I. Interest Rate Policy Under Credit Controls1
1. Vietnam, like a number of other less developed and transition economies lacking well-developed financial markets, uses direct controls to limit credit expansion.2 Vietnam has added these credit controls to a preexisting system of interest rate controls, leaving the banking system with two administratively-imposed sources of financial distortion. Despite the costs to economic efficiency of these measures, they have been highly effective in helping achieve macroeconomic stability over the past few years.
2. There are a number of issues involved in setting interest rates in the presence of bank-by-bank credit ceilings that do not arise in other contexts. Binding credit ceilings imply that the interest rate on loans will not determine the quantity of credit. Deposit rates, however, will play an important role in determining the quantity of deposits supplied by the public. In this situation, the central bank can play an important role in reducing distortions and enabling the banks to match assets and liabilities, though this is a very difficult task when good financial markets do not exist.
3. Not only do interest rates play a different rote in achieving macroeconomic balance when credit ceilings are in place, but in such less developed economies they may be called upon to play other roles as well, such as to promote financial deepening, bank soundness, and high-quality economic growth. This is a heavy burden to place on a single policy instrument. Nevertheless, policymakers in less developed countries frequently must make do with a more limited number of policy tools, which involves accepting tradeoffs. Several such tradeoffs affect interest rate policy in Vietnam:
- High lending rates that reinforce macroeconomic stability may encourage higher-risk bank lending.
- High deposit rates to encourage financial growth may attract unwanted capital inflows.
- Improving monetary control by raising reserve requirements could widen spreads and retard financial intermediation.
- Paying interest on (or sterilizing) excess reserves to encourage deposit gathering may entail a large fiscal cost.
4. The Vietnamese authorities are making steady progress in developing the tools that would allow them to address each of these concerns independently, such as prudential regulation and securities markets. In the meantime, they have a more limited set of options, and must balance macroeconomic stability and other goals through the use of interest rates. Difficult tradeoffs are involved, but good implementation, including careful consideration of the incentives that are set up for banks and borrowers by the institutional framework for monetary control, can reduce the costs of any policy choice, and help to achieve microeconomic objectives without sacrificing the macroeconomic ones.
B. Credit Controls and Monetary Policy
5. A useful starting point for examining the impact of interest rate and credit controls is to review the justification for introducing them, since the distortions that bank-by-bank credit ceilings can introduce are generally recognized and indirect instruments are widely preferred.3 The rationale for direct controls is that, under certain conditions, they may be a more effective tool for conducting monetary policy. The use of indirect instruments relies on the central bank's ability to guide interest rates and the stock of reserve money to influence the broader monetary aggregates. If the monetary authorities are unable either (i) to accurately control the growth in reserve money or (ii) if the relationship between reserve money and broader monetary aggregates is not itself predictable, then the use of indirect instruments may not be practical.
6. Monetary authorities may have difficulties controlling the growth in reserve money for several reasons. Banks may be able to overdraw their accounts at the central bank. The monetary authorities may have difficulty sterilizing international capital flows. Developing countries with poor payments systems may face large and hard-to-predict movements in float. In many cases, the central bank may be required to provide credit on demand to the government.4 Similarly, there may be different causes for an unstable relationship between reserve money and broad money. Banks may not comply with reserve requirements. Movements in inflation or interest rates may affect money demand through the currency/deposit ratio. When banks are not profit-seeking, as may be the case, for example, when major banks are state-owned, then price signals may not produce the intended changes in bank behavior.
7. These problems reduce the ability of indirect instruments to achieve particular objectives for money and credit, and provide a rationale for administered credit ceilings. Farahbaksh and Sensenbrenner (1996) have discussed ways of designing credit ceilings so as to minimize the distortions they create. But the ceilings by themselves, however well designed, are likely to be inadequate to prevent the emergence of major distortions in the economy. In this environment, interest rate policy must also be deployed in support of the objectives underlying the credit ceilings if these distortions are to be kept in check. The rationale for interest rate controls is based on the premise that some market imperfection is distorting rates. Binding bank-by-bank credit ceilings, as is shown later, can easily lead to wide interest rate spreads and create rents for the banks. Under these circumstances, interest rates are already distorted and intervention in rate setting can be readily defended.
8. In such an environment, even the most well-conceived plans to move toward indirect monetary control may require years to implement efficiently. Khan and Sundararajan (1991) point to the conditions—including a sound and competitive banking system, an active and well-functioning money market, and effective instruments to influence interest rate—associated with successful transitions. Until these conditions can be established, countries must make do with direct instruments. How that can best be done is taken up in a later section. Vietnam's efforts in this area, however, provide a good illustration of some of the issues involved.
C. The Economics of Interest Rate Policy
Banking under credit controls
9. With the introduction of credit controls, the monetary authorities have three policy variables to manage: the interest rate; the quantity of money, driven by the supply of credit; and the stock of reserve money. The interaction of these elements, particularly interest rate policy and bank-by-bank credit controls, can be clearly seen in a standard demand and supply framework. While the outcomes differ from those that would prevail in an unregulated market environment, behavior in the model is still market-driven.
10. We begin by describing the supply of deposits and the demand for loans, which are depicted in Figure 1: first, whether or not the interest rate on deposits is itself administered, the supply of deposits can be expressed as a function of the interest rate:
An increase in the interest rate will increase the supply of deposits, and a decrease will reduce them. The presence of credit ceilings will result in a deposit rate below the unconstrained rate, as banks reduce rates to discourage deposits that cannot be relent.
Second, the demand for loans is a negative function of the interest rate:
11. However, for credit ceilings, ǬL, to be binding requires that they be set at a level no higher than the demand for credit at the administered interest rate:
A competitive loan rate in the presence of ceilings would be higher than the unconstrained rate; at any rate below that, credit rationing occurs.
Credit controls with market interest rates
12. Figure 1 illustrates the situation in which credit controls are imposed without interest rate controls and in the presence of profit-seeking banks. In this case (and, it will be seen, in this case only), supply equals demand:
Figure 1.BANKING WITH CONTROLS: Flexible Interest Rates
However, interest spreads are wide, as the banks capture the rents associated with the scarcity of credit. The width of the spreads will in part be a function of how tightly the credit ceilings bind. The combination of high lending rates and low deposit rates may be seen as undesirable, and inspire interest rate controls.
Credit controls with interest rate controls
13. An administratively determined lending rate may be above or below the market equilibrium rate. Changes in loan rates will not affect the quantity of credit when the credit ceilings bind; in terms of Figure 1, when:
So long as credit ceilings are binding, however,
and credit will be rationed. Within this range, the lending rate has no effect on the expansion of credit. Even though the ceilings may play no role in regulating the amount of credit extended, however, high lending rates may be helpful in maintaining the effectiveness of the credit ceilings. By establishing a norm that minimizes the incentive to evade the ceilings, high rates may contribute to their success.
14. Even with credit ceilings and a fixed lending rate, flexible deposit rates would allow the supply of deposits to match the volume of lending.
With administratively determined deposit rates, however, the supply of deposits is predetermined; the banks must fill any gap between the supply of deposits from customers and their credit ceiling through changes in their reserves. In practice, this will mean rediscounting at the central bank in the case of a deposit shortfall, and the accumulation of excess reserve balances if deposits exceed the credit allocations. Thus, the central bank's net credit to commercial banks absorbs the mismatch:
Figure 2 depicts a situation in which commercial banks are net borrowers from the central bank. With such low deposit rates, central bank refinancing is necessary for the credit ceilings to have operational meaning. Figure 3 shows the case of excess liquidity in the banking system. The main practical difference between these two cases is the level of the deposit interest rate relative to the fixed quantity of credit. If deposit rates were freely adjusted by commercially-driven banks, large swings in deposit rates might be expected, as the full burden of balancing bank assets and liabilities would fall on deposit rates. With controls on deposit rates as well—needed, perhaps, to avoid such volatility, the burden falls to the central bank to provide the facilities to allow the two sides of the balance sheet to balance within the constraints set.
Figure 2.BANKING WITH CONTROLS: The Central Bank Provides Credit Figure 3.Banking with Controls: The Central Bank Absorbs Excess Liquidity
Setting administered interest rates
15. On what basis, then, are administered interest rates to be set? In general, tight monetary policy established in the credit ceilings will have particular implications for interest rate policy options. The tighter the policy, the larger the market-clearing spread. Thus, mimicking the market interest rate outcome would tend to be an inappropriate goal when inflation reduction is a policy priority, but might be more suitably introduced during a period of weak credit demand.5 Similarly, the less credit available, the more pressure there will be to evade the ceilings, and the higher interest rates can rise without limiting banks' ability to ration credit. High deposit rates at the same time would also support a disinflationary policy by discouraging the hoarding of goods as an inflation hedge. As credit policy becomes progressively looser, it would be appropriate to move interest rates downward.
16. Even when credit ceilings do not play the role that interest rates normally play in regulating the amount of credit extended, high lending rates may be helpful in maintaining the effectiveness of the credit ceilings. By establishing a norm that minimizes the incentive to evade the ceilings, high rates may contribute to their success. They also permit relatively high deposit rates—without putting pressure on bank profits—contributing to macroeconomic stability by stimulating demand for interest-bearing deposits. Moreover, in an economy with a history of macro-economic instability, or where currency substitution and low interest rates in the past have severely reduced the size of the banking system in real terms, sustained high real deposit rates—implying high real loan rates as well—may be a necessary condition for remonetization and financial deepening.
Central bank policy under credit controls
17. Figure 2, when the central bank needs to step in to sustain the targeted credit increase, and Figure 3, when the central bank needs to drain liquidity to forestall any spillovers from a buildup of excess reserves, illustrate why central bank action is vital to achieve the goals of macroeconomic policy. In both cases, the central bank must ensure that there is sufficient liquidity to enable banks to reach their credit goals while at the same time ensuring that excess reserves do not undermine the incentive for the banks to attract deposits, particularly if excess reserves are remunerated at a lower rate than retail deposits.
18. Draining liquidity from the banking system often presents a major challenge for developing country central banks, unless they are in a position to market securities or to run down the stock of credit to banks. If not, the central bank must reduce net credit another way. Unless net credit to government can be cut (and seldom will fiscal policy accommodate monetary policy needs in this way), it will be necessary to induce the banks to hold increasing deposits at the central bank. Raising the reserve requirement is the simplest method. Excess reserves are reduced to the desired level at a single stroke and at minimal cost to the authorities. It has the drawback, however, that it requires increasing interest rate spreads for the bank to remain profitable, increasing the implicit tax on banking.
19. An alternative strategy would be to pay interest on banks' deposits at the central bank. Paying interest on all bank deposits at the central bank-including required reserves—will improve the income position of banks and permit spreads to narrow, but if funds at the central bank are to be paid a rate competitive with bank deposits, the interest cost to the central bank can easily become excessive and result in a substantial fiscal or quasi-fiscal burden. A more economical approach, which retains the desired incentive effects of encouraging banks to pursue growth in deposits, is to pay interest only on excess reserves, or on a stock of central bank bills offered in sufficient amounts. Treasury bills could also be issued in this case, as long as the proceeds were placed in blocked Government accounts at the central bank.6
20. Whichever option is chosen, the important goal must be to balance the permitted stock of credit with the supply of deposits in a way that does not stress banks' balance sheets. It may even be possible to combine the two approaches, in order to retain incentives for banks to attract deposits while containing the interest cost of issuing securities.7
Interest rates and bank risk
21. Commercial bank lending and deposit-taking activities are the crucial link between monetary policy instruments and macroeconomic outcomes. Interest rate controls and bank-by-bank credit ceilings are meant to induce behavior on the part of the banks consistent with the macro objectives. However, while banks' behavior at the level of the individual loan is not so readily observable, these same policies are likely to affect banks' behavior, and in particularly the amount of risk in their portfolios, in important and generally predictable ways. 8
22. While low real interest rates may tend to increase the pressure on the credit ceilings from the larger pool of borrowers denied credit, there may be value to reducing rates to reduce the riskiness of credit portfolios. Problems of moral hazard and adverse selection in banking as a result of high real interest rates are well-recognized, and it is clear that some credit rationing may be desirable to improve the overall quality of bank lending (McKinnon (1991) discusses the macroeconomic implications of moral hazard problems). Problems of moral hazard on the part of banks—that they may deliberately lend to high-return, high-risk projects-are largely overcome when the positive returns to the bank are equivalent for all projects due to interest rate ceilings.9 The principal exception is when lending to related parties is pervasive. The most effective means of discouraging such “insider” loans is to maintain strongly positive real lending rates, to reduce the benefits of preferential access to credit.10
23. Adverse selection among borrowers, however, can be an even more serious problem, especially where there is no history of screening loans for credit quality: banks will lack the skills—and reliable accounting data—to effectively evaluate borrowers. Indeed, many economic models that discuss adverse selection assume that borrowers are “observationally equivalent”—that the bank is unable to distinguish between different types of borrowers. This assumption holds even more strongly than in the developed economies for which, by and large, they were conceived.
24. The interest rate at which adverse selection will become a potential problem is difficult to identify, but will be affected by the profitability of borrower projects at the micro level. This has two components. The first is the rate of growth of the economy: a fast-growing economy will tend to have a higher-return set of projects (a rule of thumb that recognizes this fact suggests the real interest rate should be equal to the rate of growth). The second component is the overall supply of credit in the economy. With low levels of credit, the economy is likely to be able (other things equal) to support higher real interest rates, as only the most profitable activities can receive credit anyway.11 In such circumstances, the health of the banks may be greatly supported by a policy that avoids very high real interest rates. Low-risk, low-return projects will then be less likely to ration themselves out of the borrowers pool. In addition, since all loans will be easier to repay at lower rates, this can be expected to contribute to a higher average loan quality as well.
D. Interest Rate Policy in Vietnam
25. In Vietnam, trust in the banking system is low following years of very high inflation and negative real interest rates. A confiscatory currency reform in 1985 and extensive depositor losses in a wave of credit cooperative failures in 1989-91 further undermined public trust. Unofficial estimates for currency substitution suggest that Vietnamese are holding several billion dollars in gold and foreign currency.12 As a result, financial intermediation remains quite limited by international standards, and much of it is in foreign currency. While some measures have shown improvement (Chart 1), bank credit to the economy is the equivalent of only 17 percent of GDP, compared with ratios of around 25 percent in India and the Philippines, 50 percent in Indonesia, 90 percent in China, and 100 percent in Thailand.
Chart 1Vietnam: Monetization, 1988-1996
26. Although Vietnam's doi moi economic reform policy has liberalized almost all prices in the economy, continued administrative control over interest rates has been an important exception, and one which has played a central role in Vietnam's successful stabilization. These controls have been effectively employed to maintain deposit rates at high real levels; deposit rates at times even exceeded loan rates (Chart 2). This has served to bolster the demand for domestic money in the face of exchange rate uncertainty, currency substitution, and deeply-held suspicion of the banking system. As macroeconomic stabilization proceeded, these high deposit rates brought large increases in local currency bank deposits, which doubled in the two years to end-1995.
Chart 2Real Interest Rates
27. At the same time that high deposit interest rates were bolstering the demand for domestic money, credit restraint was being effectively used to moderate growth in the supply of money. Since 1994, Vietnam has accomplished this by the use of bank-by-bank credit ceilings. For most of this time, these ceilings have bound very tightly, and they have played a major role in reducing inflation.
28. This policy combination resulted in more rapid growth in deposits than in loans, leaving the banks to accumulate excess liquidity. This is just the situation depicted in Figure 3 in section I.C. The State Bank of Vietnam (SBV) has worked to minimizing the negative effects of this mismatch, which the earlier section suggested is the preferred policy stance in this situation. While lacking the tools to drain excess liquidity through open market operations, it did raise interest rates on a large portfolio of outstanding credit to the state-owned banks which had previously been at concessional interest rates. This encouraged their repayment, steadily draining excess liquidity from the banking system. Thus, gross credit to banks from the SBV declined by over 12 percent during 1995 (Table 1). This approach has limitations, however, as much of the remaining credit has been relent on a long-term basis for officially-sponsored projects. As a result, the reduction in credit has not been able to keep pace with the accumulation of liquidity by the banks. Excess reserves have continued to rise.
|(in trillions of dong)|
|SBV credit to banks||7.8||7.3||6.5||7.3||6.8||6.7||6.5|
|less: Bankers' deposits with SBV||4.4||4.8||3.9||5.5||6.3||7.5||7.8|
|of which: required reserves||1.7||1.9||2.3||2.4||…||…||…|
|less: SBV bills outstanding||0.0||0.0||0.0||0.0||0.4||0.0||0.0|
|Net credit to banks||3.4||2.5||2.6||1.8||0.1||-0.8||-1.3|
|Memorandum item: reserve money||23.9||24.2||23.0||24.0||26.8||28.9||29.1|
29. Faced with this difficult situation, the State Bank began to develop its capacity for monetary operations. Efforts to sterilize by issuing central bank bills at the beginning of 1996 were quickly abandoned, however, because draining the amounts necessary required paying higher interest rates than the authorities were willing to offer: the bills were offered at a rate of 9 percent, and only half of the bills offered in the two auctions were taken up. The average maturity was only two months, but the State Bank saw the interest cost, which came to the equivalent of about $2 million, as too high to pursue this avenue further. In addition, the State Treasury opposed the SBV bill issues because they competed with treasury bill issues. Thus, despite improvements in the technical capability of the State Bank, other factors undercut the will to use monetary operations to reduce excess liquidity. This excess liquidity is an important barrier to shifting to indirect monetary control in Vietnam, as it cannot yet be sterilized and the authorities have been reluctant to eliminate it through an increase in the reserve requirement.
30. Offsetting excess reserves is particularly difficult in Vietnam because the sources of the build-up are continuing to boost bank reserves. First, foreign exchange reserves have increased rapidly over the past two years. Moreover, high interest rates attracted funds from abroad despite capital controls on portfolio investment. Although the treasury bill market was closed to foreign institutions, and there were restrictions on interest-bearing bank accounts on foreigners, some foreign funds may have found their way into the banking system. Also, there has been pressure from borrowers to extend credit in foreign currency in order to take advantage of the interest rate differential between dong and dollar loans, as the exchange rate has been stable for years. While the evidence suggests that banks are not running large open foreign currency positions in order to accommodate this demand for foreign currency credit, the demand for foreign commercial borrowing is great. Foreign borrowing is kept under control by the State Bank only with great difficulty.
31. The second source of growth in bank reserves has been a shift in, money holdings from currency to deposits as a result of the high interest rate policy. The currency/deposit ratio has declined from 1.3 at the beginning of 1995 to 0.8 in mid-1996. The demand for bank reserves has thus grown much more slowly than broad money, though the supply of reserves has kept pace: the money multiplier remained virtually unchanged over the same period. As a result, the banks have consistently held large excess reserve balances over the past couple of years.
32. As inflation fell in the second half of 1995, the authorities began moving interest rates downward. The State Bank set out to maintain real interest rates (Chart 3) at least as high as some of the higher rates seen in fast-growing neighboring countries (Table 2). However, the excess liquidity made it difficult to maintain a reasonable spread. With a 10 percent reserve requirement and high demand for transactions balances on the part of banks on top of the excess liquidity, banks were forced to maintain wide spreads simply to cover the cost of idle funds. In addition, since the interest rate reductions were widely anticipated, many depositors moved to longer-term time deposits in the closing months of 1995, saddling banks with high interest costs unmatched on the rapid-turnover loan portfolio. Through most of 1996, the spread between the 3-month loan rate and the 3-month deposit rate was close to 10 percent. This wide spread continued in spite of a cap on net interest income of the banks that was imposed at the beginning of 1996 in order to limit spreads.
|(percent per annum; annual averages)|
|spread||4.4||5.6||7.0||9.7||6.3||5.1||4.6||3.2||4.6||4.3||5.2||5.1||4.5||6.3||5.4|Chart 3.Nominal and Real Interest Rates
33. In these circumstances, the tradeoff between lowering lending rates and lowering deposit rates is stark: lower deposit rates could certainly solve the problem of excess liquidity and unwanted capital inflows. Such a solution would, however, entail great risks. Over the past five years, the share (though not the amount) of foreign currency deposits has been brought down from 60 to 25 percent of total deposits, and the monetization of the economy has been increased. These gains are still fragile, however, and could easily be reversed if rates declined too far. Yet maintaining high lending rates indefinitely, even as inflation falls to low single digits, poses risks to the banks of unknown magnitude.
34. The true condition of the Vietnamese banking system is difficult to discern. The modest share of bad debts reported by the banks largely dates to the early 1990s when many loss-making state enterprises were closed. Only one of the banks has been subject to an audit meeting international standards, but the bank's internal accounting was not adequate to allow a clear picture to emerge. The four state commercial banks that dominate the Vietnamese banking system are moving rapidly to overhaul their internal procedures to bring them up to international standards, despite continued vulnerability to political pressure. With economic growth of over 8 percent per year over the past five years, many loans of dubious quality may in fact be performing well. However, the banking system may be vulnerable to a slowdown in economic growth. Before that occurs, it would be important to establish the best possible incentives for quality lending.
35. There are a number of factors that must be taken into account in setting interest rate policy in a country with underdeveloped financial markets that do not arise in more developed markets:
- In the presence of credit controls, interest rates will be distorted such that interest rate controls may be needed as a complementary measure.
- Central bank policy in such circumstances must be sensitive to the effects of these controls on the balance sheets of commercial banks, and work to maintain a consistent set of incentives in the banking system.
- There is an inevitable tension, given the dearth of policy instruments in such a setting, between keeping interest rates high to support the credit ceilings and lowering them to improve the incentives for sound banking.
36. In each of these areas, there are real tradeoffs that must be faced. In Vietnam, as macroeconomic stabilization has taken hold, the inconsistencies of an administratively-controlled financial system are becoming more evident. This points to the importance of moving quickly to develop additional policy tools so that interest rate policy can be freed from the tradeoffs now being faced, allowing different objectives policy objectives to be pursued independently.
AlexanderWilliam ETomas J.T.Balino and CharlesEnochThe Adoption of Indirect Instruments of Monetary Policy,IMF Occasional Paper #126June1995.
AxilrodStephen H. “Transformation of Markets and Policy Instruments for Open Market Operations” IMF Working Paper WP/95/16December1995.
BisatAmerR. BarryJohnston and V.Sundararajan “Issues in Managing and Sequencing Financial Sector Reforms: Lessons from Experiences in Five Developing Countries” IMF Working Paper WP/92/82October1992.
CamardWayne “Business Groups and Structural Adjustment” U.S. Agency for International Development mimeoJune1991.
CaprioGerard P. Jr. “Bank Insolvencies: Cross-Country Experience” World Bank Policy Research Working Paper #1620July1996.
DodsworthJohn R.et al., Vietnam: Transition to a Market Economy,IMF Occasional Paper #135March1996.
Dodsworth John R.AjaiChopraChi D.Pham and HisanobuShishidoEconomic Reform in Indochina IMF Working Paper WP/96/112October1996.
FarahbakshMitra and GabrielSensenbrenner “Bank-by-Bank Credit Ceilings: Issues and Experiences” IMF Working Paper WP/96/63June1996.
Galbis Vicente “Financial Sector Reform in Eight Countries: Issues and Results” IMF Working Paper WP/95/141December1995.
JohnstonR. Barry “The Speed of Financial Sector Reform: Risks and Strategies” IMF Paper on Policy Analysis and Assessment PPAA/94/26December1994.
KhanMohsin S. and V.Sundararajan “Financial Sector Reforms and Monetary Policy” IMF Working Paper WP/91/127December1991.