Financial reform will be discussed in this paper in relation to bank regulation, or, better still, banking supervision. First, I shall define the subject. The purpose of financial policy is to stipulate conditions for the highest economic growth, lowest inflation, and lowest unemployment through the most efficient allocation of financial resources. In view of this, the purpose of financial reform is to enable the financial sector to operate along the lines described in cases where it does not. Liberalization is the key word.
Banking regulation—or supervision—is to be seen here from its prudential aspect, that is, rules and techniques that aim to protect the depositors through the protection of the soundness of the financial sector, or what is generally described as its solvency, liquidity, and profitability. Why do we associate the question of financial reforms (liberalization) with the idea of protection of depositors? We do so because banking supervision does not deal only with protection of depositors; more widely, it relates to the well functioning of the financial system and the well functioning of the financial system is indispensable to the success of financial reforms.
I would like to present the subject from the point of view of a bank supervisor, which is what I am, that is, from a pragmatic point of view. We have observed a great number of financial reforms (the transition from credit policy to more indirect tools) in various developing and industrial countries. The reforms have not always been successful and have often ended with inflation, bank crisis, or both. I shall break the subject into two parts: the lack of proper banking supervision, which, in many a case, was at the root of the lack of success, or at least one of the reasons, and a review of how to boost banking supervision to support financial reform, especially in the transition from a centrally planned economy to a market-oriented economy.
Financial Reform
The main objective of financial reform is to enhance intermediation. It should attract more deposits and remove excessive regulation on interest rates (often coupled with unduly heavy taxation) so as to offer positive interest rates to depositors; savers rely on unproductive assets or capital flight if no asset is available that can be expected to retain its value over time. It should also let lending institutions themselves select those to whom they lend, using the criterion of the highest return (taking into account the degree of riskiness attached to any lending operations). This means removal of credit ceilings and priority sectors—subsidies, if any, to be granted from budgetary resources—as well as liberalization of lending rates, plus increased competition.
In short, the object of financial reform is the removal of what is called financial repression. Financial repression is defined as policies that distort domestic capital markets through a variety of measures; consequently, the development of the financial system, savings, and investment suffer generally.
Comparison Across Countries
Among the many countries that implemented financial liberalization policies during the 1970s and 1980s, several were unable to achieve their goals; instead liberalization left them with a sharp increase in real interest rates, an outburst of inflation, a drop in output owing to bankrupt firms (both financial and nonfinancial), and a widened external imbalance. Argentina, Chile, the Philippines, Turkey, and Uruguay are the countries mentioned in this regard; whereas Korea, Malaysia, and Sri Lanka had successful reforms.
The pace at which reforms were introduced, despite economic instability or inadequate bank supervision at the time, affected each country’s achievement. Argentina, Chile, and Uruguay had severe macroeconomic imbalances when interest rate reform and financial liberalization policies were implemented. Ratios of growth of output, savings, and investment were all low; inflation rates were high; and the external current account deficits were large in relation to national income. The strategy in these three countries involved the entire and abrupt removal of interest rate ceilings and credit controls, and the relaxation of government supervision over the banking system, combined with virtually free deposit insurance, explicit or implicit. This combined strategy is regarded as especially important.
Much of the same pattern of events occurred within a relatively short time in the Philippines, Turkey, and Malaysia. In Malaysia, however, where the macroeconomic environment was stable and banking supervision was adequate, the reform took place successfully.
The reforms were also a success in countries where they were implemented more progressively, for example, in Korea (1981) and Sri Lanka (1977). In these countries, positive real interest rates were achieved and maintained through credible macroeconomic policies that successfully reduced inflation to low levels. At the same time, the efficiency of banking supervision was improved, while incremental adjustments were made in regulated nominal interest rates in order to maintain a positive real level. Positive rates stimulated deposits that, in turn, increased the amount of credit available to productive firms. Full liberalization of interest rates was implemented only when economic stability was firmly established and a permanently effective system of banking supervision was in place and enforced.
Implementing reforms over time is not sufficient if, during the period, the economy is not stabilized and banking supervision is not maintained at the required level of efficiency. For example, in Indonesia, complete liberalization of interest rates was achieved, although the measures implemented to stabilize the economy had not reached their goal; the result was inflationary pressure and destabilizing capital flows, which ended in volatile interest rates and an unstable banking system.
From the above, it appears that two components are needed for a financial reform to succeed: macroeconomic stability and effective banking regulation and supervision. In cases where one or both of these conditions are not met, there is need to improve the situation before the reform is implemented, but this cannot be done overnight, hence the need for progressive implementation. I will not elaborate on the macroeconomic aspect, which is discussed in other papers and illustrated by specific country cases. I will concentrate on the question of banking supervision. I will first explain why the lack of proper supervision may result in a failure of the financial reforms.
Link Between Lack of Proper Supervision and Failure of the Financial Reforms
Why does the lack of proper banking regulation and supervision impede the successful achievement of financial reforms?
What has generally been noticed as a reaction to the liberalization of interest rates is an increase in the lending rates coupled with a widening of bank margins. This is normal and expected, as controlled lending rates were usually set too low. However, it is normally expected that the trend toward an increase will be limited by (1) increased competition, which is generally favored as part of the reform, and (2) a reduction in reserve requirements, which is also often part of the financial reform. It has, nevertheless, occurred that the increase in lending rates reaches—in real terms—high levels, that is, levels that by themselves trigger the problem.
At this stage, a short theoretical explanation is necessary. When lending interest rates have been liberalized, banks, as they are looking for the highest return, will be induced to increase their lending rate up to the point where supply and demand for loans will meet. This point may be defined as the market-clearing equilibrium interest rate. This is the point at which the maximum number of potential borrowers can be satisfied. If lending rates were increased beyond this point, the number of applicant borrowers would decline. On the contrary, if lending rates were reduced below the market-clearing level, there would be potential borrowers who are prepared to pay more and who are not accommodated (as was the case prior to the liberalization).
Well-managed banks will not behave in this way. Wise bankers will realize that, beyond a certain level that is lower than the market-clearing level, any increase in the lending rate is counterproductive. It is counterproductive for two reasons: (1) the most creditworthy borrowers are discouraged and are likely to drop out of the market, and (2) other borrowers, those who are less creditworthy, are induced to choose the projects that are associated with the highest expected profits in order to compensate for the high level of interest they have to pay, but these projects are also the riskiest ones, that is, those that give rise to the highest probability of default.
In short, because wise bankers take into account the degree of riskiness of loans, that is, the repayment probability, they apply a lending interest rate at which the demand for credit is greater than the supply. This means that bankers are in a position to screen the applications for loans so as to retain those which appear the less risky. As a result, the losses on loans at this interest rate are expected to be limited, which means that the final return will be higher than the return obtained at the market-clearing level once the losses are deducted from the profits.
Several conclusions can be drawn from the preceding discussion. First, as the risk of default on losses increases when macroeconomic instability is a problem, one may expect that well-managed banks will fix their lending rate at a level lower than that during economic stability. Second, not all bankers are wise and prudent bankers. There is a risk that banks do not realize that fixing lending rates at the market-clearing level does not produce the highest return. This means that there is a risk that banks fix their lending rate at a level that will increase the likelihood of loan defaults; this risk is especially high if the macroeconomic environment is unstable.
Finally, from the experience of several countries, it would also appear that this risk is particularly high if bank deposits are guaranteed by a system of deposit insurance (be it implicit or explicit). Simply put, banks’ managements are not averse to making risky loans or investments if they think there is no risk to their depositors. I prefer a more elaborate explanation. Depositors are generally not aware of the respective degree of soundness of individual banks. Even if they are informed, they are not induced to shift their funds from one bank to another if they feel that the deposit insurance scheme provides them with adequate protection. As a result, management does not feel any pressure to limit the riskiness of their loan portfolio for fear of a drain on liquidity caused by withdrawals of deposits. This is what is called banks’ moral hazard, that is, the tendency to provide risky loans at high interest rates in the expectation that large losses will be covered by deposit insurance. Moral hazard is even higher when not only depositors but also shareholders are guaranteed against any loss.
When unwise bank management is coupled with moral hazard, especially in an unstable macroeconomic environment, a sharp increase in interest rates is likely, further to a sudden liberalization of these rates. The sharp increase may, in turn, make the demand for credit inelastic. This means that more and more firms are unable to pay the interest on their borrowings. Banks are forced to capitalize interest; the additional facilities themselves are charged the high level of interest that is in effect; many firms exhaust their capacity to borrow and nonperforming loans carried by banks begin to grow rapidly. The financial reform has been a failure.
Let us now turn to the role of banking supervision. In this context, we have said that not all bankers are prudent and wise. This is true. That they behave in a prudent fashion is due to an effective system of prudential supervision. Especially important in this regard is what pertains to loans classification and provisioning, interest in suspense, concentration of loans and advances, and connected lending. Countries in which financial reform was a failure have had an inadequate system of banking supervision in addition to moral hazard and an unstable economy. Banking supervision was often confined to mere control of compliance with credit policy and foreign exchange regulations; it had little to do with real prudential supervision. It is true that in Argentina and Chile, banking regulation was strengthened during the reform, but some critical rules were not included in the additions that were made and, more important, implementation of the regulation was weak. It is reported that, in Argentina, for instance, the frequency of on-site inspections declined following the reforms.
If it is now understandable why lack of proper supervision can prevent the success of financial reforms, the case is even much clearer in a situation where a sizable fraction of financial institutions are technically insolvent (sometimes these institutions have lost their capital several times over). This is present in both public and private banks, but more common in public ones. Undue concentration of risks, pressure from the government, and connected lending have been common features. This has two major policy implications: (1) By rolling over the loans made to weak firms, banks are unable to shift lending priorities to new activities and investments; and (2) Higher interest rates on deposits will impede cash flow, as interest on loans is just rolled over. In this case, the reform reveals the real situation of the banking system, as it hastens the emergence of a liquidity crisis and it transforms a bank in distress (an institution technically insolvent that goes on operating as if everything were in order) into an element of a real banking crisis.
Boosting Banking Supervision to Support Financial Reform
From the above, it would appear that before finalizing financial sector reform, especially before freeing interest rates completely, there is a need to enhance the efficiency of the prudential banking supervision system, if there is one, or to set up such a system, if there is none. This may not be sufficient for the financial reform to be a success; however, it appears to be a prerequisite.
Let us suppose that the Central Banking Department of the International Monetary Fund has been requested to provide advice on how to implement enhanced banking supervision. This may pertain either to a developing country or an Eastern European country that plans to shift from a centrally planned economy to a market-oriented one.
The first step would be to assess the present supervisory capability of the country and to get at least an impression of what the banking system comprises. From one country to another, the picture can be completely different. Here, I will discuss mainly the case of centrally planned economies shifting to a market-oriented economy, that is, a situation where banks are often in distress, or possibly where there are no real commercial banks at all, and practically no banking supervision.
Now, if we assign experts or consultants to advise the authorities, what do we expect from them? At this stage, the next step would be for them, and for the authorities, to acquire a clearer knowledge of the situation of the banking system. Prudential supervision is not a simple set of rules and regulations that can be applied everywhere and at all times in the same fashion. There are permanent principles and constant features, but they have to be adapted to the existing context. It is of the utmost importance to take into account (1) the major features of the legal framework (company law, rules—and practice—regarding collateral, bankruptcy, etc.); (2) the quality of the communication network; (3) the actual nature of the operations carried out by the banks; (4) the nature and the degree of stability of banks resources and the apparent value of their assets; needless to say, the latter is critical; (5) the accounting organization and techniques, including a detailed review of the balance sheets and proflt-and-loss accounts; and (6) the organization chart of banks, their staffing, including information on how well trained is the staff.
The review must be conducted mainly through numerous meetings and discussions held with bank representatives of various levels. These meetings should preferably be held on the banks’ premises, where it is possible to consult their books and files in order to acquire a concrete understanding of their organization and methods.
The review will frequently reveal—or simply confirm—a situation where one or several banks are in distress; a bank in distress usually keeps its real position hidden. If a significant share of the banking system is in distress, no monetary policy can be carried out. Banks in distress are unable to react to market signals and will need injections of liquidity that may be contrary to what the central bank is trying to achieve through its monetary policy. It is also an impossible task to set up an interbank money market in a situation where banks are not prepared to lend money to each other because they are aware of their respective weaknesses.
The task of consultants of the Central Banking Department at this stage is to draw the authorities’ attention to three important facts that are often ignored—not always in good faith—(1) prudential supervision cannot resuscitate a dead body and it is useless for banks in distress; (2) someone will have to foot the bill (central bank, depositors, taxpayers); the question is to decide who will bear the cost; and (3) the later the decision is made the higher the bill (high amount of interest paid versus little interest received, prospects for collection of debt vanish as time elapses). It is essential that a thorough audit of problem banks be conducted as soon as possible by a competent accounting firm if the situation is not clear, and it is essential that steps are taken in order to remove the bad assets—those that do not appear collectible—from the banks’ balance sheets. This will lead to mergers, divisions, liquidations, recapitalization, in short what is usually called “restructuring” or “rehabilitation” of the banking system. The remaining banks will be the viable ones. Needless to say, this is said more easily than done. The whole process generally requires several years. The major difficulty is to properly assess the value of a loan portfolio and to make an estimate of what will be recovered. Not surprisingly, in most cases the first assessments were too optimistic. Nonetheless, the rehabilitation process should be started as soon as possible and completed as quickly as possible. Any monetary policy instruments that are being developed at this stage should be used to absorb excess liquidity generated by the recapitalization measures.
Prudential supervisors should deal with the remaining banks, which are regarded as viable and which are expected to react to monetary policy decisions in the manner desired by the authorities. These are the banks that will permit an increased financial intermediation following the financial liberalization process.
At the end of the initial review, central bank staff and experts from the Central Banking Department have acquired an overall knowledge of the banks’ operations and of the existing reporting system (if any). Keeping in mind the nature of the information with which supervisors should be provided in order for them to carry out their supervisory function, their next step is to design a set of reporting forms through which this information will be collected. In practice, the forms may consist of a statement of assets and liabilities completed by a certain number of appendices. These periodic reports are designed to enable central bank staff to monitor the development of banks’ activities, mainly in order to discover as soon as possible any unfavorable trend that would require corrective action. In due course, as soon as various types of rules and regulations are introduced, the reports will also be used to monitor the compliance with the said rules and regulations.
The reporting forms need not be very sophisticated; they have to be adapted to the actual nature of bank operations and will be developed to include more and more detailed information when banks’ operations and the types of risks incurred become more diversified. Another possibility is to require a certain number of data, but to allow some flexibility, during a transition period, for those banks that are not yet in a position to produce the said data.
As it is important that all banks submit their data to the authorities in a similar way for comparison, a common accounting plan for banks is useful. Such a plan may be designed, if this is technically feasible, before the reporting forms. In this case, forms are based on the accounting plan. As the latter defines in detail how to enter the various operations in the books, it is easy to obtain forms that are transparent. Depending on local circ*mstances, the design of an accounting plan may require a long time; two years would not be exceptional; in addition banks must have time to adapt their internal organization to the common plan. Consequently, a common accounting plan is sometimes not to be regarded as a prerequisite for the design of the reporting forms. In the absence of a common plan of accounts, the authorities will have to describe precisely and in detail how to fill out the forms. It is true that the accuracy of the returns will not be foolproof, but it is also true that this is never achieved. Experience worldwide shows that—with or without a common accounting plan—errors in reporting forms, sometimes significant ones, are common. What is most important is that supervisors be vigilant for accuracy, so as to ascertain that banks are improving the quality of the returns when there is a need to do so.
In truth of fact, it must also be recognized that banks require some time (generally several months) to adapt their internal organization to the common plan. What creates difficulty is when banks must break down the balance of some of their accounts in order to comply with the plan of accounts (or the return forms). Then, it is often difficult not to tolerate some delay in the complete implementation of the plan. In the meantime, banks should be authorized to either leave the line blank or base the reported information on a statistical approach or on pure estimates. As a result, whether or not the accounting plan is implemented very rapidly (after a few months) or only after two or three years does not make a great difference.
It is indispensable for supervisors to be closely involved in the design of the accounting plan so that the various risks faced by banks can be easily identified and quantified in the books. This should be regarded as indispensable not only for proper prudential supervision but also for proper management. The information recorded in the books and reported on the returns should normally include the following, which is generally not available in developing and Eastern European countries before the implementation of the reform:
(1) Classification of loans and advances, depending on the repayment probability; interest in suspense; provisions for bad debts. In countries that do not have sound experience in banking supervision, it is regarded as preferable to decide on the degree of riskiness of loans and advances, not on a case-by-case basis, but in view of objective criteria, for instance, on the length of time a payment has been in arrears.
(2) Off-balance sheet items, which are designed to record all the commitments underwritten by a bank in the form of a promise to lend, extension of guarantees, and other obligations. Such commitments involve a risk for the bank in spite of the fact that no disbursem*nt has yet been made. Guarantees and commitments extended by other parties in favor of the bank also have to be recorded.
(3) Regularization accounts, which attach incomes and expenses paid in advance or to be paid at a later stage to the year to which they are related, whatever the date of the actual payment.
(4) Detailed recording of income and expenses, in order to permit a precise analysis of the profitability of a bank.
(5) Assets and liabilities, as well as off-balance sheet items, in foreign currency to be accounted for in a way that enables banks to determine their profit and loss in foreign currencies, as often as they want, and to know at all times the extent of their foreign exchange risk. The concept is clear but its implementation requires some technical adjustments in accounting procedures, especially in electronic data processing, which may need some time.
(6) The same remark applies to the breakdown of major assets and liabilities on the basis of their remaining life in order to permit the computation of a bank’s liquidity position.
At the same time as new reporting forms are designed, a set of prudential rules needs to be introduced or the existing set revised, if one exists. These two different tasks should be done concurrently and within the framework of the same approach, that is, (1) taking into account the actual situation of banks and the overall economic, legal, and psychological environment and (2) being in close contact with the banks’ representatives. Knowing that prudential rules are rules that any wise banker would adopt spontaneously, unlike credit policy or foreign control regulations, better compliance may be expected if banks are convinced that the proposed regulation is not against them but in their favor.
At this early stage, and especially for countries that have no experience at all in prudential supervision, this is not a complete and sophisticated set of regulations that can be envisaged. The various parties involved need to become familiar with new practices, step by step. It must be verified that some rudimentary rules are (1) applicable and (2) by and large correctly applied, before issuing rules that would be more detailed and more demanding.
The first step, which is of utmost importance, is to remind banks of some basic principles recognized universally in the banking industry but which are not yet familiar in the country. This should include the requirement to conduct a formal appraisal of credit applications and to make the board of directors of each institution define who has the responsibility to make decisions on such applications, depending on the size of the loans. It is also essential to define rules that will prevent excessive concentration of loans as well as connected lending. Another indispensable requirement is the careful monitoring of loans and advances (in particular overdrawn current accounts). Any arrears in the repayment of a loan or in the payment of interest should be recorded immediately so that the decisions to set aside provisions for bad debts and to keep interests in suspense (instead of including them in the income) are made in due time and with a realistic view of the probability of repayment. In addition, it is essential to convince the fiscal authorities that provisions for bad debts should be deducted from the profit before tax.
It is also necessary to require at this early stage that banks set up proper internal control measures and internal audit policies. Their attention must be drawn to the importance of having the capacity to discover errors made by their staff before they become a catastrophe. Banks should be given some time to develop a minimum level of control before entire liberalization is decided.
There is nothing against introducing a capital adequacy ratio and, for instance, a liquidity ratio, but these should be provisional ratios, introduced on a tentative basis, and the design should be as simple as possible. A capital adequacy ratio at this stage may complement stabilization policies by limiting the expansion of assets by banks. This may be useful in the framework of financial liberalization, as experience shows that the response of credit growth is generally more rapid than the response of deposit growth. Only at a later stage will it be possible to introduce a risk-weighted capital ratio. This will be possible only when the exact situation of banks is well known and when banks are technically able to carry out the computation of a sophisticated ratio. One of the requirements that should be formulated from the outset is the fact that, for the computation of any capital adequacy ratio, the shortfall in provisions for bad debts, if any, must be deducted from the amount of capital as it appears in the books. This is indispensable if a realistic view of the situation of a bank is to be had; my personal experience has shown me that this is often forgotten and that this may lead to very damaging consequences.
As far as the surveillance of liquidity is concerned, it will generally be impossible for banks to compute from the outset a ratio based on the comparison of the remaining life of assets and liabilities. This will be impossible for technical reasons. As a result, if a liquidity ratio is introduced at this stage, it should be a ratio following what is called the “stock approach,” that is, comparison of liquid or easily liquefiable assets compared with the global amount of liabilities or only with some types of liability. Needless to say, however, banks should be free to choose the type of liquid assets that they want to keep. In this regard, a liquidity ratio designed for prudential purposes is different from a rule that would prescribe a minimum investment in a particular type of asset, for example, treasury bills.
Limits on foreign exchange exposures may also be issued but, here again, only once the actual position of banks is well known. Issuing a regulation that would be inapplicable would be counterproductive. It must also be kept in mind that foreign exchange control will remain in effect for some time and this tends to limit the magnitude of the foreign exchange risk of banks (at least on the asset side).
The same applies to interest rate risk. The quantification of the interest rate risks—if one wants to take the overall aspect of this risk into consideration—involves such a degree of technicality and sophistication that it cannot be envisaged in the first step of prudential supervision. A bank’s management must be warned, however, of the existence of a risk it faces in receiving as income low, fixed rates on long-term loans, that is, for an expanded period of time, while it has to remunerate the deposits at rates that will be variable, and probably much higher.
For this reason, both foreign exchange risk and interest rate risk should be considered from the outset when assessing whether banks have to be rehabilitated or not. These two risks cannot be ignored. If the size is huge, necessary steps should be taken within the framework of the rehabilitation process.
Enhanced competition is one piece of regulation that must be defined and applied as soon as possible, and, especially, freedom of entrance. The doors should be open to new banks, both national and foreign, providing that (1) they have a strong capital base and that major shareholders have the capability to support the bank if need be in the future (capital to be subscribed by those who have money to invest, not by those who have money to borrow); and (2) management is experienced. In this regard, the feasibility study that is often required in the regulation governing entrance of new banks must be regarded less for its own merit than as evidence of management’s expertise. Enhanced competition is essential to prevent the dramatic increase in lending rates that would crowd out good borrowers and trigger a failure of the financial reform.
At the early stage of implementation, banking supervision, although it is indispensable, as already said, cannot be organized as per the textbook. Shortage of skilled staff and lack of experience limit the scope and frequency of off- and on-site supervision. The staff is still in the process of training, both theoretically and practically. In this regard, short spot checks limited to specific areas may be a useful approach to on-site inspection, keeping in mind that full-fledged examinations will not be carried out before several years. In terms of organization, it is also useful not to distinguish between those who carry out off- and on-site supervision. On the contrary, as it is the most efficient way, it is advisable to set up a structural organization where the same individual is in charge of supervising a few banks, looking at the returns as soon as they are received, and conducting on-site visits from time to time with the assistance of junior staff members. This can be complemented, at least in the early stage, and as long as permitted by the overall number of banks, by regular contacts between high level management of the central bank and commercial bank managers, in addition to the contact that will necessarily take place between these managers and the supervisors.
The main purpose of banking supervision in the early stage of a shift to a market-oriented economy is to draw the attention of those who are in charge of banks to the dangers they may face and to their responsibility. Banking supervision cannot prevent all banks from failure (this is verified everywhere, even in the most advanced countries). Its purpose should be to prevent the emergence of a general banking crisis. I believe that this is something that can be achieved. Let us repeat, however, that if prudential supervision is an indispensable element of the reform, it alone is not sufficient. This means that, in the framework of a financial reform, the liberalization process will need to be delayed if the economic situation is not stabilized.
