare quarterly projection of the balance of payments and analysis of the seasonal pattern of payment flows.
3.	Currency outside banks	As in 1. with reference to currency holdings	The currency issue to be notified daily to the monitoring unit.	Central bank to prepare quarterly forecast of currency holdings and an analysis of the seasonal pattern in data.
4.	Government accounts	Forecasts of regular receipts and payments, for example, customs duties, wage and social security payments, etc., to be prepared by the relevant government departments. Accountant General to notify large irregular payments expected during forecast period. List of securities outstanding with amounts and maturity dates and notification of new issues with dates and amounts to be sold to the public.	Actual receipts and payments through government accounts to be notified to the monitoring unit.	Request forecasts and information from government departments.
5.	Maturing central bank bonds and assistance	Lists of bonds and assistance outstanding with amounts and maturity dates.	Actual repayments to be notified to the monitoring unit.	—
6.	Central bank current transactions	Central bank staff to prepare weekly forecasts.	Actual payments and receipts to be communicated to the monitoring unit.	—
7.	Other central bank current accounts	—	Actual balances in accounts to be communicated to the monitoring unit.	—

The greater freedom given to banks as part of financial reforms can bring to the fore weaknesses in bank solvency and in the supervisory arrangements. A policy of allocating resources through directed credit often limits consideration of the quality of banks’ portfolios. When direct controls on interest rates and credit are relaxed, some borrowers that previously had access to directed credit may no longer be viable, and loans will have to be provisioned and written off. Moreover, a policy of interest rate regulation has in some countries led to excessive interest rate mismatching that is only revealed when interest rates are freed. Broad structural reforms can also bring about large changes in relative prices which can have widespread effects on the viability of borrowers. Hence, in a liberal framework, greater emphasis may have to be placed on prudential controls to safeguard the stability of the financial system.

In addition to reforming the central banks’ money market intervention facilities, the development of secondary markets may need to be encouraged through promoting market makers. The initial turnover in the markets may not make market making profitable, and transactions margins could as a result be excessively wide. Incentives for market makers would include restricting bidders at the auctions to the market makers and commercial banks; requiring purchases of treasury bills by other entities to be directed through the market makers; and by providing market makers temporary access to the central bank’s lender-of-last-resort facility. In return, the market makers could be required to provide buy and sell quotations for bills in the secondary market for various maturities, and to underwrite the primary bill issues.

a. Overview

This section reviews the uses of market-based instruments of monetary control in a sample of developing countries: Argentina, Brazil, Indonesia, Kenya, Malaysia, Mexico, Philippines, Sri Lanka and Thailand. The financial markets in these countries are at varying stages of development and the countries have faced different macroeconomic challenges in recent years. Nevertheless, all have reduced or abolished the use of direct controls in monetary policy, in particular controls on interest rates and credit, and have moved toward more market-based systems of monetary control. Some countries (Argentina, Brazil) have for a number of years operated market-based instruments of monetary control, although direct regulations on interest rates and credit have to a certain degree been retained or reintroduced. Other countries have undertaken quite determined reductions in direct controls in the last few years (notably Indonesia, the Philippines and, to a lesser extent, Malaysia, Sri Lanka and Thailand).

All the countries surveyed—except for the Philippines—have operated some form of exchange rate peg in recent years (Table 3). However, all countries have allowed a certain amount of exchange rate flexibility, and in most countries exchange controls have limited capital movements. Thus, some scope for independent monetary policy has remained, at least in the short run. Reflecting this, most countries have used monetary aggregates as intermediate targets and have also sought to achieve interest rate objectives for certain periods. Countries which have taken the most determined steps from a regulated toward a market-based monetary control system have adapted their intermediate targets. Following its financial reform, Indonesia changed from targeting domestic credit of the banking system to targeting base money and broad money. In Malaysia and Sri Lanka, the focus shifted from interest rates to broader monetary aggregates. Also, the Philippines shifted from targeting net domestic assets to targeting base money following the move to a floating exchange rate in late 1984.

Table 3.

The Policy Framework for Monetary Control in a Sample of Nine Developing Countries

Table 3.

The Policy Framework for Monetary Control in a Sample of Nine Developing Countries

Country	Exchange rate arrangement	Intermediate targets	Operating targets
Argentina	Austral. Introduced in June 1985, has in some periods been fixed vis-a-vis the U.S. dollar, in other periods adjusted on a continuous basis.	Monetary aggregates.	Interest rates on central bank bills, debt sales
Brazil	Cruzado. Frequent adjustments vis-a-vis the U.S. dollar, in line with relative inflation. Temporarily fixed against the U.S. dollar under the Cruzado Plan Feb.-Oct., 1986.	Monetary aggregates, real interest rates.	Base money, overnight interest rates
Indonesia	Rupiah. Managed float, basket of currencies used as indicator.	Monetary aggregates and monetary base. Domestic credit before October 1984.	Exchange rate, interest rates.
Kenya	Shilling. Adjustable peg to a basket of currencies. Managed flexibly with the aim of obtaining a suitable real effective exchange rate.	Money plus quasi-money (M2), real interest rates.	Bank credit, interest rates.
Malaysia	Ringgit. Determined on the basis of a basket of currencies.	Broad monetary aggregates.	Exchange rate, long-term interbank rates, recently overnight interbank rates.
Mexico	Peso. Two-tier system, with a free and controlled foreign exchange market. The rate in the controlled market depreciated daily, by predetermined amounts.	Monetary aggregates.	Interest rates, debt sales
Philippines	Peso. Floating. (Linked to U.S. dollar until October 1984.)	Base Money. (Until October 1984 net domestic assets of Central Bank.)	Sales and interest rates on treasury bills.
Sri Lanka	Rupee. Crawling peg, nominal rate adjusted for inflation differential with major trading partners.	Broad and narrow money	Net domestic assets of central bank. (Interest rates until early 1988.)
Thailand	Baht. Determined on the basis of a basket of currencies.	Multiple monetary targets are used as a guide for policy	Reserve money, but loosely implemented.

View Table

Table 3.

The Policy Framework for Monetary Control in a Sample of Nine Developing Countries

Country	Exchange rate arrangement	Intermediate targets	Operating targets
Argentina	Austral. Introduced in June 1985, has in some periods been fixed vis-a-vis the U.S. dollar, in other periods adjusted on a continuous basis.	Monetary aggregates.	Interest rates on central bank bills, debt sales
Brazil	Cruzado. Frequent adjustments vis-a-vis the U.S. dollar, in line with relative inflation. Temporarily fixed against the U.S. dollar under the Cruzado Plan Feb.-Oct., 1986.	Monetary aggregates, real interest rates.	Base money, overnight interest rates
Indonesia	Rupiah. Managed float, basket of currencies used as indicator.	Monetary aggregates and monetary base. Domestic credit before October 1984.	Exchange rate, interest rates.
Kenya	Shilling. Adjustable peg to a basket of currencies. Managed flexibly with the aim of obtaining a suitable real effective exchange rate.	Money plus quasi-money (M2), real interest rates.	Bank credit, interest rates.
Malaysia	Ringgit. Determined on the basis of a basket of currencies.	Broad monetary aggregates.	Exchange rate, long-term interbank rates, recently overnight interbank rates.
Mexico	Peso. Two-tier system, with a free and controlled foreign exchange market. The rate in the controlled market depreciated daily, by predetermined amounts.	Monetary aggregates.	Interest rates, debt sales
Philippines	Peso. Floating. (Linked to U.S. dollar until October 1984.)	Base Money. (Until October 1984 net domestic assets of Central Bank.)	Sales and interest rates on treasury bills.
Sri Lanka	Rupee. Crawling peg, nominal rate adjusted for inflation differential with major trading partners.	Broad and narrow money	Net domestic assets of central bank. (Interest rates until early 1988.)
Thailand	Baht. Determined on the basis of a basket of currencies.	Multiple monetary targets are used as a guide for policy	Reserve money, but loosely implemented.

Nearly all of the countries surveyed have made use of regular auctions—either for central bank or treasury bills—for monetary control purposes (Table 4). ^30/ Under these auctions, the public submits bids containing the volume demanded and a bid price to the central bank for treasury/central bank bills on auction, and the central bank accepts or rejects bids on the basis of a cut-off yield determined to be consistent with their desired operating target for monetary policy. The most widely used operating target has been short-term interest rates; other operating targets have included various liquidity measures (Argentina, Brazil, Mexico) and net domestic assets of the central bank (Sri Lanka). Some countries (Mexico, Philippines) have instituted formal limitations on the cut-off yield in the auctions, in order to promote orderly bidding and avoid large, short-run fluctuations in interest rates.

Table 4.

Instruments for Monetary Control in a Sample of Nine Developing Countries

Table 4.

Instruments for Monetary Control in a Sample of Nine Developing Countries

Country	Primary issues of securities	Instruments of daily money market management	Other market-based instruments	Refinance facilities	Direct controls	Reserve requirements
Argentina	Weekly auctions of participations in government papers in the form of central bank bills.	Repurchase/reverse repurchases, foreign exchange swaps.	—	Several, including subsidized facilities.	Interest rate regulations. Quotas on nonregulated rate deposits, acceptances and swaps.	Actively varied.
Brazil	Weekly auctions of central bank bills (and treasury bills, but mainly for budgetary finance).	Repurchases/reverse repurchases, and outright sales.	—	Several, including subsidized facilities.	Selective credit controls.	Not actively varied.
Indonesia	Auctions of central bank bills. The frequency of auctions has varied between daily and weekly.	Repurchase agreements and daily auctions of central bank bills.	—	Several, including subsidized facilities.	Interest rate and credit controls lifted as part of financial reform.	Not actively varied.
Kenya	Weekly auctions of treasury bills.	Discount window.	—	Several including discount window and O/D facility.	Maximum lending rates, minimum deposit rates, credit ceilings actively used.	Not actively varied for monetary control purposes.
Malaysia	Weekly auctions of treasury bills solely for government finance purposes.	Discount window, outright sales and purchases of government securities, repurchase/reverse repurchases, foreign exchange swaps, recycling of government deposits.	—	Several, including subsidized facilities.	Interest rates liberalized, selective credit controls remain in effect.	Not actively varied.
Mexico	Weekly auctions of treasury certificates.	Outright sales and purchases and repurchase/reverse repurchases.	Weekly auctions of fixed deposits and placements from the central bank.	…	Credit ceilings.	Actively varied.
Philippines	Weekly auctions of treasury bills.	Repurchase/reverse repurchases, foreign exchange swaps.	—	Single facility covering an array of purposes. Interest rate charged is linked to market interest rates.	Interest rates have been liberalized.	Actively varied.
Sri Lanka	Weekly or more frequent auctions of treasury bills.	Repurchase/reverse repurchases, foreign exchange swaps. Secondary window for sale and purchase of treasury bills.	—	Several, including subsidized facilities.	Interest rates liberalized. Some credit controls remain but are not strictly enforced.	Actively varied.
Thailand	Weekly auctions of treasury bills. (Mainly to satisfy demands for liquid assets; of limited importance for monetary management).	Repurchase/reverse repurchases; discount window.	Central bank bond issue.	Several, including subsidized facilities.	Interest rate ceilings, direct controls on credit last used in 1984.	Not actively varied.

View Table

Table 4.

Instruments for Monetary Control in a Sample of Nine Developing Countries

Country	Primary issues of securities	Instruments of daily money market management	Other market-based instruments	Refinance facilities	Direct controls	Reserve requirements
Argentina	Weekly auctions of participations in government papers in the form of central bank bills.	Repurchase/reverse repurchases, foreign exchange swaps.	—	Several, including subsidized facilities.	Interest rate regulations. Quotas on nonregulated rate deposits, acceptances and swaps.	Actively varied.
Brazil	Weekly auctions of central bank bills (and treasury bills, but mainly for budgetary finance).	Repurchases/reverse repurchases, and outright sales.	—	Several, including subsidized facilities.	Selective credit controls.	Not actively varied.
Indonesia	Auctions of central bank bills. The frequency of auctions has varied between daily and weekly.	Repurchase agreements and daily auctions of central bank bills.	—	Several, including subsidized facilities.	Interest rate and credit controls lifted as part of financial reform.	Not actively varied.
Kenya	Weekly auctions of treasury bills.	Discount window.	—	Several including discount window and O/D facility.	Maximum lending rates, minimum deposit rates, credit ceilings actively used.	Not actively varied for monetary control purposes.
Malaysia	Weekly auctions of treasury bills solely for government finance purposes.	Discount window, outright sales and purchases of government securities, repurchase/reverse repurchases, foreign exchange swaps, recycling of government deposits.	—	Several, including subsidized facilities.	Interest rates liberalized, selective credit controls remain in effect.	Not actively varied.
Mexico	Weekly auctions of treasury certificates.	Outright sales and purchases and repurchase/reverse repurchases.	Weekly auctions of fixed deposits and placements from the central bank.	…	Credit ceilings.	Actively varied.
Philippines	Weekly auctions of treasury bills.	Repurchase/reverse repurchases, foreign exchange swaps.	—	Single facility covering an array of purposes. Interest rate charged is linked to market interest rates.	Interest rates have been liberalized.	Actively varied.
Sri Lanka	Weekly or more frequent auctions of treasury bills.	Repurchase/reverse repurchases, foreign exchange swaps. Secondary window for sale and purchase of treasury bills.	—	Several, including subsidized facilities.	Interest rates liberalized. Some credit controls remain but are not strictly enforced.	Actively varied.
Thailand	Weekly auctions of treasury bills. (Mainly to satisfy demands for liquid assets; of limited importance for monetary management).	Repurchase/reverse repurchases; discount window.	Central bank bond issue.	Several, including subsidized facilities.	Interest rate ceilings, direct controls on credit last used in 1984.	Not actively varied.

The type of instrument used and the frequency of the auction have varied between the countries. Typically, the papers used for monetary control through auctions have been short-term, with maturities from 7 days to 12 months. In some cases, an inflationary environment has put downward pressure on maturities (Argentina, Mexico, Philippines). Most countries have made use of some form of central bank securities, but the use of government papers has also been quite widespread. Auctions have normally been held weekly, although the frequency has varied and even daily auctions have been held in one country (Indonesia). Mexico has also held weekly auctions of central bank offers to place or accept deposits to supplement the auctions of treasury bills.

In some countries, participation in the auctions has been formally restricted to financial institutions (Indonesia) or appointed dealers in the papers in question (Mexico, Philippines). Appointed dealers have sometimes been obliged to make secondary markets in these papers (Philippines).

The day-to-day operating procedures and the design of other supporting instruments have varied among countries. Some countries (such as Sri Lanka) have operated a secondary window for the purchase and sale of treasury bills at the latest auction rate with small margins between buy and sell rates; several countries operate a loan or rediscount window carrying varying degrees of penalty (for example, Kenya); and some countries have used foreign exchange swaps (Argentina, Philippines), and repurchase/reverse repurchase schemes for public sector papers (Argentina, Indonesia, Sri Lanka, Malaysia, Thailand) to manage money market conditions from day to day. Interest rates on most of these facilities are determined by the central bank. Some countries (most notably Indonesia and Philippines) have also reformed the central bank’s refinance facilities for priority sector borrowing, inter alia, in order to restrict the financial institutions’ access to central bank credit and to rationalize rediscount facilities. Despite the reforms, however, substantial elements of credit subsidization have remained in several countries. Moreover, reserve requirements have continued to be used, although less actively, in most of the countries; liquid asset ratios tend to be relatively high and banking competition is in general still weak. ^31/

b. Countries’ detailed operating procedures

a. Approach and limitations

Our ability to assess experiences with the market-based instruments is constrained by a number of factors. First, while certain effects on interest rates and monetary aggregates might be expected a priori following financial reforms—for example, the use of interest rates rather than direct credit controls for macro monetary management could require an increase in the average level of real interest rates—there were many influences on financial markets and the broader economy that were in some countries more important than the particular reform measures under consideration here. Indeed, a major problem is to distinguish the impact of the new procedures per se from larger influences operating on the economy. This difficulty is particularly acute given the relatively short period of time since the new procedures were implemented in several countries. ^47/

Second, the position prior to the introduction of the new procedures and the targets of monetary policy varied greatly between countries; therefore, similar experiences are not to be expected. In some countries, market interest rates were already free to vary and the active management of money market liquidity aimed at a stabilization of money market rates; in other countries, the liberalization of rates meant increased interest rate volatility. In some countries the operational target was the money market interest rate, which as a result followed a smoother path than in the countries which used the volume of debt sales or reserve money as the operational target.

Third, many aspects of financial reform, such as the authorities’ willingness to accept an adequate degree of interest rate flexibility, the development of active markets with wide participation, and—related to this—the building of institutions, technical infrastructure and human skills in the area of financial intermediation, are very difficult to assess. The experiences therefore have to be analyzed from a somewhat narrower perspective.

The majority of the countries surveyed have faced adverse external conditions in the 1980s, linked both to cyclical developments in the world economy and to weak terms of trade as well as problems in the conduct of economic policies domestically. These conditions have required changes in economic policies in order to contain domestic demand and, at the same time, to provide for sustainable economic growth. Several of the countries have implemented adjustment programs supported by Fund resources, involving a tightening of monetary and fiscal policies aimed at improving their balance of payments and reducing inflation. Related to these adjustment programs, and the subsequent improvements in external positions in some countries, there have been marked shifts in nominal and real interest rates. Short-term interest rates increased abruptly in Brazil in late 1984 and again in 1986, in the Philippines in 1984 and 1985, and in Indonesia in 1984. In some cases (Indonesia, the Philippines) this coincided with the implementation of new monetary control techniques. On the other hand, interest rates have fallen in Thailand and Malaysia in response to improved external positions.

In subsection b. below, the role of monetary policy and the operational role of market-oriented instruments in achieving monetary policy objectives are examined within a more general macroeconomic policy context. This section presents tables showing the counterparts to the change in reserve money along the lines of the reserve money identity developed in section 4a. The main counterparts are the governments’ domestic financing requirements less sales of government debt outside the central bank, including auctions of treasury bills where appropriate; the increase (on a transactions basis) of the central bank’s net holdings of foreign assets; autonomous net lending by the central bank related to preferential rediscount and special liquidity support measures; and the central bank’s discretionary use of money market instruments such as repurchases, foreign exchange swaps, central bank bills, etc. for money management purposes. Market-oriented instruments of monetary control would be represented both by the central bank’s use of these discretionary instruments and the line for treasury bill sales (where appropriate). Given this information on the quantitative use of market-oriented instruments, subsection c. focuses on the results for short-term interest rates and monetary aggregates following the introduction of new operating procedures.

b. Reserve money developments and counterparts

In Argentina and Brazil, monetary instruments have been developed against the background of large government domestic financing needs (Appendix Tables 1 and 2). Despite the change in monetary instruments in Argentina in 1985 and Brazil in 1986, the authorities did not use the new instruments to exert additional control over the growth of reserve money. In Brazil the Central Bank increased its uptake of government debt at the same time as it began issuing central bank bills. It appears that the austral and cruzado plans, which relied to a large extent on de-indexation of financial assets and wages, lacked sustained or substantive support from fiscal and monetary policies. Real interest rates in Argentina and Brazil remained generally negative on average in the period after the plans were introduced and in Brazil reserve money growth continued at a rapid pace.

In Indonesia nominal and real interest rates increased significantly following the financial liberalization in 1983 and as part of their concerted adjustment efforts (Chart 1). The major monetary disturbances have been associated with instability in the balance of payments and domestic financing of the government budget deficit (Appendix Table 3). Up until 1987, the reserve money effects of the changes in foreign exchange reserves were largely offset by the central bank’s net purchases of government debt; however, increasing use was also made of open-market-type operations, especially in 1986 and 1987. As the liquidity situation tightened in late 1986 and early 1987, the central bank supplied liquidity through central bank bills. Subsequently, the central bank allowed interest rates to increase, and in the year to March 1988, capital reflows contributed to an increase in reserve money, which was partly offset by central bank operations.

Indonesia Developments In Interest Rates And Monetary Aggregates

(In percent)

Citation: IMF Working Papers 1989, 048; 10.5089/9781451972160.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Indonesia Developments In Interest Rates And Monetary Aggregates

(In percent)

Citation: IMF Working Papers 1989, 048; 10.5089/9781451972160.001.A001

• Download Figure
• Download figure as PowerPoint slide

Indonesia Developments In Interest Rates And Monetary Aggregates

(In percent)

Citation: IMF Working Papers 1989, 048; 10.5089/9781451972160.001.A001

• Download Figure
• Download figure as PowerPoint slide

In Kenya, the impact of the domestic financing needs of the Government on reserve money has been moderated through the domestic sale of treasury bills and bonds outside the Central Bank (Appendix Table 4). However, monetary policy has not been active. Nominal interest rates have been maintained at a stable level while real interest rates have fluctuated. In 1984, the authorities did not sterilize the external surpluses and reserve money growth increased sharply. In 1987, a substantially greater proportion of the larger government domestic financing requirement was financed by the Central Bank, and, as a result, reserve money grew very rapidly.

Market-oriented monetary instruments have been increasingly utilized in Malaysia. In 1986 and 1987 such instruments were used largely to offset the accumulation of foreign exchange by the central bank and to moderate downward pressure on interest rates (Chart 2). The central bank has also been able to maintain a fairly stable reserve money growth (Appendix Table 5). The main instruments of money market management have been foreign exchange swaps, “recycling” of government deposits, secondary transactions in government securities, and, beginning in 1987, issues of central bank bills. The central bank has been able to restrict its domestic financing of the Government, although these financing needs have been increasing; the Government has met its domestic financing requirement largely through sales of securities.

Malaysia Developments In Interest Rates And Monetary Aggregates

(In percent)

Citation: IMF Working Papers 1989, 048; 10.5089/9781451972160.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Malaysia Developments In Interest Rates And Monetary Aggregates

(In percent)

Citation: IMF Working Papers 1989, 048; 10.5089/9781451972160.001.A001

• Download Figure
• Download figure as PowerPoint slide

Malaysia Developments In Interest Rates And Monetary Aggregates

(In percent)

Citation: IMF Working Papers 1989, 048; 10.5089/9781451972160.001.A001

• Download Figure
• Download figure as PowerPoint slide

As in Argentina and Brazil, monetary policy in Mexico has been implemented against the background of large government domestic financing needs and instability in the external position. The auctions of CETES have financed only a small percentage of the Government’s domestic financing needs and a substantial percentage of government domestic borrowing has been provided through the central bank (Appendix Table 6). Auctions of deposits with the central bank have reduced financial sector liquidity, but have been insufficient to prevent increases in the growth rate of reserve money.

In the Philippines. market-oriented monetary instruments have been used actively in macroeconomic management. In 1984, nominal interest rates were increased sharply (see Chart 3), influenced by sales of central bank bills aimed at moderating the expansion of reserve money (Appendix Table 7). Slippages occurred in 1986 and there was a substantial loss of foreign exchange reserves; however, in 1987 monetary instruments, especially the weekly auctioning of treasury bills and the associated increase of government deposits at the Central Bank in special sterilized accounts were used actively to constrain the growth of reserve money. The treasury bill sales and the sterilization of proceeds more than offset the maturing of central bank bills which were being phased out as a monetary instrument.

Philippines Developments In Interest Rates And Monetary Aggregates

(In percent)

Citation: IMF Working Papers 1989, 048; 10.5089/9781451972160.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Philippines Developments In Interest Rates And Monetary Aggregates

(In percent)

Citation: IMF Working Papers 1989, 048; 10.5089/9781451972160.001.A001

• Download Figure
• Download figure as PowerPoint slide

Philippines Developments In Interest Rates And Monetary Aggregates

(In percent)

Citation: IMF Working Papers 1989, 048; 10.5089/9781451972160.001.A001

• Download Figure
• Download figure as PowerPoint slide

There has been a deterioration in economic conditions in Sri Lanka since 1984, with a loss of foreign exchange reserves and widening government domestic financing needs (Appendix Table 8). Sales of central bank bills were used in 1984-86 to offset the increase in financial sector liquidity associated with the accumulation of foreign assets in 1984 and the uptake of domestic debt by the Central Bank in 1985 and 1986. In 1987, the maturing of central bank bills offset the impact of the loss of foreign exchange reserves on the growth of reserve money; thus, monetary policy was geared toward lowering interest rates, and real interest rates became negative. The adoption of the net domestic assets target required increased interest rate flexibility and a less accommodating stance for monetary policy. In early 1988 interest rates in the treasury bill auction were allowed to increase.

Until 1986, the monetary authorities in Thailand did not pursue an active policy with regard to managing money market liquidity. However, following the substantial improvement in the external balance beginning in 1986 and an associated build-up in financial sector liquidity, the authorities actively drained liquidity through government bond repurchases and issues of central bank bonds (Appendix Table 9). These operations have not been sufficient, however, to prevent a sharp increase in the growth of reserve money in 1987 and a fall in interest rates (Chart 4).

Thailand Developments In Interest Rates And Monetary Aggregates

(In percent)

Citation: IMF Working Papers 1989, 048; 10.5089/9781451972160.001.A001

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Thailand Developments In Interest Rates And Monetary Aggregates

(In percent)

Citation: IMF Working Papers 1989, 048; 10.5089/9781451972160.001.A001

• Download Figure
• Download figure as PowerPoint slide

Thailand Developments In Interest Rates And Monetary Aggregates

(In percent)

Citation: IMF Working Papers 1989, 048; 10.5089/9781451972160.001.A001

• Download Figure
• Download figure as PowerPoint slide

c. Implications for interest rates and monetary aggregates

To examine the impact of the new operating procedures on interest rates and monetary aggregates, we first specified for each country a date for the introduction of the new operating procedures (see Table 5, footnote lc). These dates were selected on the basis of our judgment about when a substantive shift occurred in operating procedures, rather than necessarily the date on which reforms were initiated. For the pre - and post-reform periods, we conducted simple statistical tests, comparing monthly data on nominal and real interest rates, and on twelve-month growth rates of reserve money and broad money. The tests include a comparison of means and variances of the data series and the residuals from autoregressive regressions for nominal and real interest rates in the sample sub-periods. ^48/ Data availability has restricted this quantitative analysis to only four countries: Indonesia, Malaysia, Philippines and Thailand. ^49/

Table 5.

Financial Reform and Structural Changes in Interest Rate Determination ^1/

Sources: IFS. data provided by the authorities, and staff estimates.

^1/Note on methodology:

a. Data. Monthly money market interest rates [IFS. line 60b, except for Philippines (60c)]; deflated by change in CPI [IFS. line 64] over previous 12-month period.

b. Tests. Standard t and F tests are carried out to check whether there have been significant changes in interest rate levels and variability, respectively, following monetary control “reform.” Regressions of the form X_t = aX_t–1 (X_t is interest rate in period t) are run for pre- and post-reform periods, and F-tests are used to gauge whether there have been structural changes in interest rate determination.

c. The nature and time of reform, and pre- and post-reform periods for each country were:

			Sample periods
Country	Reform measure(s)	Time	Pre-Reform	Post-Reform
Indonesia	Comprehensive financial reform initiated	June 1983	78:1-83:5	83:6-88:5
Malaysia	Freer money market interest rate determination	February 1987	82:2-87:1	87:2-88:1
Philippines	Floating of currency	October 1984	78:1-84:9	84:10-88:8
Thailand	More active use of repurchase transactions initiated	May 1986	78:1-86:4	86:5-88:2

^2/One asterisk indicates that the change has been significant on the 5 percent level, two asterisks on the 1 percent level.

^3/Residual Sum of Squares in the regression, divided by the degree of freedom.

^4/Nominal interest rate was unchanged during the whole pre-reform period.

Table 5.

Financial Reform and Structural Changes in Interest Rate Determination ^1/

		Nominal Interest Rate				Real Interest Rate
		Pre-Reform	Post-Reform	Change	Test Statistics 2/	Pre-Reform	Post-Reform	Change	Test Statistics 2/
Indonesia
	Mean (percent)	6.00	15.56	+	t=17.51**	-7.57	7.27	+	t=13.30**
	Standard deviation	—	4.23	+	**4/	6.02	6.43	+	F=1.19
	RSS 3/	—	2.8	+	**4/	2.6	3.7	+	F=1.45
Malaysia
	Mean (percent)	8.07	3.35	-	t=16.03**	5.29	2.21	-	t=7.66**
	Standard deviation	1.95	0.53	-	F=13.54**	2.86	0.55	-	F=5.20**
	RSS 3/	3.6	0.2	-	F=17.76**	4.2	0.54	-	F=10.30**
Philippines
	Mean (percent)	16.55	20.18	+	t=2.82**	-0.02	7.42	+	t=3.96**
	Standard deviation	4.18	8.24	+	F=3.89**	9.99	10.38	+	F=1.08
	RSS 3/	0.87	4.0	+	F=5.71**	4.5	7.9	+	F=1.76*
Thailand
	Mean (percent)	13.58	6.43	-	t=19.02**	6.50	2.75	-	t=3.99**
	Standard deviation	2.53	0.96	-	F=6.95**	5.75	3.48	-	F=2.73**
	RSS 3/	1.2	0.8	-	F=1.46	2.4	0.8	-	F=2.93**

Sources: IFS. data provided by the authorities, and staff estimates.

^1/Note on methodology:

a. Data. Monthly money market interest rates [IFS. line 60b, except for Philippines (60c)]; deflated by change in CPI [IFS. line 64] over previous 12-month period.

b. Tests. Standard t and F tests are carried out to check whether there have been significant changes in interest rate levels and variability, respectively, following monetary control “reform.” Regressions of the form X_t = aX_t–1 (X_t is interest rate in period t) are run for pre- and post-reform periods, and F-tests are used to gauge whether there have been structural changes in interest rate determination.

c. The nature and time of reform, and pre- and post-reform periods for each country were:

			Sample periods
Country	Reform measure(s)	Time	Pre-Reform	Post-Reform
Indonesia	Comprehensive financial reform initiated	June 1983	78:1-83:5	83:6-88:5
Malaysia	Freer money market interest rate determination	February 1987	82:2-87:1	87:2-88:1
Philippines	Floating of currency	October 1984	78:1-84:9	84:10-88:8
Thailand	More active use of repurchase transactions initiated	May 1986	78:1-86:4	86:5-88:2

^2/One asterisk indicates that the change has been significant on the 5 percent level, two asterisks on the 1 percent level.

^3/Residual Sum of Squares in the regression, divided by the degree of freedom.

^4/Nominal interest rate was unchanged during the whole pre-reform period.

View Table

Table 5.

Financial Reform and Structural Changes in Interest Rate Determination ^1/

		Nominal Interest Rate				Real Interest Rate
		Pre-Reform	Post-Reform	Change	Test Statistics 2/	Pre-Reform	Post-Reform	Change	Test Statistics 2/
Indonesia
	Mean (percent)	6.00	15.56	+	t=17.51**	-7.57	7.27	+	t=13.30**
	Standard deviation	—	4.23	+	**4/	6.02	6.43	+	F=1.19
	RSS 3/	—	2.8	+	**4/	2.6	3.7	+	F=1.45
Malaysia
	Mean (percent)	8.07	3.35	-	t=16.03**	5.29	2.21	-	t=7.66**
	Standard deviation	1.95	0.53	-	F=13.54**	2.86	0.55	-	F=5.20**
	RSS 3/	3.6	0.2	-	F=17.76**	4.2	0.54	-	F=10.30**
Philippines
	Mean (percent)	16.55	20.18	+	t=2.82**	-0.02	7.42	+	t=3.96**
	Standard deviation	4.18	8.24	+	F=3.89**	9.99	10.38	+	F=1.08
	RSS 3/	0.87	4.0	+	F=5.71**	4.5	7.9	+	F=1.76*
Thailand
	Mean (percent)	13.58	6.43	-	t=19.02**	6.50	2.75	-	t=3.99**
	Standard deviation	2.53	0.96	-	F=6.95**	5.75	3.48	-	F=2.73**
	RSS 3/	1.2	0.8	-	F=1.46	2.4	0.8	-	F=2.93**

Sources: IFS. data provided by the authorities, and staff estimates.

^1/Note on methodology:

a. Data. Monthly money market interest rates [IFS. line 60b, except for Philippines (60c)]; deflated by change in CPI [IFS. line 64] over previous 12-month period.

b. Tests. Standard t and F tests are carried out to check whether there have been significant changes in interest rate levels and variability, respectively, following monetary control “reform.” Regressions of the form X_t = aX_t–1 (X_t is interest rate in period t) are run for pre- and post-reform periods, and F-tests are used to gauge whether there have been structural changes in interest rate determination.

c. The nature and time of reform, and pre- and post-reform periods for each country were:

			Sample periods
Country	Reform measure(s)	Time	Pre-Reform	Post-Reform
Indonesia	Comprehensive financial reform initiated	June 1983	78:1-83:5	83:6-88:5
Malaysia	Freer money market interest rate determination	February 1987	82:2-87:1	87:2-88:1
Philippines	Floating of currency	October 1984	78:1-84:9	84:10-88:8
Thailand	More active use of repurchase transactions initiated	May 1986	78:1-86:4	86:5-88:2

^2/One asterisk indicates that the change has been significant on the 5 percent level, two asterisks on the 1 percent level.

^3/Residual Sum of Squares in the regression, divided by the degree of freedom.

^4/Nominal interest rate was unchanged during the whole pre-reform period.

(2) Monetary aggregates

As regards the likely outcome for monetary aggregates, we would anticipate, ceteris paribus, an inverse relationship between nominal interest rates and the growth of non-interest-bearing reserve money. This expectation is supported in the results for three of the countries (Table 6). The relationship between interest rates and broad money is a priori complicated since a part of broad money is interest bearing and could expand with a general rise in interest rates; the relevant interest rate term for broad money demand is therefore the differential between the own rate on money and the interest rate on nonmonetary assets. In addition to interest rates, the demand for reserve and broad money would be affected by developments in income or wealth; however, high frequency data on these variables are not available and have not been taken into account.

Table 6.

Financial Reform and the Growth Rates of Monetary Aggretates ^1/

Sources: IFS and staff estimates.

^1/Twelve-month growth rates. Note on methodology:

• a.Data. Monetary observations: Reserve money as reported in IFS (line 14); broad money the sum of money and quasi-money in IFS (lines 34 + 35).

• b. Tests. Standard t and F tests are carried out to check whether there have been significant shift in money growth and variability, respectively.

• c. The nature and time of reform (see footnote lc, Table 5).

^2/One asterisk indicates that the change has been significant on the 5 percent level, two asterisks on the 1 percent level.

Table 6.

Financial Reform and the Growth Rates of Monetary Aggretates ^1/

		Reserve Money				Broad Money
		Pre-Reform	Post-Reform	Change	Test Statistics 2/	Pre-Reform	Post-Reform	Change	Test Statistics 2/
Indonesia
	Mean (percent)	19.7	16.7	-	t=1.56	28.5	24.0	-	t=3.07*
	Standard deviation	13.3	7.7	-	F=2.98**	10.8	4.6	-	F=5.51**
Malaysia
	Mean (percent)	7.6	3.0	-	t=4.49**	11.6	7.4	-	t=3.21**
	Standard deviation	4.9	2.8	-	F=3.06*	3.8	3.4	-	F=1.25
Philippines
	Mean (percent)	21.2	20.3	-	t=0.36	20.1	13.1	-	t=8.63**
	Standard deviation	18.5	10.0	-	F=3.42**	4.8	4.2	-	F=1.31
Thailand
	Mean (percent)	11.3	13.9	+	t=2.16**	19.2	14.4	-	t=6.96**
	Standard deviation	4.7	5.2	+	F=1.22	3.8	2.7	—	F=1.98*

Sources: IFS and staff estimates.

^1/Twelve-month growth rates. Note on methodology:

• a.Data. Monetary observations: Reserve money as reported in IFS (line 14); broad money the sum of money and quasi-money in IFS (lines 34 + 35).

• b. Tests. Standard t and F tests are carried out to check whether there have been significant shift in money growth and variability, respectively.

• c. The nature and time of reform (see footnote lc, Table 5).

^2/One asterisk indicates that the change has been significant on the 5 percent level, two asterisks on the 1 percent level.

View Table

Table 6.

Financial Reform and the Growth Rates of Monetary Aggretates ^1/

		Reserve Money				Broad Money
		Pre-Reform	Post-Reform	Change	Test Statistics 2/	Pre-Reform	Post-Reform	Change	Test Statistics 2/
Indonesia
	Mean (percent)	19.7	16.7	-	t=1.56	28.5	24.0	-	t=3.07*
	Standard deviation	13.3	7.7	-	F=2.98**	10.8	4.6	-	F=5.51**
Malaysia
	Mean (percent)	7.6	3.0	-	t=4.49**	11.6	7.4	-	t=3.21**
	Standard deviation	4.9	2.8	-	F=3.06*	3.8	3.4	-	F=1.25
Philippines
	Mean (percent)	21.2	20.3	-	t=0.36	20.1	13.1	-	t=8.63**
	Standard deviation	18.5	10.0	-	F=3.42**	4.8	4.2	-	F=1.31
Thailand
	Mean (percent)	11.3	13.9	+	t=2.16**	19.2	14.4	-	t=6.96**
	Standard deviation	4.7	5.2	+	F=1.22	3.8	2.7	—	F=1.98*

Sources: IFS and staff estimates.

^1/Twelve-month growth rates. Note on methodology:

• a.Data. Monetary observations: Reserve money as reported in IFS (line 14); broad money the sum of money and quasi-money in IFS (lines 34 + 35).

• b. Tests. Standard t and F tests are carried out to check whether there have been significant shift in money growth and variability, respectively.

• c. The nature and time of reform (see footnote lc, Table 5).

^2/One asterisk indicates that the change has been significant on the 5 percent level, two asterisks on the 1 percent level.

In Indonesia and the Philippines, the significant increases in interest rates and their variability after the introduction of monetary reforms were associated with significant reductions in the variability of the growth of reserve money and reductions in its average rate of growth, but the latter tests were insignificant at a 5 percent confidence level. This could suggest that other factors, such as developments in income, are important in addition to the level of interest rates in determining the trend in reserve money growth. Following the financial reforms, there were significant reductions in the variability and rate of growth of broad money in Indonesia. In the Philippines the average rate of growth of broad money fell significantly following the introduction of the new monetary policy operating procedures.

In Thailand, there was also an inverse relationship between interest rates and reserve money. The average rate of growth and the variability of the growth rate of reserve money increased while average interest rates and their variability declined following the introduction of the more active money market policy of the central bank. However, the effects on the variability of reserve money growth were not significant, which may again suggest that the change in Thai operating procedures was not particularly active. There is also other evidence to suggest that this was the case, such as the fact that the Thai authorities placed limits on the volume of their money market intervention. In contrast, the growth rate and the variability of the growth rate of broad money fell significantly in the post-reform period.

Despite the reduction in interest rates and their variability in the post-reform period in Malaysia, both the growth rate and the variability of the growth rate of reserve money fell significantly—results which are anomalous to the simple inverse relationship hypothesis between interest rates and the growth of reserve money. However, the Malaysian post-reform data period, which begins in February 1987, is the shortest in our sample. These results also highlight the importance of taking into account factors other than interest rates in the determination of the path of reserve money and the complexity of the response of monetary variables to interest rates.