6 Speeding Up by Slowing Down: A Market-Building Approach to Financial Sector Reform

Laura Wallace
Published Date:
May 1997
  • ShareShare
Show Summary Details
David C. Cole and Betty F. Slade

Given the many serious disruptions and financial crises that have been associated with recent attempts to modernize, privatize, and market-orient the financial systems in African countries, it would seem appropriate to consider some alternative approaches that focus on improving the basic components of the existing systems and moving toward the desired objectives over one or two decades rather than a few years. This paper outlines a more gradual, market-building approach to four major aspects of financial development: managing monetary policy; strengthening prudential regulation; improving financial services in rural areas; and promoting capital markets.

Managing Monetary Policy

Developing countries have recently been advised and encouraged to adopt market-based systems of monetary management—using open market operations with treasury bills or similar instruments to control the supply of reserve money, to influence interest rates, and to replace direct controls over credit allocation.1 This market-based approach is seen as preferable to having monetary authorities set credit ceilings or guidelines, as well as interest rate ceilings and rate structures, that commonly fail to keep pace with changes in the rate of inflation or to reflect differences in credit quality or usage.

A number of countries have attempted to implement such market-based systems, but have encountered serious problems due to lack of fiscal discipline, absence of efficient, competitive markets for the new monetary instruments, and generally weak infrastructures of the financial system. Consequently, real interest rates have become unreasonably high or unstable, the monetary authorities have found it difficult to control key monetary variables, and, while some institutions and individuals have been hit hard, others have been able to exploit the new systems and markets to their own advantage.

Interest Rates and Exchange Rates

A transitional and less demanding alternative to trying to develop a “market-determined” level and structure of reference interest rates would be for the central bank to set fairly broad ranges for bank interest rates, and let individual banks set their own rates for different types of deposits and loans within the prescribed ranges. In the absence of a market-deter mined reference rate, the best available guide for interest rates is likely to be the recent rate of inflation, as measured by a consumer price index. The range for allowable bank deposit rates could be, for example, 0–10 percentage points above the national inflation rate of the most recent 3–6 months. A similar range of allowable bank loan rates should be prescribed, somewhat above the range for deposit rates, to allow for intermediation spreads (e.g., 5–20 percentage points above the inflation rate of recent months). Permitting banks to set their own rates for deposits and loans would give the banks an incentive to analyze their markets and costs of operations and to begin to explore strategies for increasing their profitability, important steps in preparing them for more market-based operations.

Similarly, instead of having a market-determined exchange rate, the central bank could adopt a crawling peg exchange rate linked to the differential rates of inflation of the home country and its main trading partners over the past 6–12 months. The central bank could set bands around the central rate at which it would be prepared to buy or sell foreign exchange. The central bank could widen the bands over time to allow more room for the foreign exchange market participants to trade with each other and to determine the daily rate within the bands.

Supply of Reserve Money

Rather than attempting a rapid transition from reliance on direct credit controls to use of open market operations for monetary management, the central bank could rely on the methods used in most developed countries over many decades—that is, control of the supply of reserve or base money and setting of legal reserve requirements as the primary constraint on total banking system credit and monetary expansion. A new approach to legal reserve requirements that distinguishes among the solvency, liquidity, and clearing roles of such reserves is outlined below in the discussion of prudential regulation.

For reserve money management, instead of open market operations or treasury bill auctions, the central bank could use much simpler instruments that have a direct impact on the supply of legal reserves and do not require elaborate new market instruments and mechanisms. For example, the central bank could participate directly in the interbank market, if one exists, by borrowing reserves from, or lending reserves to, individual banks at approximately the going interbank rates. It should avoid being a continuous borrower or a continuous lender of reserve funds in the interbank market because this would discourage development of trading between banks. Instead it should generally enter the market in an active way when it wishes to influence the level of reserve money and put some pressure on liquidity and interest rates. If an active interbank market does not exist, the central bank may be able to help get one started by offering to borrow or lend short-term funds from and to banks.

Through such participation in the interbank market, the central bank would not only be able to monitor directly the liquidity position of the banking system and of individual banks but also would be aware of pressures on the interbank interest rates. Utilization of the interbank market to manage the supply of reserve money, if done properly, would help to develop that market and provide an alternative to premature attempts to introduce bill auctions and the complex mechanisms required for trading such paper.

The central bank could also have authority to transfer deposits of government, or some government agencies, between the commercial banks and the central bank as a means of controlling the supply of reserve money. Such transfers could be done on a daily basis and distributed across the whole banking system in some equitable manner (as is the practice in Canada), or they could be done only on certain occasions when the other interbank and foreign exchange market instruments were not sufficient to achieve the needed adjustments in the supply of reserve money. Transfers of government deposits or leaving government deposits in individual banks should not, however, be used as a mechanism for subsidizing or penalizing individual banks.

The central bank should also have a rediscount window for banks that, for various reasons, could not be accommodated through the interbank market. For short-term borrowings the rediscount rate should be somewhat above the interbank rate, but it should rise steeply for longer-term users to discourage such demands.

The central bank could, and should, also participate directly in the foreign exchange market, as a buyer and seller at its own initiative and within the bands, in an effort to influence both the supply of reserves of the banking system and the exchange rate.

This approach to monetary management is designed to move the banking system in the direction of more market-oriented operations in gradual steps rather than quickly. It would provide room for increased bank discretion over both interest rates and sources and uses of funds, but within limits that were controlled by the monetary authorities. It would also give the central bank instruments for controlling the supply of reserve money that were not dependent on fully functioning money and capital markets. At the same time, it would open the way for developing money markets—including foreign exchange—which would subsequently facilitate use of indirect monetary policy instruments. Finally, it would give the banks more opportunity and experience in operating in such markets and eventually build the expertise to do so profitably and without excessive risk.

Strengthening Prudential Regulation

Elimination of interest rate and credit controls, especially if accompanied by either implicit or explicit government guarantees of bank liabilities and weak prudential supervision, has contributed to institutional failures and systemic crises in both developed and developing countries in recent years. A number of responses have been tried or proposed for resolving these problems, including strengthening prudential supervision, limiting the range of liabilities covered by deposit insurance or other government guarantees, limiting the kinds of assets that banks can hold to relatively low-risk instruments, and increasing required bank capital and linking it to the riskiness of the bank’s assets. All of these suggestions have potential advantages, but they also have limitations and generally are very difficult to implement and enforce, especially in less-developed financial markets.2

There appears to be general agreement that prudential supervision of banks should be based upon active monitoring of the assets and the capital of individual banks by the central bank, or other designated agency, and that the regulatory system should be designed to provide incentives for prudential behavior by bank owners, managers, and regulators. The primary instrument for achieving this in many countries is the minimum required ratio of capital to risk-weighted assets. The more advanced countries have been working almost continuously to improve the original guidelines contained in the Basle Capital Accord of July 1988 so that the guidelines can better fit the requirements of today’s increasingly complex financial markets.3 However, the fundamental and very serious problem for many developing countries is not more sophisticated rules, but rather their very limited capability for supervising financial institutions and the related deficiencies in financial infrastructure (e.g., accounting, legal, and communications systems).

In developing countries, it is very difficult to monitor the quality of a bank’s assets. It is also difficult to be sure that a bank has any real capital or even to be sure that it has liquid reserves to meet a sudden withdrawal of deposits. When a bank starts running into difficulty, the bank managers often attempt to hide the problems not just from the regulators but also from the depositors, and often even from the owners. The first sign of problems that the regulators are likely to see is when the bank is unable to meet its daily clearing obligations. When that happens, it usually means that the bank has already exhausted its liquid reserves and probably much of its capital.

One way to try to prevent this kind of liquidity and solvency crisis is to recast the traditional legal reserve requirement into three components that are specifically designed to meet the three separable objectives: solvency, liquidity, and clearing of daily settlements.

  • Solvency is a medium- to long-run requirement for capital, or net worth, sufficient to meet all liabilities after liquidating assets.
  • Liquidity is a short-run requirement for liquid assets that can be sold or pledged to obtain cash to meet temporary, adverse withdrawals of funds.
  • Clearing balances are the normal working balances held at the central bank to meet daily settlements in the clearinghouse.

A powerful way to reinforce both the capital and liquidity requirements for banks would be to require banks to hold a solvency reserve and a liquidity reserve at the central bank. The solvency reserve should be linked to the bank’s capital, while the liquidity reserve should be linked to the bank’s deposit liabilities. The solvency reserve requirement might be equal to 50 percent of the bank’s capital. The liquidity reserve requirement might be 10 percent of total deposit liabilities. Both reserve accounts could consist of domestic and/or foreign government securities that would be held by the central bank on behalf of the individual bank, which would receive the interest income derived from the securities. Both the solvency and liquidity reserves would be totally separate and distinct from a bank’s required (clearing) reserve deposits at the central bank.

The solvency reserve would only be available for use by the central bank to help meet the obligations of the depositing bank in the event of closure and liquidation, or as part of a merger or acquisition of the bank. Holding the solvency reserve at the central bank would not prevent a bank from becoming insolvent, but it would at least give the central bank some good assets to help offset that insolvency. One objective of supervision would be to try to prevent a bank from reaching a level of insolvency such that the solvency reserve would not be able to meet the claims of creditors and depositors if liquidation became necessary. The required solvency reserve for individual banks could even be increased if, upon examination, it appeared that the bank was getting into difficulty. An increase in the required solvency reserve ratio would force a bank to transfer some securities from its liquidity reserve to its solvency reserve and then make adjustments in its assets and liabilities to bring its liquidity reserve back up to the required level.

The solvency reserve system is an alternative to, but very different from, deposit insurance. The main difference is that the solvency reserve is linked to the individual bank and is a partial guarantee of that bank’s liabilities to its creditors. It is not a payment into a fund that serves to guarantee the claims of depositors at all participating banks, as insurance does.

After the end of each year, banks would be required to adjust their solvency reserve at the central bank to maintain the specified ratio. The central bank could withhold interest payments on the deposited securities if necessary to bring the reserve up to the required level.

The liquidity reserve is intended to meet extraordinary liquidity needs of a bank. Normal liquidity needs should be met out of other liquid assets that are held for such purposes and are not part of the formal liquidity reserve.

Each bank should be required to maintain the required balance in the liquidity reserve on average over a period of perhaps one month. It should be permitted to draw out up to perhaps 50 percent of the liquidity reserve for short periods of up to 15 days, but failure to bring the liquidity reserve account back up to the required level within that period or to maintain the average monthly requirement should be subject to severe penalties, including withholding of interest earnings on the securities in the account. Drawing the liquidity reserve down below 50 percent of the required level should require prior discussion with, and approval by, the central bank.

The clearing reserve is the clearing balance that banks need to maintain at the central bank to meet their clearing obligations at the end of the day. For purposes of uniformity, it may be desirable to set a minimum required clearing reserve of perhaps 2 percent of all third-party liabilities. Banks should be required to maintain this minimum on average over a two-week period. They should be subject to penalties if the average falls below the minimum. They should also be prohibited from having a negative clearing reserve at the end of any day. It would be much easier to have accurate information on each bank’s clearing reserve position if such reserves were defined as deposits at the central bank and did not include vault cash or any other items.

Introduction of the solvency and liquidity reserve requirements could be facilitated for existing banks by a concurrent reduction in the current legal reserve requirement that would free up an approximately equivalent amount of reserve deposits that could be transferred to the solvency and liquidity reserve accounts. Most developing countries have relatively high required reserve ratios. A reduction in these ratios would be warranted in its own right to reduce the amount of non-earning assets of the banks. Transferring the funds to the solvency and liquidity reserve accounts would turn them into earning assets, but it would not result in any immediate change in the liquidity of the banking system that might call for neutralizing action by the central bank. Once the initial adjustment had been made, banks would henceforth be required to maintain their several reserve deposit accounts on a continuing basis.

New banks should be required to meet the established solvency, liquidity, and clearing reserve ratios from the inception of their operations. Initially the solvency reserve would account for a relatively large portion of the new bank’s total assets, but as third-party liabilities and other earning assets grew, the solvency reserve would diminish in relative importance until it stabilized at about 5–6 percent of the bank’s total assets. The solvency reserve might be increased for individual banks to 10 percent or even 20 percent of total assets if those banks were perceived to be heading for insolvency. The liquidity and clearing reserve deposits, on the other hand, would initially be a small share of total assets but would then rise, as third-party liabilities grew, to become a relatively constant share of total assets commensurate with the required reserve ratios.

The suggested system of solvency and liquidity reserves is not a substitute for bank supervision, but rather an attempt to reinforce effective supervision and at the same time to provide an alternative to deposit insurance that is less susceptible to moral hazard and exploitation by unscrupulous bank owners and managers. Such reserve requirements should help to open the way for less restricted entry of new banks and expansion of existing banks that might be expected to occur with the removal of direct controls over banking activities. If some banks, especially state-owned banks, could not be expected to conform to the new decontrolled system, they should be excluded from the privileges and requirements of that system and treated as special cases.

Improving Financial Services in Rural Areas

The financial services that urban-based officials tend to believe are most needed in rural areas are production credits and means of making payments between rural and urban areas. The experience of many developing countries that have viable rural financial institutions, both formal and informal, is that the main need is for a secure form of savings and, secondarily, for short-term consumption credits. Payment services are more often needed by larger traders and by urban residents who wish to transfer funds to their relatives in the rural areas. Production credits are often part of government-sponsored programs and have generally been shown to be unnecessary to gain acceptance of “good” programs. Moreover, they are likely to benefit the better-off rural inhabitants and those who manage the programs, rather than the rural poor.

The primary focus of rural financial service improvement programs should, therefore, be on improving savings facilities. This entails making such services convenient, accessible, and safe. Whether such savings facilities should be provided by an entity that also extends credit is a difficult issue. It is the extension of credit that generally leads to nonperforming loans and inability to reimburse depositors. If safety, convenience, and liquidity of deposits are the primary objective, then only those deposit-mobilizing institutions that have a viable capability for administering loan programs effectively should be permitted to use the depositors’ funds to make such loans. Otherwise the funds should be invested in other ways that would assure safety and a reasonable return. A central bank that wished to promote such deposit-mobilizing institutions could even consider accepting deposits from those institutions and paying them a positive real rate of interest.

The postal service is an entity that might be encouraged to concentrate on deposit mobilization. Unfortunately, in some countries that have postal savings banks, those institutions are attempting to become lending institutions serving the needs of small borrowers. This is more than likely to lead to nonperforming loans, insolvency, and a call upon the central bank for a bailout. The alternative of concentrating on deposit mobilization and lending the proceeds to the central bank or to sound commercial banks at a viable interbank rate should be considered.

Promoting Capital Markets

The experience in most Asian developing countries has been that capital markets began to play a significant role only after the banking system was reasonably well developed. Capital markets are not a substitute for an inefficient banking system, as Popiel has suggested,4 but, in fact, require the services of an efficient banking system to be able to function efficiently themselves. Therefore, in the early stages of financial development the focus should be on the banking system and creating an environment in which it can grow and provide the many services for which it is intended and generally best suited.

As a rough rule of thumb, based on Asian experience, it is probably unrealistic to expect capital markets to play a significant role in mobilizing and allocating savings in a country until the ratio of M2 to GDP has reached at least 40 percent, and the ratio of M2 to Ml is in the range of 3 or 4 to 1. Such ratios tend to indicate that the banking system has managed to become a significant mobilizer of financial savings, which also suggests that inflation is not a major problem and that the public has some confidence in the safety and soundness of the banking institutions, as well as the monetary and fiscal policies of the government. It also is likely to indicate that money markets, especially the interbank markets, are already functioning.

Capital markets have two important “hidden” costs that are not widely appreciated. The first is that stock market indexes, no matter how small or inefficient the market, often become “barometers” of overall economic performance that can affect capital flows, international borrowing costs, and even the tenure of senior government officials. As a consequence, there is a strong temptation for governments to promote rises in the index, and to resort to various nonmarket measures to prevent excessive declines.

The second hidden cost is the time of both senior government officials and financial specialists in government agencies and private institutions that are expended on setting up capital market institutions, attempting to provide adequate regulation, and dealing with the problems and crises that inevitably arise. These costs are in addition to the more obvious real resources that go into building the stock exchanges, trading and clearing systems, and communication facilities that efficient capital markets require, as well as the often highly trained human resources that are directly involved in operating the capital market institutions. All these costs need to be weighed against the positive benefits that may be derivable from the capital markets in assessing when it is justifiable for individual countries to begin promoting their development.5


Current efforts to transform quickly the financial systems of developing countries from directly controlled, often insolvent institutions to healthy, market-oriented, and indirectly regulated institutions are encountering serious problems. Efficient financial markets, well-managed financial institutions, reliable information and communication systems, competent monetary policy managers, and prudential regulators cannot be created in a brief period, especially in countries with unstable political systems, and limited human capital and physical infrastructure.

The alternative is to follow an evolutionary strategy that seeks to address the most serious problems sequentially with measures and instruments that not only are practical within the existing constraints but also will help to open the way for further refinements in the future. This paper has suggested approaches to managing monetary policy (including determining interest rates and exchange rates, setting reserve requirements, and managing reserve money), strengthening prudential regulation of banks, improving rural financial services (especially savings facilities), and developing money and capital markets that are consistent with this kind of evolutionary strategy for financial restructuring in countries in the early stages of financial development.


See, for example, William E. Alexander, Tomás J.T. Baliño, and Charles Enoch, The Adoption of Indirect Instruments of Monetary Policy, IMF Occasional Paper No. 126 (Washington: International Monetary Fund, 1995).


See, for example, Gerard Caprio, Jr., “Bank Regulation: The Case of the Missing Model,” in Sequencing? Financial Strategies for Developing Countries, ed. by Alison Harwood and Bruce L.R. Smith (Washington: Brookings Institution, 1997), pp. 109-26.


See the “Proposal to Issue a Supplement to the Basle Capital Accord to Cover Market Risks,” issued by the Basle Committee on Banking Supervision, Basle, April 1995.


Paul A. Popiel, Financial Systems in Sub-Saharan Africa: A Comparative Study, World Bank Discussion Paper, Africa Technical Department Series, No. 260 (Washington: World Bank, 1994).


Professor Kapur, in commenting on this paper, suggested that Yoon Je Cho, in his “Inefficiencies from Financial Liberalization in the Absence of Well-Functioning Equity Markets,” Journal of Money, Credit and Banking, Vol. 18 (May 1986), had developed a strong theoretical argument for the early introduction of equity markets in order to achieve more efficient allocation of capital and risk sharing in a liberalized financial environment. As we read Cho’s argument, he is really making a case for venture capital companies to mobilize funds from limited numbers of high-risk, high-return investors and channel such funds into matching types of enterprises. We would agree that this could be a useful form of intermediation in some countries, and that equity markets may eventually play a role in distributing shares of successful venture capital investments. But they are not necessary to the process and, if introduced prematurely, their costs may more than offset any benefits deriving from this role.

    Other Resources Citing This Publication