6 Introducing Open Market Operations, Reforms in Markets, and Policy Instruments
- Tomás Baliño, and Lorena Zamalloa
- Published Date:
- September 1997
Open market operations have become the major instrument of monetary control in industrial countries. In those countries, the evolution of well-developed, active interbank, money, and securities markets has enabled central banks to undertake outright or repurchase agreements in securities or foreign exchange as needed to provide or absorb bank reserves.
The advantages of open market operations are well known. The instrument provides great flexibility in the timing and volume of monetary policy operations at the initiative of the central bank; permits an impersonal, market-oriented business relationship with counterparties; and avoids the economic and market inefficiencies of direct controls.
In today’s globalized financial world, the flexibility of open market operations is also becoming increasingly important to countries with developing or transitional economies as their markets evolve, becoming more deregulated, more competitive, and more closely integrated into world markets. At earlier stages of market development—when highly flexible two-way open market transactions by the central bank in well-developed money and securities markets are not practical—certain o pen-market-type instruments can be, and have been, employed to secure many of the advantages of more fully evolved market operations.
Such instruments would generally be employed in primary markets and include auctions of a central bank’s own securities, auctions of treasury securities for monetary policy purposes, and credit auctions through the open market function. They have some of the features of broader open market operations in that the monetary authority gains more initiative in its operations, interest rates are or can be market determined, and there is a more impersonal business relationship with counterparties.
However, as a country continues to grow and its markets become deeper and more sophisticated, the central bank should become increasingly able to employ open market operations even more flexibly in active secondary markets for government securities and money market instruments. Because of the added flexibility gained for monetary policy from the presence of active, competitive markets in such instruments, it would also be desirable for the central bank as well as other arms of the government to take regulatory and other steps that may be needed to encourage market development. Such steps would not only help create the conditions for a stabilizing and market-enhancing monetary policy, but would also help encourage, through demonstration effects, the development of competitive markets in and across other financial and also nonfinancial sectors.
Countries that lag in developing primary and secondary markets must rely for the most part on other, more direct instruments of monetary control, such as reserve requirement changes, terms and conditions for access to the central bank’s discount window, or changes in quantitative credit restrictions on commercial banks and other institutions. These may be quite effective for a time, especially for countries at earlier stages of development and without exposure to a broad and sophisticated spectrum of markets, investors, lenders, and borrowers.
However, as a country’s economy and markets expand, policy implementation that is basically limited to the more direct instruments of control tends to become less effective because markets eventually find a way around them, especially in a globalized world. And even if such an approach turns out to be highly effective, at least temporarily, it may well be associated with stop-go policies or with uncertainties generated by the government’s role in credit allocation that are damaging to long-run business and financial planning.
Use of market instruments in policy implementation would enable the central bank more flexibly and smoothly, and with less distorting side effects, to phase in adjustments to the bank reserve and monetary base that may be needed as the country’s financial structure changes, usually unpredictably, in the process of economic growth and transition. More generally, market instruments, because their use is entirely at the central bank’s initiative, provide a relatively reliable tool for controlling the longer-run growth in the country’s money or credit aggregates and inflation if the bank has the will to do so. On a more technical level, a market instrument also gives the central bank added flexibility for adjusting the size of its reserve operations in light of the ongoing market response to its policy intervention—perhaps a particularly important advantage in emerging markets, which are less likely to have the experiential basis for gauging the response in advance.
This paper focuses on the practical role in monetary policy implementation that can be, and has been, taken by various forms of open market and open-market-type operations in developing and transitional economies; on their advantages and disadvantages; and on the market conditions that may influence, and abet, the transition from operations in primary markets to more flexible two-way operations in secondary markets.
For empirical analysis, the paper draws in part on a survey of open market or open-market-type instruments used in 15 developing or transition economies and more detailed studies of market instrument development in 5 countries (Brazil, Indonesia, Mexico, Philippines, and Poland).1 The survey and individual studies help demonstrate how open-market-type and open market operations have evolved in nations with differing financial systems and traditions (Axilrod, 1995).
Open Market Operations in Relation to Other Monetary Instruments
If open market operations are to be the principal instrument of policy implementation, other monetary instruments obviously need to be adjusted supportively so as to ensure that day-to-day operating objectives are attainable and that the chances of reaching intermediate-term monetary guides, such as the money supply, are maximized. The need for adjustment applies particularly to the discount window, or any other central bank lending facility where the banking system can obtain reserves at its own initiative. There are also reserve requirement issues. The structure of requirements affects not only the predictability of the multiplier relationship between bank reserves and the money supply, but also banks’ speed of response to open market operations.
The complementary adjustments in other monetary instruments that are most effective depend in part on the strategy adopted for conducting day-to-day open market operations. In general, they can be conducted in one of two ways. First, they can be actively aimed at a given quantity of reserves, letting the price of reserves (a money market or interbank interest rate) fluctuate freely along with fluctuations in the pressure on banks’ liquidity. Second, they can be aimed at a particular market interest rate, letting the amount of reserves provided at the central bank’s own initiative be determined passively as a function of demand at that price.
Given the inherent volatility of money demand and a desire to avoid policies that risk destabilizing markets, central banks often follow a passive approach to open market operations, providing whatever reserves are needed to accommodate the demand for bank deposits and required reserves that actually emerges during a particular operating period (anywhere from a week to a month depending on the span of time over which banks must meet their reserve or clearing balance requirements). If the monetary authorities were not following such an approach, the banking system, in a period when money demands were perhaps temporarily strong, would have to draw down excess reserves, increase borrowings at the central bank’s discount window, or undertake portfolio adjustments, with consequent upward pressure on money market interest rates as banks’ liquidity positions came under pressure.
While accommodative in the short run, use of a passive approach to open market operations is nonetheless consistent with a monetary policy that stresses either interest rates or monetary and credit aggregates as intermediate objectives. With the money supply as an objective, the central bank would ordinarily focus on control over a three- to six-month, or possibly longer, period and would accommodate to the week-to-week or month-to-month variations that are generally economically insignificant. However, control over the longer-term policy period would depend on the central bank’s ability to estimate the rather uncertain and indirect relationship between a money market rate or bank liquidity conditions and the responses of the banks and the public that influence the money supply.
It is unclear whether an active approach to open market policy that takes a particular reserve level—derived from, say, some longer-run path of desired money growth—as a direct target in an operating period would achieve more adequate control of money supply measures. It would depend in part on the predictability of the multiplier relationship between reserve aggregates and the money supply. That relationship too may be variable and unstable, especially when financial markets are expanding and modernizing. Moreover, a rapidly changing financial structure would also tend to alter the economic significance of particular measures of money supply themselves.
Industrial countries have normally followed a more passive approach to providing reserves through open market operations. A notable exception to such an approach was the policy followed by the U.S. Federal Reserve System from late 1979 to late 1982. At that time, the Federal Reserve chose a particular volume of nonborrowed reserves (the reserves provided through open market operations) as its short-term operating objective, with the amount determined by the presumed multiplier relationship to the policy’s money supply objective and an initial assumption about the amount of borrowed reserves.
The procedure was adopted in inflationary circumstances on the thought, first, that it would result in more assured money supply control. Second, after a while, as the policy approach was successfully sustained, it was expected to restore the Federal Reserve’s lost anti-inflation credibility. The highly developed U.S. markets were considered to be resilient enough to absorb the more volatile interest rate movements that were likely to, and did, occur.
The policy was abandoned after inflation was reduced and as it became apparent that changes in financial technology and institutional structure were making the relationship between various measures of money and overall economic and financial conditions increasingly uncertain. Reserves provided through open market operations were then once again determined passively, guided in effect by money market interest rate pressures.
In countries with markets at earlier stages of development, the absence of adequate secondary or interbank markets to provide signals about reserve needs and to convey the results of policy would be one reason for taking an active approach to reserve provision. Another reason would be to clarify and better define the central bank’s policy objectives, especially when control of an existing inflationary episode is the overriding goal.
Such an active approach is in effect embodied in a number of IMF-supported programs for particular countries where control of inflation may be the predominant concern or where competitive markets may be at an early stage of development. For particular policy reasons, though—for example, the importance to the country of exchange rate and balance of payments stability—such programs may use the net domestic assets of the central bank as a guideline rather than a bank reserve or monetary base aggregate.2
In the end, whether a central bank uses a more passive or more active approach to open market operations will depend on the structure of its markets and economy, on the stability and predictability of the relationship between money supply measures and national economic objectives, and on the economic or financial issues that are of most concern to policy at the time.
Discount Window Policy
For open market operations to be an effective instrument of monetary control from either an active or a passive perspective, certain limitations need to be placed on access by banks to the discount window and other central bank lending facilities. Obviously, if banks could borrow at will from the central bank, open market operations could not be used effectively in an expansionary situation to control the money market rate, or to influence total reserves or the monetary base for that matter. Any positive gap between banks’ demand for reserves and the reserves supplied through market operations could be readily filled through borrowing at the discount window. As a result, the money market rate would never rise above the discount rate, and any growing demand for reserves would be fully accommodated. In an expansionary situation, open market operations simply could not function as an effective policy lever to exert a restraining influence on money and credit conditions, and the discount rate rather than the money market rate would serve as the anchor rate for the economy.3
Thus, to make open market operations an effective policy instrument, the discount window should be designed to deter access to the central bank’s credit, for example, through a high penalty rate or restrictive guidelines. A penalty rate from one perspective can be construed as an upper limit on money market rates. Whether or not it is a limit in practice would depend in part on how it is set (e.g., administratively reset on a periodic basis, automatically set at some premium to a going market rate, and the time lag, if any, in setting the rate) and on whether access to the window at that rate is restricted or unrestricted. Some countries, such as Germany, maintain a dual rate structure comprising a basic discount rate and a penalty Lombard rate and employ open market operations to guide money market rates within the channel.
If a penalty rate is set well beyond current market rate conditions, there is a risk of sharp market reactions to liquidity demands that are unanticipated in open market operations, a situation most likely to arise when such operations are aimed actively at a predetermined reserve aggregate. In such circumstances, as well as more generally, use of guidelines that restrict access to the window may permit a smoother adjustment to developing reserve shortages (relative to demand) in the banking system, at least compared with a window that is effectively closed by a very high penalty rate. With the window administered to restrict access, short-term interest rates will rise in a tight money period when reserves provided through open market operations are persistently in short supply. However, the rise will be buffeted to a degree by the ability of banks to make more orderly portfolio adjustments by borrowing from the central bank for a limited period (and only after searching actively for funds in the interbank market).
Short-term, temporary borrowing that helps the banking system and markets adjust more smoothly to changes in money market conditions initiated by open market operations should be differentiated from longer-term structural borrowing at the discount window, such as emergency advances to institutions in severe operating difficulties. Longer-term borrowing at the discount window may also be an important means of adding reserves to the banking system in the normal course for countries with markets at an early stage of development, such as China and Russia.
Reserves provided through structural, longer-term borrowing—whether for emergency or other reasons—should generally be considered to be practically the same for monetary policy purposes as reserves provided through open market operations because they represent long-term funds that determine money market or bank liquidity pressures, rather than the short-term, temporary funds that are determined by such pressures. Their expansionary impact on the reserve base would need to be constrained to be consistent with the central bank’s money-supply or interest rate objectives. If a market instrument of some sort were available to the central bank, it could be used to absorb any excess creation of what are in effect “permanent” reserves through the discount window.
In any event, at relatively early stages in their market development, a number of developing and transition economies have limited access to the discount window through penalty rates, guidelines, or moral suasion in the interest of making open market operations more effective. These countries have also been influenced by the need to encourage the evolution of a competitive interbank market to help provide a better framework for the further development of open market policy instruments.
Of the countries examined, some allow both short- and longer-term borrowing through their credit facility (Brazil, Ghana, Indonesia, Malaysia, and Poland), while others provide only short-term credit (Czech Republic, India, and Thailand). Countries employing open market operations as a principal monetary instrument have generally taken steps to ensure that short-term adjustment-type credit available through the discount window is used judiciously, if at all.
In the Czech Republic, for instance, the discount window comprises a short-term credit facility at a penalty rate and a rediscount facility with restricted access. In practice, neither facility is used actively, with the central bank’s discount rate serving as an indicator of the stance of policy implemented through both open market and open-market-type operations. In India, discount window credit is provided through a standing facility to two specialized institutions, with frequent recourse discouraged by the practice of scaling up loan rates. In Mexico, the central bank does not have a short-term adjustment facility, although it maintains an emergency lending window and, more recently, has instituted daily credit auctions.
If the discount window is effectively shut for short-term adjustment purposes, its market-smoothing or safety-valve functions are lost, and the full burden of adjustment to shortages of reserves from open market operations falls on excess reserves, overdrafts, or even stronger (than otherwise) portfolio adjustments by banks to extinguish required reserves. In that context, the availability of short-term adjustment borrowing from the discount window can helpfully smooth the market response to open market operations by allowing enough access to provide some cushion against unexpected day-to-day financial shortages, but not so much as to offset the basic thrust of monetary policy or, at a more micro level, to hamper the necessary development of interbank markets.
Reserve requirements can be considered either as an alternative instrument of monetary control implemented through changes in required reserve ratios or as a relatively stable fulcrum that may enhance the effectiveness of open market operations by making it easier to predict the money supply or interest rate effects of open market operations. Since open market or closely related operations have become more widespread, changes in reserve ratios have been used less and less as an instrument of monetary control.4
In any event, when used for monetary control, reserve requirement ratios have the disadvantage of being a relatively crude tool that also tends to complicate internal bank reserve management procedures. Moreover, in modern open and highly competitive markets, reserve requirements may place banks at a significant competitive disadvantage compared with institutions that provide the same or similar services, but that are not subject to the reserve requirement tax. In that context, many countries have tended either to lower reserve requirements to minimal levels or to eliminate them.5
The minimal level of reserve ratios that would be useful for implementing policy through open market operations would depend partly on how closely a central bank wishes to control a particular measure of the money supply. Relatively close control is facilitated if the deposit components of the money supply are subject to required reserve ratios that are at least high enough to be binding in that they force banks to hold more required reserves than they would otherwise voluntarily hold for transactions or clearing purposes. Such voluntary reserves are likely to be subject to more slippage and volatility in relation to deposit levels as banks respond to changing liquidity pressures and other circumstances.
The reserve ratio might also be best expressed as an average pertaining to a relatively short period—say, two weeks. The averaging process would permit banks to manage their reserve positions more flexibly and with less need to establish a large excess reserve buffer that might make the relationship between reserves provided by the central bank and the money supply less predictable. A relatively short period would help ensure that responses to the goals of monetary policy were reasonably prompt. Almost all of the countries sampled average reserves, generally over a period ranging from two weeks to a month.
The argument for reserve ratios that are binding in the above sense is considerably less strong if the central bank’s open market operations are not actively targeted on a particular level of reserves derived from the multiplier relationship to some measure of the money supply, but are determined passively in response to market interest rate pressures. Even in the case of passive reserve provision, however, binding reserve requirements may have some practical use.
For instance, the need for a relatively predictable reserve base to make implementation of open market policy more effective is greater when the money market is less developed and is unable to provide relatively unambiguous signals about the degree of a shortage or surplus of funds in the market. Even in the United States, with its highly developed money market, reserve requirements, while being steadily lowered and reduced to zero on time and savings deposits, apparently remain binding on transactions deposits. The ability to make relatively predictable estimates of required reserves seems to help the Federal Reserve make decisions about the size and timing of open market operations during a short-run operating period. In developing countries, although reserve requirements have for the most part been lowered over time, they also appear to remain at binding levels (with the exceptions of Indonesia, Mexico, and Tunisia), at least as judged from the 15 countries surveyed.
Because of statistical difficulties in securing deposit data that are current enough to estimate required reserves, some countries base, or have based, reserve requirements wholly or in part on deposits in a prior period. By basing reserve requirements on earlier deposits—a so-called lagged reserve requirement system—the amount of required reserves and the fulcrum for open market operations during a current open market operating period would be known with reasonable certainty. Deposit revisions and banks’ current attitudes toward excess reserves, however, would always be problematic.
Such a system has certain statistical and operating benefits, although it has disadvantages if the central bank is attempting to control the money supply by aiming directly at a reserve aggregate, because it tends to delay the market response needed to bring under control undesired changes in money and credit. Under a lagged reserve requirement system, any undesired expansion in deposits, for example, would not immediately increase required reserves and thus would not automatically add to upward interest rate pressures in the money market. However, the disadvantages appear significant mainly if the lag between deposits and required reserves is fairly lengthy.
A substantial number of countries examined—including the Czech Republic, Egypt, India, Malaysia, Thailand, and Tunisia—employ a lagged or partially lagged reserve requirement structure. For the most part, however, the length of the lag is reasonably short, about 14 days—a period that probably would not give rise to excessive technical difficulties in controlling the money supply over the medium or longer term.
In sum, when open market or open-market-type operations are crucial in policy implementation, the most effective complementary reserve requirement structure depends in part on whether the central bank adopts an active or a passive attitude toward reserves provided through market operations. An active attitude, to the extent that it presumes high priority for money supply control, would argue for a binding reserve requirement structure. It would also argue for uniform reserve ratios applicable to the various deposit components of the money supply measure to be controlled—a condition met in relatively few of the countries studied—and for applying the ratios to deposits contemporaneously or at least with a relatively short lag. Efficiency and equity in markets and the economy would be encouraged if the reserve ratios were set as low as possible while continuing to be binding.
As markets evolve, with attitudes toward money changing as the financial structure becomes more sophisticated, it becomes more practical and perhaps even necessary for central banks to adopt a more passive stance toward open market operations. An overnight money market rate or the banking system’s marginal liquidity position would then be taken as the day-to-day operating guide for open market operations. Under such circumstances, there are still some advantages to a binding reserve requirement structure as a benchmark for judging the timing and size of open market operations. This is especially important if the money or interbank market is not yet developed and competitive enough, with adequate breadth and depth, to provide reliable signals about emerging tightness or ease.
More broadly, however, consideration also needs to be given to retaining the ability to vary reserve requirements as an optional policy instrument when bank liquidity needs to be expanded or contracted rapidly and clearly signaled. The availability of a reserve requirement instrument may be especially useful in countries where markets still remain relatively thin and may not be able to absorb the substantial open market operations that would sometimes be required to offset large and sudden changes in liquidity conditions.
Market Prerequisites and the Role of the Central Bank
In the course of a country’s economic growth, its financial markets will evolve generally in the direction of greater breadth and complexity, although the particular institutional form—whether, for example, banks are dominant or share a role in varying degrees with independent securities firms, savings institutions, or others—depends on historical circumstances and government attitudes. As the financial system expands, the market prerequisites for using open market instruments in monetary policy implementation should gradually fall into place. Experience has shown, however, that the pace and pattern of market development may well need guidance from a country’s monetary and other government authorities to facilitate a timely shift from direct to indirect instruments of control if policy is to be most effective as markets become deregulated, unsegmented, and, in today’s world, open externally.
The transformation to indirect instruments of control usually involves two stages: (1) a shift away from direct instruments toward more reliance on open-market-type operations in the primary market, and (2) a further shift toward greater use of more fully flexible two-way operations as active secondary markets develop in more mature financial systems. In that context, in addition to their monetary policy function, open-market-type operations should be viewed in their long-run structural role as a useful, if not necessary, prelude to the evolution of active secondary markets.
While such operations should be designed to encourage growth in markets and discourage dependence on the central bank for either meeting liquidity needs or laying off surplus funds, other actions—by the central bank itself (e.g., affecting the discount window), by other government bodies, or within the private sphere—may be more crucial for appropriate market development. This section will focus mainly on how developing and transition economies can move toward the market structures that are most useful for highly flexible open market operations and on the role that the central bank can play in contributing to the development of such markets.
Suitability of Markets for Policy Implementation
Highly flexible open market operations require, ideally, a market structure that features a secondary market for an instrument in which a large number of daily transactions occur on a continuous, competitive basis and involve numerous different buyers and sellers or borrowers and lenders. This ideal situation exists in few developing or transition economies and was late to develop even in more advanced economies. However, as countries around the world have begun to appreciate the advantages of a deregulated market environment, competitive price determination, and a discretionary monetary policy aimed at reasonable price stability, the practical usefulness of active secondary markets has become clearer, and their establishment has taken on more urgency.
Open market operations of one sort or another can and should be undertaken, of course, in markets that may not be ideal, but that are moving in that direction. In markets lacking depth and continuity, such operations may need to be limited in size or employed only on a periodic basis. The participation of the central bank should hasten the development of the market not only by enlarging its size but also because the bank’s stature and authority can be used to guide the development of the infrastructure needed to continue the market’s expansion.
In addition to affecting the usability of open market operations, the pace and pattern of market development also affect the liquidity and riskiness of the central bank’s balance sheet and thus, potentially, its credibility and stature. Without credibility, central banks find it much more difficult to maintain overall financial stability, as experience has amply demonstrated in both developed and emerging markets.
A central bank’s credibility is reinforced if markets perceive that its portfolio of assets is highly liquid and essentially risk free. Thus, the most appropriate market structure for open market operations would be one that permits the central bank to acquire such securities over time. If the central bank, in the course of supporting economic growth, has instead acquired securities that the market believes are illiquid and cannot be sold as needed to avert an inflationary expansion of the nation’s monetary base, then confidence in the bank is likely to falter.
Markets by Maturity
The markets most suitable for sizable and highly flexible open market operations are normally those where the shortest-term instruments are traded, although it should be technically possible and sometimes desirable on policy grounds at times to employ all maturity sectors in operations. The short-term sector, however, is most capable of absorbing frequent in-and-out official transactions.
When well developed, the short-term sector is characterized by a large and continuous volume of activity resulting from use of the market by the government, financial institutions, and other businesses for day-to-day management of liquidity and cash flow. The central bank is thus able to undertake open market transactions on its own initiative and in adequate size either to shift policy or—what is generally the more frequent type of operation—to offset day-to-day changes in market liquidity that might otherwise signal policy shifts when policy has in fact not changed.
Longer-term markets have less capacity than short-term markets to absorb a large volume of open market operations, partly because they usually have less breadth and depth and partly because they are basically riskier. Although operations in such markets may be useful when the central bank is providing reserves or is attempting more directly to influence conditions in the long-term market sector, for the most part, they tend to be avoided when the central bank wishes to drain reserves from the banking system.
Open market operations undertaken to absorb reserves in longer-term markets risk a marked upward interest rate response, at least in the short run, given the relative thinness of such markets and the inherently larger price risk in longer-term than in shorter-term securities. Even when operations to constrain reserve growth are undertaken predominantly in short-term markets, there would still be some effect on longer-term rates, but it would be more measured. The response of longer-term rates would be unaffected by overreactions to a sudden change in existing demand-supply conditions within the longer-term market itself. As a result, these rates would be a much more useful indicator to the authorities of the financial community’s attitude toward the overall thrust of monetary policy.
The longer-term market would present a clearer picture of the underlying cost of capital in nominal terms, thereby providing some basis for estimating the longer-run inflationary premium that is one useful measure of the credibility of monetary policy. It would also provide businesses with an effective basis for gauging the present value of future income streams, thereby permitting them to make economically more rational judgments about capital spending plans, given their own estimates of the real return on investment and the likely course of inflation.
Markets by Sector
Looking at markets by type, rather than by maturity, of instruments, three sectors present the best opportunities for effective open market operations—the markets for national government and central bank securities, for interbank debt, and for very short-term debt instruments issued to the public by financial institutions and other corporate entities (encompassing such instruments as commercial paper, finance company paper, and bank certificates of deposits). Of the three, the government securities market presents the fewest complications and is generally preferred for operations, with availability dependent on the role of the government in the economy and the amount of debt outstanding resulting from the cumulative net impact of budgetary deficits and surpluses.
The government market, as well as a market for a central bank’s own instruments, is generally thought to be free of the kind of credit risk found in markets for instruments issued by institutions that depend on cash flow from the sale of their services. The government’s taxing power is considered to be insurance against such risk.
In political and economic situations that are highly unstable, however, government debt can be subject to a credit risk in market eyes—a risk of default—that is similar to other instruments. The possibility of such a risk would make it almost impossible to maintain a viable primary market for issuing debt, much less the two-way secondary market that is essential for fully flexible open market operations. Thus, political stability and a sustained government record of meeting interest payments and redemption schedules are essential for the use of the open market instrument.
The market for government securities can be effectively destroyed not only by a government’s failure to meet its contractual obligations. It can also dry up if the central bank pursues an inflationary policy that drives investors, domestic and foreign, out of the market by eroding the real value of the outstanding debt. Thus, a government policy designed to keep inflation within acceptable bounds is a precondition, along with political stability, for developing and sustaining markets suitable for the two-way open market operations of the central bank.6
The development of short-term markets in government and private instruments in countries such as Mexico, the Philippines, and Poland—among those studied—facilitated the use of open-market-type and later, in certain cases, full-fledged open market operations. In a number of countries, market development was itself spurred by the central bank’s introduction of open-market-type operations. The transition to flexible open market operations has, however, proved difficult in some countries, especially those without a sizable volume of government debt outstanding to provide the basis for a risk-free, liquid market.
In Indonesia, where the government is not permitted to run domestically financed budgetary deficits, the central bank took steps to institute an indirect system of monetary management by issuing its own bills once the more direct monetary controls proved to be unworkable. The central bank also promoted a short-term commercial paper market with a view to developing a two-way instrument that would provide reserves over time. However, secondary market activity in both commercial paper and the central bank’s own bills has so far remained thin, and transactions are dominated by central bank activity.
In general, short-term private7 debt markets are less suitable for flexible open market operations than the market for government debt, in large part because the inherent credit risk in those sectors tends to make for less resilient and thinner markets. But in addition, the central bank can unnecessarily be placed in an awkward position because of issues raised by the presence of credit risk. For instance, if the central bank happens to be on the buyer’s side of the market, counterparties may very well take the opportunity to unload riskier paper. The central bank would find it difficult to avoid purchasing at least some of the paper; if it refused, the market itself would turn away from the paper on the thought that the central bank had access to information unavailable to the market generally (as in practice it may well have).
Such a quandary might be resolved if the central bank were to confine operations to paper carrying ratings above a certain level from an independent rating agency. In Indonesia, the central bank buys commercial paper issued by banks (so-called SBPUs) that are endorsed by third parties, presumably to reduce credit risk.
If the central bank finds that its open market operations must be carried out in the private money market because either there is no government debt or the amount of the debt is declining, operations in commercial bank instruments or in the interbank market may raise fewer credit risk issues than transactions in other money market instruments, such as commercial paper. In contrast to other financial institutions, commercial and often other more specialized banks have what can be considered a close “business” relationship with a central bank because they hold required reserve, clearing, or conventional working balances at the institution and because they have access to the central bank’s discount window, or, if there is no such window, to the central bank’s overdraft facility. Moreover, the central bank can easily refrain from operations with any banking institution that borrows from the discount window—whether for emergency or other longer-term purposes—thus establishing an objective and understandable basis for distinguishing among counterparties.
The central bank of Malaysia, for one, undertakes the bulk of its operations in the interbank market, mainly through short-term borrowing to absorb liquidity created by capital inflows. In Indonesia, however, the central bank has chosen not to intervene in the active interbank market, but to undertake primary market operations instead in SBPUs to provide reserves and in its own certificates (SBIs) to absorb reserves, with the hope of fostering a secondary market.
Central Bank’s Balance Sheet
While a central bank in a country without a government securities market should be able to develop other channels for open market operations, there is the risk that, over time, acquiring a securities portfolio dominated by private commercial paper and banking assets would make the central bank appear less liquid and indeed a bit less safe than if government debt were the primary source of the nation’s monetary base. Private paper may prove difficult to sell at times, especially in crises when a central bank may wish to undertake open market sales to at least partly offset the potentially inflationary reserve impact of increased emergency lending through the discount window.
In such circumstances, the central bank may not be viewed as any stronger than the private banking and financial system itself. Under those conditions, the central bank would also have difficulty selling its own newly created securities in an open-market-type operation to absorb reserves. It might not be viewed, therefore, as a truly safe guardian against either systemic risk or inflation, thereby increasing the odds on such adverse developments as capital flight at the first sign, or rumor, of financial difficulties in key institutions. Financial crises could become at least marginally more difficult to control.
The advantages for operations and market perceptions in the central bank’s having a relatively liquid and risk-free balance sheet would be an important reason for confining operations mainly to government securities when possible.8 If a significant government debt market does not exist, consideration might also be given to creating a similar balance sheet effect by developing a special government debt instrument that can be used to add reserves to the banking system over time. In contrast to open-market-type instruments that have been used to absorb reserves by being sold into the market, it would remain as a permanent and liquid addition to the central bank’s balance sheet, substituting to a degree for private assets and in effect “dressing up” the central bank’s books, as will be discussed further in the section Conduct of Open Market Operations.
Regulatory Role of the Central Bank in Market Development
Open market operations, or monetary policy as a whole for that matter, will work best in a market environment that not only permits competitive determination of interest rates in short-term and government securities markets, but also allows variations in those rates—the basic liquidity and/or risk-free rates in the financial system—to be promptly reflected in other markets. Establishment through law and regulation of continuously functioning competitive markets is thus basic to an effective monetary policy with a more predictable impact.
Normally, the central bank would not have a significant direct regulatory input into or authority over those markets in which it has no operating interest—such as those for equities or private longer-term debt instruments. However, it needs to be clearly understood that open market operations, and monetary policy generally, are most effectively implemented when transactions in the country’s various financial markets take place in a regulatory framework that minimizes the chances of micromarket developments—such as a breakdown in the clearing and payments mechanism or major institutional failures—that would prevent monetary policy from focusing on its macroeconomic goals.
In the early stages of a country’s development, the central bank, given its expertise in and concern for markets, would generally be called upon to provide expert advice as the government formulates financial legislation and regulations. However, questions about the extent of the central bank’s continuing and direct responsibility for regulatory matters would normally be confined to financial areas of immediate concern in its operations—such as the banking system or the government securities market. Even for those markets, however, it is not clear whether the central bank should or should not be a direct regulator.9
With regard to the government securities market, the central bank and the government need a reliable marketplace in which participants can feel secure that counterparties will perform and that is transparent and “safe” enough to encourage broad customer participation. The central bank, like any other institution, should establish performance standards for its counterparties based on its business needs. Its chief need is for policy to attain its bank reserve and/or short-term interest rate objectives through open market operations. The central bank is also the natural focal point for market surveillance through gathering information from its counterparties and publishing aggregate market statistics.
However, the central bank may not wish to go beyond these functions and also assume direct regulatory responsibility and oversight for the market, which may unduly tax its limited personnel resources. Moreover, these additional functions may unnecessarily expose the bank to a loss of stature and adversely affect its credibility in monetary policy should scandals erupt in the government securities market, as history suggests they inevitably will.
In that sense, it may serve a country’s national interest best to maintain a division of labor between monetary policy operations in markets and regulatory authority over the markets, with an institution other than the central bank having regulatory power over the government securities market. Nonetheless, whatever the exact regulatory role of a central bank, the public will generally look to it as bearing some oversight responsibility for the markets in which it operates and will assume that the continuing presence of the central bank in the market validates the market and perhaps also the various counterparties to the bank’s transactions. From that perspective, if not from its own need for markets that make policy operations more effective, the central bank should take steps that help rationalize the market’s architecture and enhance its performance.
Market Architecture and Performance
To implement open market operations, a central bank would generally prefer a market that is designed for continuous rather than periodic transactions and for immediacy in the execution of trades. In addition, the monetary authority would also prefer a market that is transparent and where communication of its operations is prompt so that its activities and purposes are well understood.10 Continuity and transparency would also encourage the participation of businesses and individuals that is needed if markets are to become adequately broad and deep.
The central bank can consider a number of steps to encourage the development of an active and broad domestic market for open market operations, although the size of the market and the potential for fully flexible operations, relative to more limited open-market-type operations, would necessarily be affected by a country’s stage of development. These steps include promoting an interbank market; helping in the design of market instruments and a trading infrastructure; providing financing facilities; establishing criteria for doing business with the central bank’s open market function; collecting and disseminating market statistics; and encouraging a safe clearing and payments mechanism.
Apart from its potential as an instrument for open market operations when government securities are not adequately available, an active interbank market is more particularly important to monetary policy because it serves as a pressure gauge for helping to judge the timing and volume of security market operations. As it is extended to encompass both bank and nonbank dealers in securities, it may also be instrumental in establishing a viable securities market by making day-to-day financing available on a competitive basis.
A large number of the countries surveyed—those both with and without effective government securities markets—have developed or are working toward a competitive interbank market as they move toward an increasingly market-oriented monetary policy with a principal role for open market operations. Among these are the Czech Republic, India, Indonesia, Malaysia, Mexico, the Philippines, and Russia.
Most countries have helped encourage the interbank market’s development through adjustments in policy instruments as well as through moral suasion and other approaches. While progress in establishing the market has been variable depending on institutional circumstances, the more successful countries so far—India, Indonesia, Malaysia, and Mexico—have incorporated discount window policies that discourage, penalize, or forbid short-term borrowing at the central bank.
On a more technical level, the central bank can encourage market development by using its transfer and settlement mechanism to ensure the integrity of interbank fund flows on its books. Since the central bank would in the process be guaranteeing that the balances transferred represented good funds on receipt (i.e., that they could be used immediately), it would have to consider the extent to which it wishes to be exposed to so-called daylight overdrafts in the process or wants the commercial banks involved to hold satisfactory collateral with the central bank.
Market Instruments and Infrastructure
Experience suggests that using competitive price auctions to market debt obligations may be one of the more crucial mechanisms under the control of the central bank and the treasury for encouraging the development of a well-functioning, competitive securities market. In the Philippines and Mexico, initial efforts to establish a government securities market foundered because early offerings took the form of tap issues (issues available to the market on demand from the central bank’s portfolio at a price set by the bank) at below-market rates. The primary market later developed, followed by a surge in secondary market activity (relatively strong in Mexico, not so strong in the Philippines) when the government shifted to competitive auctions.
It should go without saying that the maturity and other characteristics of the debt instrument (e.g., minimum denomination and call provisions) not only need to be adjusted to a country’s particular institutional structure, but should also be set so as to attract an increasing number of participants into the market. In a country with substantial inflation, experience has shown (as in Brazil) that the market is unlikely to develop unless bonds are indexed in one way or another to retain their real value.
With regard to market infrastructure, the central bank should take the lead, along with the treasury, in encouraging market practices and organization most conducive to the evolution of competitive trading. When markets are at an early stage of development and are relatively inactive, this role could take the form, for example, of having the market establish a daily, computerized system of bids and offers that protects anonymity, with the market probably closing at a relatively early hour. As activity increases, trading is more and more likely to occur in the course of the day (rather than mainly just at or near the close), and the market could stay open longer. If the market is to be transparent and informative, trading should be discouraged from taking place outside “established” markets—whether in the form of exchanges, over-the-counter markets, or both—that can be monitored.
The treasury should have at least as great an interest as the central bank in developing both a primary and a secondary market. For one reason, the cost of the national debt will be reduced to the degree that government securities become more and more liquid. Thus, at an early stage, a joint working group of treasury and central bank experts might well be established to encourage, monitor, and oversee market development.
Official Financing Facility
The availability of an official financing facility would help market development—especially in its early stages prior to the emergence of an active interbank and money market—by encouraging market-makers to take positions and carry adequate inventory, a necessary condition for a liquid market. Such a facility could take a number of forms depending on a country’s financial structure and traditions.11
In India, the central bank provides market financing through two semiprivate institutions, part of whose function is to provide liquidity to the market and to stimulate market making by other market participants. The Philippine central bank finances market activity through a broader group of authorized dealer banks. Poland has set up a group of primary dealers through whom open market operations are undertaken, but who also function as a core group for market development and financing through their obligation to participate in auctions, distribute securities, and provide two-way quotations.
Most countries have been moving toward the use of repurchase agreements as the most flexible and convenient form of financing. In that transaction, the price at which the central bank buys the security and the price at which the seller agrees to buy it back would be set so as to yield an appropriate loan rate for the period involved, overnight or somewhat longer. The rate can be established competitively, or it can be offered by the central bank based on its assessment of, and policy desires with respect to, current money market conditions.
The establishment of repurchase agreements as an effective money market instrument would smooth the way for a broader development of the money market among private sector transactors facing temporary shortages or surpluses of funds.12 Ultimately, the central bank would begin to have more scope for active two-way open market operations. To reach that point as soon as is practicable, the central bank would need to be sure that it administered any financing facility in such a way as to encourage market participants to seek financing elsewhere, such as from corporations with temporary surplus funds or from banks and other financial institutions.
The central bank should make it clear that availability of financing from it would depend on monetary policy, rather than strictly market, considerations. Nonetheless, in the early stages of market development a bit more consideration might be given to market needs, especially at times when market-makers have clearly made an effort to reduce their inventory in face of a financing shortage but, because of the underdeveloped state of the market, have been unable to find a buyer despite substantial price reductions.
In the early stages of market development, the central bank may also wish to consider whether a relatively favorable financing rate should be offered to encourage the emergence of active market-makers. If such an approach is taken, it should be clear to the market that any implicit subsidy is purely transitional and that financing costs will be based on market conditions as soon as the market is reasonably well established. But in considering any such temporary arrangement, the central bank should take into account the political problems that always arise with subsidies and the potential inconsistency of favorable treatment with efforts to develop the spirit of market competitiveness. It seems probable, all things considered, that market development may be most effectively enhanced by the availability of a form of official financing facility at a competitive rate rather than by favorable rate treatment, even for a transitional period.
Criteria for Business Relationships
In addition to establishing a financing relationship with the market, the central bank could also encourage market development by establishing ground rules for entities with which it is willing to undertake open market transactions. A number of countries conduct open market operations through so-called primary dealers, who have an obligation to make reasonable bids and offers when the central bank enters the market and also to bid realistically in treasury auctions in an attempt to win securities with some frequency and as a backup at other times. The primary dealers in turn help broaden and liquefy the market through their sales force and willingness to undertake transactions with other market participants and also by seeking sources of financing outside the central bank. Brazil, the Czech Republic, India, Malaysia, the Philippines, Poland, and Russia have all introduced, or are introducing, a primary dealer structure within the market.
In smaller countries or in markets at an early stage of development, the number of active market-makers or even of total participants may be so small that encouragement of a primary dealer market would be impractical and unnecessary. However, even in those instances, the monetary authority should make it clear to the market that it expects a certain level of performance if the entity expects to continue participating in transactions with the central bank. For example, entities would be expected to respond fairly regularly and realistically to the central bank’s bids and offers in open market operations and also to try to develop the customer base that would help buttress their participation and market knowledge.
When a market becomes large enough, however, there is much to be said for confining open market operations to a group of dealers, probably a relatively large one to minimize the potential for charges of favoritism. The central bank may wish to designate a minimum capital for such institutions, but otherwise regulatory and supervisory responsibilities over them are probably best left, as noted earlier, to authorities outside the central bank. A central bank should be willing to continue undertaking transactions with a dealer as long as the institution is participating actively enough as a counterparty in open market operations. However, before the market has become relatively well established, the central bank should probably also encourage dealers to develop their customer base since that will facilitate both open market operations and treasury debt management.
In addition, the central bank can use its business relationship with the market to encourage participants to establish market-making standards in transactions among themselves, such as the minimum size of a transaction to which a dealer is obligated when quoting a bid or an offer price. As a market matures, these and similar standards would ordinarily be set by the dealers as a group through a trade association or a self-governing regulatory body.
A central bank is the natural focus for the collection and dissemination of market statistics, especially for the government securities market, in view of the bank’s normal fiscal agency function for the treasury and its own open market concerns. In addition, the central bank would be the most relevant agency to assemble data on other short-term markets, particularly those in which commercial banks are the principal participants.
Data collection—including daily figures on positions, transactions volume, and financing by type of issue from individual institutions—should begin at the earliest stages of market development. The data should be collected on a confidential basis—that is, with assurance that an individual institution’s data will not be revealed to the public.13
The figures would provide the basis for surveillance of the market by the central bank’s market group. While the bank should not interfere in a market institution’s internal business decisions and assessment of the market, its ability to monitor a market-maker’s activity would permit it to make informal suggestions that may spur market development. In addition, of course, the availability of detailed data would help officials assess responsibility should questions about appropriate market practices arise.
As a market develops and the number of participants and market-makers increases, the central bank should be able to publish aggregate data. Publication should be as quick as is practical to enhance transparency of the market and public confidence. However, aggregate data cannot be published until there are enough transactors to avoid the risk of revealing the position and business of individual firms. In any event, publication should be undertaken with a sufficiently long lag—perhaps of a week to a month depending on the particular series—to avoid market overreactions.
Clearing and Payments Mechanism
No market functions effectively without reasonable assurance that securities will be delivered and paid for as agreed upon. In the initial stages of development, a market may be dominated by bilateral transactions in which the parties agree on the timing of delivery and payment. Over time, however, a market will function better—and be more “user-friendly”—if time lags are generally agreed upon, perhaps varying by type of security, with the shortest-term issues having the quickest turnaround time.
Because of the central bank’s general interest in promoting the development of a secondary market and its specific interest in ensuring that the reserve effect of its transactions is well timed in relation to market and policy needs, the bank should take an early lead in encouraging the market to set delivery and payment standards.14 The terms set by the central bank on its own government security or money market transactions with the market—whether for delivery and payment the same day, the next day, two days later, or even longer—would be one key element in setting overall market norms in those areas.
The speed and reliability of clearings and payments ultimately depend on the market’s technical capacity and institutional arrangements. Through their own actions, government bodies can encourage appropriate developments within the private sector. For instance, the central bank and the treasury can work together to establish up-to-date technology in the government securities market—such as a book-entry system to record security ownership and a procedure for simultaneous delivery and payment through the central bank’s deposit accounts.
The central bank can also play a powerful role in galvanizing effective and safe market clearing and payments mechanisms because of the leverage it has from its role as lender of last resort. Private clearing institutions will be needed as markets develop to minimize transaction costs and improve market functioning through multilateral netting procedures and through their guarantee of timely deliveries and payments. The central bank’s potential role as lender of last resort to the payment system—together with its, and the market’s, interest in minimizing the chances that it would ever be called upon to fulfill that role—gives it an important voice in ensuring that business arrangements and practices are as risk free as possible.
In particular, the monetary authority would have to be sure that clearing institutions obtain adequate credit lines from banks to backstop potential obligations stemming from delivery and payment failures. In that context, while the central bank’s lending facility provides assurance against systemic crises, both market participants and banks should be made to understand, if they do not already, that all efforts must be made to avoid actual use of the facility since such necessitous borrowing—whether at a penalty rate or not—could in and of itself further reduce public confidence in a market’s viability.
Conduct of Open Market Operations
The conduct of open market operations in developing and transitional economies has varied with market conditions, the structure of the banking system, the degree of financial system deregulation, and the availability of government securities or private money market instruments. It has also been affected by the central bank’s state of readiness, including the availability of a trained staff, an adequate statistical base, and a financial framework for guiding operations.
Information and Criteria for Guiding Daily Operations
As previously described, open market operations can be conducted either actively by aiming at a quantity of reserves15 or passively by aiming at a money market interest rate or at a closely related measure of the banking system’s liquidity position. Whatever the approach to implementing open market operations, however, the central bank must organize itself as early as possible to collect figures on the supply of and demand for bank reserves, including clearing balances.
Data on the public’s deposits in banks have the greatest impact on the day-to-day demand for reserves. In addition, when a money supply measure is serving as a strong intermediate policy guide, an up-to-date flow of deposit data would permit the monetary authority to implement changes in money market or bank reserve conditions sooner rather than later as needed to offset undesired trends in the money supply and related measures of aggregate bank reserves. However, in economies undergoing rapid growth or transition accompanied by deregulation, the central bank would need to be especially alert to changes that affect the significance of various measures of money (narrow and broader) as the public’s saving opportunities and habits are altered. Such a situation reinforces the argument for collecting data on close substitutes for bank deposits early in the market development process.
But even when the monetary authority is not especially concerned with money behavior, the prompt availability of deposit data will enable the central bank to make better projections over an operating period of the demand for the required reserves and clearing balances that serve as a fulcrum for gauging the market effect of open market operations. To determine how many of the reserves demanded should be supplied through the open market function, estimates would also have to be made of other elements that affect the supply of bank reserves, including the government’s deposits, currency in circulation, foreign exchange, and float (arising from timing differences between crediting and collecting funds in the clearing and payment process at the central bank). Many of these estimates require close cooperation with the treasury.
Each day, revised estimates of these elements, together with estimates of banks’ required reserves and clearing or other balances at the central bank, will yield an up-to-date figure for the amount and direction of open market operations needed to tighten or ease pressure on the money market and banking system or to keep it unchanged. The statistics put together on the demand for and supply of bank reserves would provide a basis for judging the extent to which these pressures are likely to change from factors other than open market operations. They would be the principal guide for the monetary authority’s daily decision about the timing and extent of its own market intervention, with the amount depending on whether or to what extent the central bank may want to encourage or tolerate changes in money market or bank reserve conditions.
In looking to the particular reserve statistic that would be a key to daily action, the central bank may take a targeted level of aggregate reserves provided through open market operations (or that part represented by net domestic assets in those countries that focus on that measure) and let the pressure on bank reserve positions be what it will, given banks’ demand for reserves. Alternatively, it may use a predetermined level of pressure on bank reserve positions and let the amount of reserves provided through open market operations be what it will in response to the demand for reserves.
In determining the size, timing, and direction of open market action, the accuracy of the reserve estimates, which are always subject to some uncertainty and to revision, would in practice need to be judged against incoming evidence from the interbank or money market about emerging pressures as seen in interest rate behavior. The interpretation of factors behind money market rate movements should be aided by the central bank’s continuing contacts with the market. Traders from the central bank’s open market function should be continually speaking with market traders in an effort to understand the forces behind ongoing market conditions and also to enable monetary policy decision makers to better assess market psychology.
Apart from the role of the money market rate, and the tone of the market, in gauging the accuracy of statistics on factors affecting bank reserves, a short-term market rate, in particular an overnight rate, may also usefully serve as the primary guide for day-to-day open market operations once the interbank part of the money market becomes sufficiently well developed. An overnight money market interest rate has a technical advantage over a measure of bank reserve pressures (such as excess or free reserves) as a guide because it would permit the central bank automatically to accommodate changes that may occur in banks’ demand for excess or free reserves, given market rates. If such shifts in demand were not accommodated, they could adversely affect growth in money and bank credit, holding it back perhaps procyclically when there was an unexpected upward shift in the demand for free or excess reserves and accelerating it when there was a downward shift.
Using a money market rate as a day-to-day operating guide would not lessen the need for prompt collection of statistics on factors affecting the supply of and demand for reserves. Given the generally close relationship between money market rates and the degree of pressure on bank reserve positions, the estimate derived from those statistics of what would happen to free reserves (the difference between the reserves provided through open market operations and the required reserves and clearing balances demanded by banks) in the absence of open market operations would provide the starting point for determining the likely amount of open market operations needed during a particular operating period. Without an adequate statistical base, the central bank would have difficulty judging whether daily money market rate movements are or are not temporary and would be at greater risk of providing reserves more suited to the market’s convenience than to the central bank’s own monetary objectives.
The precise day-to-day operating target used in practice in the developing and transition markets examined, and the degree of flexibility in their use, is not always clear from available information. Nonetheless, it would appear that a short-term interest rate or some measure of bank liquidity pressure represents the predominant day-to-day policy guide in those countries where markets have attained some degree of maturity.
Bank liquidity or reserve projections are apparently a key guide to operations in the Czech Republic, India, Indonesia, Mexico, Malaysia, and Poland. Poland employs a free reserve measure, while the Czech Republic uses banks’ excess reserves. In the latter country, the discount window is effectively closed, as noted earlier, with the discount rate chiefly serving as an indicator of the central bank’s monetary stance. In India, interest rate liberalization in recent years has led to a shift in operating targets toward a measure of bank reserve pressure and away from an earlier emphasis on maintaining orderly conditions in the market for government securities to facilitate debt management.
Other countries appear explicitly to target a short-term interest rate. Thailand, the Philippines, and Brazil employ an interbank rate. Egypt, which does not have an active interbank market, aims at a treasury bill rate.
This rather general focus on using what might be broadly termed money market conditions (whether represented by an interest rate or by bank reserve pressures) as a day-to-day operating guide is widespread although different countries use different intermediate policy guides. Net domestic assets have been taken as an intermediate guide in Poland and Mexico, base money in the Philippines and Brazil, M2 in the Czech Republic and Indonesia, M3 in India and Malaysia, and the foreign exchange rate in Egypt. Thus, central banks in emerging markets that have reached a certain level of competitiveness and flexibility have—like central banks in major markets—generally decided to conduct open market operations on a passive basis, allowing themselves considerable flexibility in affecting the degree of pressure on the banking system and the basic cost of liquidity in the economy.
Instruments for Open Market and Open-Market-Type Operations
The experience of developing countries and economies in transition with using specific market instruments to achieve their operating objectives has depended on a variety of factors specific to their individual circumstances. These have included the availability of government securities, the relative size of primary and secondary markets, the existence of a competitive market for interbank funds, and more generally the extent to which governments have been willing to deregulate and rely on market processes for encouraging and channeling saving into productive investments.
Without a developed secondary market in securities, central banks would be, and have been, limited to open-market-type operations in the primary market, such as auctioning newly issued securities to absorb reserves or auctioning central bank credit to provide reserves. Instruments that have been and might be used in the absence of suitable secondary markets and a sizable government debt will be evaluated first in this section of the paper.16
As has been noted, such instruments represent the initial steps that can be taken as policy implementation evolves toward the use of more fully flexible open market operations. In practice, they are best accompanied by actions discussed in the section “Market Prerequisites and the Role of the Central Bank” if market participants are to be made aware of the central bank’s interest in furthering the development of markets in an effort to achieve greater flexibility in operations.
An open-market-type operation that the central bank uses to absorb excess liquidity involves issuing either new treasury or central bank securities into the primary market. The Czech Republic and Ghana employ both. Egypt auctions only treasury bills to absorb reserves, but also mops up liquidity through commercial banks’ placement of time deposits directly with the central bank.17
The Philippine central bank has issued treasury securities and sold its own bills in the primary market as monetary instruments, but using both instruments created some market confusion and difficulties, as will be briefly discussed later. In Indonesia, a country without government debt, the central bank, as noted earlier, auctions its own bills as a way of absorbing excess liquidity and purchases bank-endorsed paper to provide reserves to the banking system.
If the central bank offers a new treasury security, it should be considered a monetary operation, rather than a government debt-management operation, only if the incoming funds are not available to the government for spending. This can be accomplished in several ways, one being to cede to the central bank control over an equivalent amount of the treasury’s deposit balance at the bank. However, the cleanest approach is to place the funds in a special account that is created purely for monetary policy purposes and that cannot be used to finance spending by the treasury, or the central bank for that matter, under any circumstances.
Such an account would ensure that bank reserves were “permanently” reduced by the operation. The treasury could not increase reserves by drawing down the balances. If it should turn out that the central bank had in fact overestimated the reserve surplus when it initially issued the securities and subsequently needed to provide reserves temporarily, it could buy the securities back before maturity and credit banks’ reserve accounts, leaving the special account balance unchanged.
Such repurchases before maturity, perhaps followed by subsequent resales, would have the ancillary advantage of helping to create the basis for a secondary market among private sector participants. Secondary market trading in such securities as a means of redistributing reserves around the banking system and developing an active money market should be encouraged as a step toward further market development.
A special treasury account set up in connection with these open-market-type securities should receive interest at a market rate. An explicit interest rate could be set for the account based on market conditions. However, if the central bank distributes basically all of its net earnings to the treasury, the government would, in due course, receive the equivalent of market interest, and an explicit rate for the account would be redundant.
The drain on government revenues from sales of special monetary policy securities to the public is roughly the same as occurs when the central bank sells any security from its portfolio, auctions new securities of its own, or reduces its loans to the banking system. Thus, use of the special security would not make treasury revenues less than they would otherwise be, assuming that the central bank does not as a result alter its monetary target from what it would otherwise be.
Since the cost to the government is about the same for securities issued by the central bank to the public for reserve-absorption purposes and offerings of special treasury issues for monetary policy purposes, the choice between the two depends on institutional and market considerations. A central bank’s own security can be a useful, if not necessary, open-market-type instrument in countries like Indonesia, where there is no domestic government debt, or the Philippines in the first part of the 1980s, when the central bank did not have access to sufficient government debt. Experience in the latter country illustrates some of the difficulties in a situation in which both governmental and central bank market instruments coexist. The principal problem would appear to be public confusion about the relationship between the two. In the Philippines, the government came to view the central bank’s use of its own instruments as complicating interest rate and debt-management policies by encouraging market segmentation in a thin overall market. The development of an active government securities market was apparently being retarded, rather than aided, by large-scale issuance of the central bank’s own bills.
Moreover, the central bank was taking large losses from other operations (such as foreign exchange transactions and the restructuring of weak commercial banks) that were being exacerbated by payment of interest on its own securities. Thus, the credibility of the central bank as a viable institution for anchoring the financial system was coming under question. In the end, by 1993, the central bank was restructured and the new bank received a broad portfolio of treasury securities to facilitate market operations—but with operations in such securities confined to the secondary market rather than being undertaken in the primary market.
Experience in Brazil contrasts with that in the Philippines. The overall market in the former country is relatively broad, and central bank securities issued for monetary control purposes have traded well along with government securities issued for debt-management purposes.
Open-market-type operations in the form of issuing treasury or central bank securities in the primary market are most practically used when excess liquidity is piling up in the banking system—for example from large capital inflows. The securities issued might also be employed to adjust to the ebb and flow of liquidity needs in the course of a year as the central bank adjusts issue volume and, conceivably, buys back before maturity. Still, such operations do not provide the same flexibility as open market operations in the secondary market for coping with fluctuations—unexpected as well as expected—in liquidity and the demand for bank reserves. For one thing, in the absence of an active money and interbank market, the central bank’s open market function would still be deprived of the ongoing information about actual and emerging liquidity conditions that is conveyed through continuous competitive interest rate determination. Thus, it would be more difficult for the monetary authority to plan the size and timing of operations to absorb reserves, and the actual outcome of operations may be subject more to the vagaries of a primary market bidding process than to the central bank’s prior intentions about a particular interest rate or a particular amount of reserves.
In addition, open-market-type operations do not provide the reserves needed by the banking system over time to support expansion in money supply and bank credit for economic growth. For that purpose, countries without, say, sufficient foreign exchange inflows have turned to such market-type instruments as credit auctions in the absence of a developed secondary market in government securities or interbank instruments.18 Of the countries surveyed, the Czech Republic, Mexico, Russia, and Tunisia have used credit auctions at one time or another and to one degree or another. Indonesia’s regular purchases of SBPUs might also be viewed as a form of daily credit auction.
The auction of credit through the central bank’s open market function differs from credit made available through the discount window in crucial ways. First, the central bank can use a credit auction to control the amount of reserves it supplies, while the banks determine the interest rate through bidding. At the discount window, the central bank normally sets the rate, whereas the amount lent, and therefore the amount of bank reserves, would be at the initiative of the commercial banks, and not the central bank. Second, the central bank might be able to resell paper acquired in the auction at its initiative into a secondary market; loans at the discount window would, by contrast, normally be repaid on a schedule or at the initiative of the borrowing bank. In general, in contrast to the discount window, a credit auction gives the central bank more initiative on the timing, amount, and price at which reserves are supplied and could also provide some additional flexibility for draining reserves once the instruments are created. The Indonesian experience, though, suggests that it may be difficult to develop a secondary market in such bank paper.
Credit auctions play a useful role in the transition to more flexible markets in part because they can help introduce marketlike processes at an early stage. However, at times, they appear to have impeded the development of an active interbank market by encouraging banks to look directly to the central bank as the principal counterparty in managing their liquidity needs. Nonetheless, in a country like Malaysia, where an active interbank market has evolved, the central bank has been able to function in the market without drying it up. Indeed, the Malaysian central bank employs direct borrowing and lending with banks on a bilateral basis as a principal instrument of monetary control.
While credit auctions and other means of acquiring private debt may be useful monetary tools for providing reserves in countries without a significant floating supply of government debt, reliance on them does raise certain questions about the liquidity of the central bank’s balance sheet and the institution’s stature with the public as a truly safe guardian against both systemic risk and inflation.
Admittedly, confidence in the central bank and the government is based more on past policies and an assessment of their current posture than on balance sheet structure. Nonetheless, any questions about balance sheet structure might be lessened in countries without a significant government securities market if expansionary open-market-type operations were conducted at least in part in special treasury debt issues offered for such purposes. Special treasury obligations have been used to absorb reserves, as described earlier.
A different kind of special treasury obligation with certain special features could also be used to expand reserves. Once created, such an obligation would remain on the asset side of the books of the central bank to support growth in the banking system’s reserves if the central bank were given the right to transfer to private banking institutions an associated special treasury deposit liability. The deposit liability, which would not be controlled by the treasury, would be created simultaneously with the special security at the time the central bank wished to expand reserves. At that point, the deposit liability would be transferred to the banking system, either through an auction of a predetermined amount or through a distribution based on, say, bank size.
An auction is the preferable approach—in effect an open-market-type operation in the form of a primary market auction of bank reserves conveyed through treasury deposit liabilities. Whether auctioned or not, the special treasury deposits would be priced to yield a market return to the treasury and would determine the interest rate on the special treasury security held by the central bank. This market return would represent seigniorage on money creation that the treasury would have received in other circumstances if the central bank had been able to purchase securities in the open market (thus, in effect, retiring debt and saving the interest cost). However, in the case of the special obligation, the seigniorage would in effect be paid by private banks holding the deposit liability and not by the central bank.19
The interest rate should be based on a short-term maturity of, for example, three months or so. Banks would have the opportunity to bid for the treasury deposit liabilities every three months; as an alternative to reduce uncertainty for the banks, the central bank could reset the rate on the basis of market conditions and give existing liability holders the right to roll over at that rate.
Both the special issue and deposits could be marketable, although secondary market transactions should probably be limited to those consistent with the instruments’ basic reserve-supplying function. Banks could sell the special deposits to each other, thereby redistributing existing reserves. These deposits would become one instrument, among others, in a developing interbank market.20 But it would seem to be undesirable in practice for the central bank to also engage in secondary market transactions in special deposits with banks—for example, to buy them back to absorb reserves. Since the basic purpose of creating the special deposits would be to provide permanent reserves backed by highly creditworthy government securities, transactions in special deposits to reduce reserves would tend to confuse the issue.
Conceivably, the central bank could also sell the special debt instrument into the market to mop up excess reserves, but that too should be kept to a minimum practice for the reason noted above. However, without jeopardizing its fundamental market role, the instrument could well be used as collateral to absorb reserves temporarily through reverse repurchase agreements.
The treasury would of course need to guarantee payment of interest on the security for it to serve as a highly liquid, safe instrument. Otherwise, because the interest on the issue would depend only on banks’ ability to pay for the special deposit, the security would be no more creditworthy than ordinary bank paper and would not enhance the central bank’s credibility.
Flexible Open Market Operations
As their economies expand and markets mature, most countries have tended to place more emphasis in policy implementation on the use of highly flexible open market operations in the secondary market. Of the countries surveyed, Brazil, the Czech Republic, India, Malaysia, Mexico, the Philippines, Poland, Russia, and Thailand have, with varying degrees of progress, moved in that direction. Open-market-type operations have often continued to play a role, however, with the extent depending in part on the state of secondary market development. Brazil, Malaysia, Mexico, and the Philippines are examples of countries operating policy with a blend of open market and open-market-type instruments.
In each of these countries, however, secondary market transactions—mainly in the form of repurchase agreements and reverse repurchase agreements or in the interbank market directly with banks—appear to be a key instrument of monetary control, if not the principal one. For a while, Thailand and Poland relied more or less entirely on repurchase agreements in the secondary market to implement policy, abjuring primary market operations. Poland, however, recently introduced outright operations as the secondary market developed further. Thailand has also recently reduced its heavy reliance on repurchase agreements for implementing monetary policy with the introduction of auctions of the central bank’s own bills.
With the exception of Malaysia, these countries have generally developed a fairly active government securities market that has facilitated the transition to greater use of open market operations. The authorities in Malaysia, faced with insufficient government securities for the full scope of operations, undertake operations directly in the interbank market with banks, as noted earlier, and also employ an auction system to allocate government deposits among commercial banks.
Other countries have developed operations in the private market. Indonesia has taken steps to develop a secondary private market, but it has not yet moved significantly beyond open-market-type operations in the primary market for policy implementation. However, the operations it carries out are quite flexible and attempt to replicate the functioning of a secondary market.
Bank Indonesia absorbs reserves by auctioning its own issues to the market daily and provides reserves by purchasing various forms of bank paper in daily auctions. Because the secondary market has remained thin for a variety of reasons, including the wide diversity of issues with different risks and the absence of a requirement on market-makers for minimum bids and two-way quotations, Bank Indonesia has been forced to act as an intermediary between borrowers and lenders by adjusting the relative size of its daily auctions. The awkwardness of this technique and the absence of a well-functioning secondary market mean, however, that policy implementation is quite complicated. This may at times have contributed to Indonesia’s difficulties in achieving policy objectives and effective coordination, with other factors and instruments affecting the bank reserve base (such as capital inflows and foreign exchange swaps). Of course, even with well-developed secondary markets, countries often have difficulties in coordinating operations to balance interest rate and exchange rate objectives.
In the transition to open market operations as a major policy instrument, repurchase agreements in government securities are clearly the most useful and prevalent type of operation. Compared with outright operations, they interfere less with the further development of secondary market trading in outstanding securities because they essentially provide temporary financing of reserve shortages or surpluses and do not directly influence the basic supply of and demand for the underlying security that serves as collateral.21 More positively, they enhance the liquidity of the underlying securities and in that way help develop a more active secondary market with enough depth and breadth to accommodate outright transactions by the central bank. While repurchase agreement transactions are generally quite short term, the underlying collateral would comprise both longer-term and short-term securities, thus adding liquidity to all sectors of the market.22
The use of repurchase agreements with a short maturity should also make it clear to the market that the central bank is encouraging participants to develop as many alternative short-term borrowing or lending sources as possible. The development of what is in effect a money market would facilitate the redistribution around the banking system of the aggregate reserves supplied by the central bank—obviously a necessary function if open market operations are to replace the less impersonal and more difficult to administer one-on-one relationship between the central bank and depository institutions.
Even apart from their value in contributing to market development, repurchase agreements and reverse repurchase agreements can be expected to be the dominant form of open market operation, as country experience has shown. They are ideally suited to offsetting the short-term fluctuations in factors affecting bank reserves that are the major influence on day-to-day market operations. The central bank can set the maturity of repurchase agreements, which can therefore be timed automatically to reverse themselves as circumstances change. The central bank could also permit them to be withdrawn before maturity if it wished. That would be most appropriate when there is particular uncertainty about the actual size of operations that may be needed; it would be a way of letting the market itself adjust automatically if, for example, a reserve surplus (that the central bank is absorbing through reverse repurchase agreements) turned out to be smaller than expected.
Because they are generally much larger than outright transactions, repurchase agreements are also useful for offsetting large shifts in liquidity conditions that might be caused, for instance, by a wave of capital inflows or outflows. In that case, maturities would generally be longer than the maturities of up to one week that are most useful for smoothing out the money market. Even when maturities are short, though, the central bank can continuously roll over the transaction as needed to meet its liquidity objective.
Experience with use and maturity varies among the countries examined, but short-term operations tend to dominate. In Brazil, where repurchase agreements are the main instrument of monetary control, operations are undertaken daily through informal auctions, with maturities generally overnight. Mexico and Poland also undertake relatively frequent operations with short maturities. Thailand employs an elaborate auction process twice a day for repurchase and reverse repurchase agreement transactions, with maturities ranging from overnight to six months, although the most popular maturities are at the short end of the market. Maturities appear to be the longest in the Philippines, where reverse repurchase agreements are used to absorb liquidity, with maturities commonly between one week and one month and with a maximum of one year. It might be noted in passing that, although the central bank in Argentina announced the suspension of rediscounts and outright open market operations in shifting toward a currency board arrangement for monetary policy, it continues to use repurchase agreements and reverse repurchase agreements to smooth out day-to-day market liquidity conditions.
Outright purchases and sales of treasury securities in the secondary market are also used in most of the countries examined as their secondary markets have developed. They are often used, as in Brazil, to provide or indeed absorb reserves on a more permanent basis. In India and the Philippines, outright transactions are considered to be an important instrument of monetary control and are undertaken daily—in treasury bills in the Philippines.
In general, when secondary markets are in the early stages of growth and trading is comparatively thin, outright transactions by the central bank run a high risk of dominating the market and impeding its further development. This is particularly likely for longer-term market sectors.
As explained in the preceding section, sales outside the short-term area of the market might well be avoided in an effort to keep market risk, from the viewpoint of the participants, to reasonable proportions. And excessive purchases in the longer-term area risk weakening the market’s incentives to develop a broad customer base, not to mention diluting the market’s capacity to serve as an independent source of information for the central bank about the public’s attitudes and expectations. Thus, outright purchases and sales of securities should generally be centered on the short-term sector of the treasury market where there is normally a more continuous and sizable volume of activity.
In all transactions, outright or repurchase agreements, market development would be most encouraged by use of a competitive price bidding technique involving all eligible market participants. More limited go-arounds of a few participants may be useful at times, but they tend to be viewed as unfair and do little to enhance a broadening of the market.
There are both advantages and disadvantages to indicating the size of the operation contemplated by the central bank. The chief advantage is that markets are given a key piece of information for pricing and sizing their offers or bids. The main disadvantage is that the central bank loses the flexibility to adjust the size of its operation on the basis of market response, although it can, as debt managers often do, reserve the right to vary the aggregate size within a small range (such as 10 percent of the initial offering).
On balance, it can be argued that the amount should be indicated for outright transactions, which would in any event probably be relatively moderate in size and designed to provide reserves over the longer run. For repurchase agreement transactions, on the other hand, the central bank may not wish to indicate an operation’s size to the market in advance. Because the average interest rate on such transactions would generally be a closely watched indicator of the central bank’s current policy stance, the central bank may need to leave itself enough room to adjust transaction size so as to avoid the potential of an interest rate outcome that misleads market participants.23 Moreover, on more technical grounds, the central bank may also conclude that the intensity of market response to the proposed transaction would in some circumstances be a better measure of market shortages or surpluses than the bank’s own statistical estimates.
Even if the amount bid or offered by the central bank in a repurchase agreement transaction is not indicated in advance to the market, the approximate size—based on current statistical information and expected market performance—should first be decided internally by those in the central bank responsible for policy (as should also be the case for outright operations of course) before the manager of the open market function is given permission to enter the market. This would ensure appropriate lines of responsibility within the policy process. The manager could be given the option of varying the amount within a range, but substantial variations should require a further policy decision.
In the interest of transparency, the results of an open market transaction—or indeed an open-market-type transaction—ought to be announced to the market as promptly as possible. Depending in part on the technique of the offering, this would include the amount taken, the average price, and the stop-out (the highest or lowest price accepted depending on whether the central bank is a buyer or a seller). Especially in the early stages of market development, a central bank whose operations are transparent may acquire more of the public stature that would enable it, partly through moral suasion, to spur those developments in the microstructure of markets that would increase activity and the market’s reliability as a source of information to policymakers.
Government Deposits and Debt Management
Government decisions about debt management and use of its deposit balance have a number of implications for open market operations and vice versa. It is recognized that they can excessively complicate or facilitate operations. In all countries surveyed, the treasury and the central bank work together on these issues with varying degrees of tension and power distribution.24 On pure debt-management decisions, the treasury in most cases makes the final decision and the central bank serves as fiscal agent. In areas that have a direct effect on the bank’s reserve base, the central bank normally has a stronger input.
The particular structure of the working relationship between the treasury and the central bank differs in accord with the traditions and financial history of the country. In India, for example, the Reserve Bank has acquired and held in its portfolio ad hoc treasury bills at a below-market interest rate to finance the government deficit—a practice that is planned to be phased out in a couple of years—and has made offsetting sales of other instruments bearing market interest rates at its discretion. In Brazil, the treasury manages the debt in consultation with the central bank; with regard to debt issued in the form of notes, the treasury sets the overall size, and the central bank decides on the amount to be kept in its portfolio, with the rest auctioned. China is moving toward a new debt-management system to help break the direct debt-financing relationship between the central bank and the government. In the new approach, the finance ministry will be responsible for issuing bonds, but the central bank will take responsibility for placing them by organizing a national underwriting syndicate.
Regardless of the form of the operating relationship between the treasury and the central bank, open market operations will be most effective when the central bank has control over factors that affect the reserve base of the banking system. Thus, when the central bank does in effect underwrite treasury debt, it must be free to sell or auction the debt into the market as required to meet monetary policy objectives. It would be most desirable, though, if the treasury sold the debt directly into the market, avoiding any potential conflict between debt-management and monetary policy needs.
Moreover, such sales should be in the form of auctions, as they are in most countries, to promote the development of a competitive and deregulated market system and also to avoid pressure on the central bank to hold policy on an “even keel” for the purpose of facilitating primary market issues when offered at a predetermined rate. Market-makers or dealers should be required to make reasonable bids in the auctions as part of their market function and to retain their role as counterparties with the central bank in its open market operations. As noted earlier, this would help further secondary market development as the market-makers in turn seek to expand their customer base.
From the perspective of day-to-day open market operations, it is particularly important for the central bank to be able to influence, if not control, the treasury’s operating cash balance with the bank. It is highly unusual, however, for a central bank to be given substantial discretionary control over government deposits. Notable exceptions are the Bank of Canada, which has the right to transfer government deposits between itself and commercial banks; Bank Negara Malaysia, which as noted earlier auctions such deposits to banks as an instrument of policy; and the Bundesbank, which has a veto on the government’s ability to hold deposits outside the central bank.
The problem of coordination arises, of course, because variations in the government’s balance held with the central bank affect the reserve base of the private banking system. Central banks use open market operations in the money market to offset the impact of variations in treasury receipts and expenditures on the treasury balance and hence on bank reserves. Sometimes, however, sufficient collateral may not be readily available for the size of such operations. Thus, it is often simpler if the government keeps its balance at the central bank around some agreed operating level to the extent practicable and draws on or redeposits funds with commercial banks to offset the impact on the balance of variations in tax receipts and expenditures. Such an approach will have particular appeal when the money market is not very highly developed, but it is also followed in the United States.
The treasury’s operating cash balance with the central bank can also serve as a day-to-day monetary instrument for hitting reserve or money market rate objectives, as in Canada. But without a clear agreement between the central bank and the treasury, the instrument can become a two-edged sword and lead to market confusion. Agreement is needed about its use, for instance, to avoid conflicts between the treasury and the central bank or, even worse, the emergence of a market belief that the central bank cannot control its policy objectively if variations in the cash balance happen to lead to an easing, or avert a tightening, of money market conditions at a time when a treasury financing is in the market.
In general, open market operations will function most effectively when the government abides by, and the public believes in, a clear division between debt-management and monetary policy operations. Markets in a deregulated, competitive environment will develop most effectively when the division between monetary policy operations and the government’s debt- or cash-management policies is clear and well adhered to.
With regard to the treasury’s operating cash balance, this would involve agreement either to neutralize its monetary effect or to delegate substantial control over it to the central bank. With regard to debt management in general, this would involve use of public auctions in one form or another as the basic technique for primary market financing, leaving the central bank free to determine monetary policy on purely economic grounds. For coordination purposes, and given the central bank’s expertise in markets and its usual role as the government’s fiscal agent, in virtually all countries debt-management decisions are made with ongoing input, both informally and through formal committee structures, from the central bank and often from active market participants.
There is much to be said for establishing the basis for the central bank’s use of open market instruments as soon as possible. Deregulation of markets domestically together with the widening impact of globalization make it almost impossible for any country in the process of growth or transition to employ direct instruments of monetary control for long without adverse side effects. While the central bank cannot transform its own modus operandi much ahead of ongoing changes in market structure, it can and should encourage those changes in market structure that would make it better able to conduct open market operations. That should also help, through example supplemented by moral suasion, to accelerate the transition of domestic markets as a whole in the direction of greater consistency with major world markets.
While this paper has not undertaken the econometric research that would be required to determine whether countries with open market operations perform economically better than those without—probably a nearly impossible task given all the variations in circumstances that would need to be allowed for—experience suggests that in practice countries act as if they have little choice except to begin the transition toward greater use of market instruments in monetary policy. Both developing and transition economies have taken steps in that direction consistent with the market’s stage of development, although some have lagged behind market developments and others have more actively attempted to lead them.
Countries where central banks appear to have lagged behind in adapting policy instruments to market developments or in encouraging market deregulation have found that they were hampered in meeting their own policy objectives. This has often been a spur to official action to hasten the transformation by adapting market instruments for policy implementation in line with market evolution or to speed up the market deregulation in order to make such instruments more usable and effective. Indonesia in the early 1980s took the latter approach. The restructuring of the Philippine central bank in the early 1990s is more of an example of the former, as it shifted away from monetary policy intervention in the primary markets to sole use of developing secondary markets.
The availability of market instruments to the central bank is no guarantee that monetary policy or the country’s economic performance will be continuously satisfactory. Mexico’s recent experience is only one of a number of cases in point. In that country, active secondary markets had developed, and the central bank could implement policy through highly flexible operations in those markets as well as through use of open-market-type instruments in the primary market. Clearly, an arsenal of open market instruments in tune with the evolving market structure should be considered as at most a necessary, but not sufficient, condition for an effective monetary policy.
Most countries that are moving away from direct instruments of monetary control to market instruments have also begun to make the complementary adjustments in reserve requirement structure and in discount window and rate administration that are needed to make market operations more effective in policy implementation. With regard to reserve requirements, adjustment has involved a gradual reduction in reserve ratios to reduce the implicitly greater tax on the banking system relative to other financial institutions. For the most part, however, reserve ratios appear to remain at “binding” levels that permit banks’ reserve balances to better serve as a fulcrum for policy implementation and a signal to policymakers. The discount window, insofar as very short-term adjustment borrowing is concerned, has begun to be administered more restrictively to be consistent with an emphasis on policy implementation through reserves provided by open market operations. But it has for the most part remained available as a safety valve, so that its use may reflect the degree of pressure on markets and help smooth the process of market adjustment. In some countries, such as Mexico, reserve requirements have been eliminated and the discount window effectively closed except for emergency lending.
It is not inconsistent with countries’ experience to suggest that a complementary mix of instruments, rather than exclusive reliance on open market or open-market-type operations, may best suit countries with emerging markets or that are in transition. In those countries, markets are still being formed and are generally not especially robust. Moreover, the markets and the banking system may be prone to sudden surges of liquidity excesses or deficiencies, resulting from shifts in international capital flows and in domestic psychology that unfortunately may be generated by uncertainties that accompany rapid transformation of a country’s economy and financial system. The central bank in such circumstances would generally be well served by a balanced set of instruments—with open market instruments playing an increasingly important role as markets develop—to cope with contingencies that threaten to overwhelm markets and their capacity to act as a direct conduit for policy operations.
The transformation to effective use of open market instruments in policy implementation usually involves two stages of market development. In the first, secondary markets are weak or virtually nonexistent, and the central bank implements policy through open-market-type operations in primary markets. In the next stage, secondary market activity has evolved to the point where the monetary authority can increasingly employ the more highly flexible operations in those markets through either repurchase agreements or outright transactions. Many central banks have continued to use a mix of primary and secondary market operations as markets have matured. Others have shifted to secondary market operations only, as is done by central banks in major developed markets.
Virtually all countries examined have accomplished the initial transition from direct monetary instruments to open-market-type operations in primary markets as they began to deregulate markets. That transition has been relatively simple in part because the market is not, and need not be, significantly larger than the auction in which the central bank plays a key role. But the next step in the transformation of the central bank’s policy instruments—to using highly flexible open market instruments in secondary markets—has proved to be more difficult because it requires substantial private market activity independent of the central bank.
To have well-functioning secondary markets, the volume of activity must be quite large and must develop almost wholly within the private sector itself. Such a development depends on a host of factors largely outside the central bank’s control—including in particular the country’s stage of economic development or transition—but also on some that it can at least strongly influence—such as the legal, regulatory, and payments infrastructure needed to inspire confidence in potential market participants. Steps that the central bank can take to help encourage a positive market infrastructure are discussed at some length in an earlier section. A central bank’s open-market-type operations in the primary market should also be viewed as a way station toward operations in the secondary market and be designed, therefore, to minimize continuing reliance by individual institutions on the central bank for adjusting their liquidity positions.
It is difficult, however, for the central bank to accelerate development of the secondary market by undertaking actual transactions, especially outright transactions, in that market in the early stages of its development. If attempted in a thin market, such transactions, particularly if sizable, risk dominating the market and forestalling its development. In practice, the central bank’s open market operations in secondary markets would need to remain a relatively minor part of total market volume if the open market instrument is to be truly flexible in its use—that is, permit the central bank to enter the market on either side on a more or less continuous basis, with size and timing at its own initiative and with the interest rate determined competitively. Repurchase agreements and reverse repurchase agreements would appear to be the most effective route for encouraging further market development through operations since they add to the liquidity of the underlying securities collateral and do not directly interfere with the market supply of and demand for the securities.
An active secondary market is most likely, and easiest, to develop in government securities and at the short end of that market. Assuming that the government and central bank are pursuing a stable financial policy, such a market has the advantage to the central bank and private participants of being essentially free of credit or default risk, permitting the securities to be readily traded and used as collateral. Absent a sizable outstanding volume of such securities, however, the central bank would have to conduct open market operations in short-term private debt. For those purposes, development of an active interbank and money market is crucial.
Indeed, it is crucial for the central bank to attempt to develop an interbank market in any event, whether or not a government securities market exists. The monetary authority can help market development by, for example, adjusting its policy instruments to discourage banks from interacting with the central bank rather than with each other and taking other steps to undergird market structure, such as encouraging a reliable payment mechanism. Once developed, the market can be employed for policy operations. But equally, or more, important, it would provide the central bank with signals about the degree of liquidity pressure in markets, which are needed to help interpret the significance of incoming statistics on factors affecting bank reserves for the amount and timing of open market operations.
There are, nonetheless, risks for a central bank if it relies excessively on operations in private money market paper, presumably mostly bank paper, to support expansion of the monetary base needed for economic growth. The central bank may have no choice in the absence of a sizable open market in government securities or a sufficient and growing volume of foreign exchange. But the risks are that the central bank’s portfolio may become quite illiquid and that over time, as a portfolio of private paper builds up, the market may come to believe that the central bank is no stronger than the private sector itself. The market credibility and stature of the monetary authority might well then be reduced, and the risks of financial crises heightened to some degree—although, admittedly, the basic influences on a central bank’s stature are its past policy record, the credibility of its current leadership, and assessment by the market of the government’s basic commitment to a stable financial policy and to support of the central bank.
In the absence of an active government securities market, use of a special treasury debt issue to support expansion of the monetary base might be considered as a supplement to private paper to strengthen and diversify the central bank’s market portfolio. To absorb reserves, a number of central banks in developing countries have used open-market-type operations through primary market auctions of treasury issues or of the central bank’s own obligations. As explained in the previous section, a special treasury issue can also be used to provide reserves for growth—substituting in some degree for the acquisition of private paper—by giving the central bank the authority to auction to banks the special treasury deposit that would be created along with the special debt issue. In the end, the central bank’s asset portfolio, with the special debt issue guaranteed by the government, may appear more creditworthy to the market.
AlexanderWilliamE.TomásJ.T Baliño and CharlesEnoch1995Adoption of Indirect Instruments of Monetary Policy IMF Occasional Paper No. 126 (Washington: International Monetary Fund).
AxilrodStephenH.1995“Transformation of Markets and Policy Instruments for Open Market Operations,”IMF Working Paper 95/146 (Washington: International Monetary Fund).
BaliñoTomásJ.T.J.Dhawan and V.Sundararajan1994“Payments System Reforms and Monetary Policy in Emerging Market Economies in Central and Eastern Europe,”Staff PapersInternational Monetary FundVol. 41 (September) pp. 383–410.
Board of Governors of the Federal Reserve System1996Federal Reserve Bulletin (Washington).
DattelsPeter1995“The Microstructure of Government Securities Markets,”IMF Working Paper 95/117 (Washington: International Monetary Fund).
JohnstonR.Barry and Odd PerBrekk1989“Monetary Control Procedures and Financial Reform: Approaches, Issues, and Recent Experiences in Developing Countries,”IMF Working Paper 89/48 (Washington: International Monetary Fund).
QuintynMarc1994“Government Securities Versus Central Bank Securities in Developing Open Market Operations: Evaluation and Need for Coordinating Arrangements,”IMF Working Paper 94/62 (Washington: International Monetary Fund).
Note: The author wishes to acknowledge the helpful comments provided by Tomas J.T. Baliño and his colleagues in the Monetary and Exchange Affairs Department, and by the IMF’s area departments, as well as in discussions with Mitra Farahbaksh and Gonzalo Caprirolo.
The 15 countries are Argentina, Brazil, China, Czech Republic, Egypt, Ghana, India, Indonesia, Malaysia, Mexico, the Philippines, Poland, Russia, Thailand, and Tunisia.
For a discussion of the implications for IMF-supported programs of adopting indirect instruments of control, see Alexander and others (1995).
Open market operations could still be used to encourage expansion with an open discount window by driving the money market rate below the discount rate.
See Chapter 2 in this volume for a detailed discussion of the purpose and design of reserve requirements.
Another approach would be to pay banks a market interest rate on required reserves, which virtually eliminates the tax effect. However, commercial banks would still be restricted to some extent in their ability to invest funds based on their own assessment of market conditions and risks. Payment of interest on reserves would also of course reduce central bank revenues commensurately, as well as governmental revenues to some degree, depending on the difference between the government’s effective marginal tax rate on the central bank and the marginal tax on commercial banks.
However, it is possible to protect investors against inflation by issuing such obligations as indexed or variable interest rate bonds. Brazil, for example, succeeded in developing a market for securities in a context of high inflation.
In this context, the word "private" is used for simplicity to connote institutions whose income and cash flow depend on the sale of their services, whether the institution is state or privately owned.
To maximize flexibility at all times and particularly if faced with potential systemic crises, major central banks generally try to maintain a highly liquid portfolio of government securities. In the United States, for example, almost 86 percent of the central bank’s assets consisted of outright holdings of treasury securities as of January 31, 1996. Of these, almost 30 percent were highly liquid, maturing in three months or less, and about three-fifths matured in one year or less. See Board of Governors of the Federal Reserve System (1996).
The hotly debated issues about whether a central bank should be the sole regulator of the banking system in all its aspects, share such responsibilities, or not be involved will not be discussed in this paper.
The role of the monetary authority and other agencies of government in helping to develop markets is discussed in some detail in Dattels (1995).
See Chapter 3 in this volume for a discussion of refinance instruments in industrial countries.
Repurchase agreements are in effect collateralized loans. But, depending on a country’s legal system, they may provide more protection for lenders. If the borrower goes into bankruptcy, the lender under a repurchase agreement may be able to sell the security and realize the proceeds immediately. If the transaction were instead in the legal form of a collateralized loan, the lender would be delayed in receiving the proceeds for some time while the rights of the various lenders were being adjudicated.
Presumably, they could be revealed to other regulatory arms of the government when there is cause, although the conditions may be a negotiable point in the data-gathering process.
For a discussion of the relationship between reforms of the payment system and the development of money markets and of the monetary policy instruments used by central banks, see Baliño and others (1994).
The discussion here focuses on bank reserves rather than on the monetary base (basically bank reserves plus currency in circulation) as an operating target. Whatever its merits as a longer-run guide, as a day-to-day operating target such use of the monetary base risks introducing into markets a high degree of interest rate volatility. The reason is that currency essentially has a 100 percent reserve requirement while deposits bear only a fractional requirement, if any. Thus, any errors in estimating short-term variations in currency demand would have multiple effects on the deposit component of the nation’s money supply, leading to an undesirably large and volatile excess or deficiency of money relative to economic conditions. As a result, virtually all central banks tend to accommodate to changes in the public’s demand for currency in the course of their day-to-day operations—treating them for operating purposes as a factor affecting reserves and offsetting the reserve-draining effect when currency in circulation rises and the reserve-supplying effect when it declines.
For a discussion of the use of monetary instruments in developing financial markets, see, for example, Johnston and Odd Per Brekk (1989).
See Quintyn (1994) for a discussion of the use of government securities versus central bank securities in developing countries.
The potential negative real impact on the government’s budgetary position should a bank with a special deposit fail would be no different from the situation in which the central bank instead held such a bank’s commercial paper. In any event, the special deposit would in the end be taken over by another banking institution that may acquire part or all of a failed institution, or the deposit could be returned to the central bank and included in a subsequent reserve-supplying auction of such deposits.
However, banks should be limited in the amount of special deposits they can hold to some relatively small percentage of their total liabilities to ensure that the deposits do not drift toward weaker banks that would not be considered good risks in the interbank loan market.
The discussion here is confined to repurchase agreement transactions for domestic policy purposes, with domestic securities as the underlying collateral. In the absence of sufficient domestic securities, foreign exchange swaps can also be used to affect domestic liquidity as, in effect, a substitute for repurchase agreements. It is another matter, however, if swaps or outright foreign exchange transactions are used to influence the exchange rate. The central bank’s domestic open market operation would then be affected differently depending on whether it was being guided by a bank reserve aggregate or by its net domestic assets. If the latter, it would not "automatically" sterilize the domestic market effect of a foreign exchange operation; if the former, it would.
The buyer of the security in a repurchase agreement transaction would, of course, need to be alert to the greater potential for price variability in longer-term relative to short-term securities in ensuring that an adequate margin was maintained against the chance that the seller would fail to repurchase.
A central bank could also announce an interest rate for repurchase transactions and let the market bid solely for an amount. This would make the central bank’s rate intention very clear. However, it has the disadvantage of tending to leave the determination of the size of reserve supply or absorption to the initiative of the market and also does not encourage the evolution of a competitive process of market interest rate determination.
For a discussion of coordination of domestic public debt, government balances, and monetary management in economies in transition, see Sundararajan and others (1994).