Information about Asia and the Pacific Asia y el Pacífico
Asian Financial crises

Chapter 22 Banking Fragility, Effectiveness, and Regulation in Less-Developed Countries

International Monetary Fund
Published Date:
January 2001
  • ShareShare
Information about Asia and the Pacific Asia y el Pacífico
Show Summary Details

Banks in less-developed countries can offer borrowers a stable source of loans, deposits secure from loss, and competitive provision of financial services. These objectives can be achieved by a regulatory system based on adequate required capital and structured early intervention and resolution that permits unrestrained entry and costless (to taxpayers) exit of domestic and foreign banks.

Alan Meltzer begins his critique of the IMF by observing that “[t]he frequency and severity of recent international financial problems in a period of growth, economic progress, and low inflation should raise a series of questions.” The first of these is “Why is there so much financial fragility?” The other questions are directed to the ability of international institutions, particularly the International Monetary Fund, to deal with this situation. The balance of his article provides answers to those questions, with which I largely agree. Because he does not analyze or answer the first question (probably so that I would have something about which to write), I concentrate on it. I begin by outlining the problem—bank failures in less-developed countries. I then describe three features that, I assume, less-developed countries want their banking systems to have: (1) maintenance of banks as sources of funds for borrowers; (2) protection of depositors against loss of their funds; and (3) enhancement of competition among bank providers of financial services. I propose and describe a system of banking regulation that would enhance these objectives at almost no potential cost to domestic taxpayers or the IMF.


Judging from recent history, the banking systems in Asian countries, as well as countries in South America and Europe, appear to be fragile.1 Banks in these countries have experienced losses sufficient to render many of them insolvent.2 Because many of these countries are relatively small and are not served by many banks, the failure of several banks is sufficient to cause considerable economic distress. Were it not for implicit, as well as explicit, government-provided deposit insurance, the banks would have experienced runs, which probably would have disrupted these countries’ payments systems. Depositors as well as stockholders and other creditors would have lost their investments. Further, were it not for government policies that allow insolvent banks to remain open by explicitly or implicitly guaranteeing deposits, borrowers who depend on banks for funds would have had their loans called or been unable to obtain promised or expected funds.

Deposit insurance and forbearance by governments, though, is costly. When depositors believe that they are protected from loss and, as a result, bankers do not fear runs on their banks, ex ante bank owners have incentives to maintain lower levels of capital and take greater risks—the well-known moral hazard effect. Ex post, after their banks have incurred losses that reduce capital, the owners of these weak and insolvent banks would forbear from raising more capital and would continue to pay dividends as long as their governments did not resolve them. Because these banks often are poorly supervised, they might be tempted to make even riskier, negative-present-value loans and investments, since they would have little or nothing to lose and much to gain. The cost of government-provided deposit insurance and forbearance by governments in closing insolvent banks and their failure to deal effectively with weak banks are borne by taxpayers and, possibly, by the IMF.

Some people have pointed to inadequate regulation and supervision of banks as the reason for their recent and past apparent fragility. However, I suggest that a major cause for banks’ losses and insolvencies is regulations imposed by governments that have restrained banks from diversifying effectively, and supervision by government agents that has directed loans towards rather than away from excessively risky loans and investments. Banks in many Asian countries have been required or pressured into making loans to businesses and individuals favored by the government in power. These loans often are made at less than market rates and in amounts that are excessive, particularly considering the low level of the borrowers’ stockholder equity. In addition, the fixed exchange rates maintained by several of these countries have encouraged banks to borrow foreign currency at low rates of interest and lend or invest the funds in domestic currency at nominally high interest rates. If the fixed exchange rate were maintained, this would be a riskless arbitrage. But, disparities in exchange and nominal interest rates cannot be maintained. The result, therefore, is subsidized borrowing and importing which, eventually, must lead to a devaluation of the local currency. Banks and other borrowers, then, must repay their foreign-currency denominated loans with the depreciated domestic currency.3

An additional dysfunctional aspect of bank regulation is that it restrains entry of both domestic and foreign banks in order to protect favored domestic banks from competition. As a result, consumers are disadvantaged and the banking system is more fragile. As Meltzer points out, this situation is exacerbated in small countries, because domestic banks are unable to diversify their investments adequately. These banks invest in domestic businesses, both because they have a comparative disadvantage in lending and investing in other countries and because their banking supervisors encourage, if not require, them to support favored domestic enterprises.

The problems can be solved by a relatively easy-to-implement capital requirement and a system of structured early intervention and resolution. I first outline the assumptions underlying the proposed solution.


I assume that less-developed countries want their banking systems to have three basic features or objectives: (1) banks should be maintained as sources of funds for borrowers; (2) depositors should be fully guaranteed by the government against loss; and (3) there should be competition among providers of lending and deposit services. Achievement of these goals will safeguard both the banking system and most individual banks from collapse, save for the negative effects of substantial macroeconomic shocks imposed by incompetent government policies or very bad fortune (e.g., natural disasters and war). If the proposed bank regulatory system were adopted, the three goals and stability of the banking system can be achieved without costs being imposed on taxpayers. It should be noted that I am assuming that the objective of a country’s banking system should not be to increase the wealth of government-favored projects, enterprises, or people. Indeed, if adopted, the proposed system would make this goal very difficult to achieve.


Unlike the situation in developed and large countries, consumers and businesses in several Asian countries do not have many alternative sources of funds. In these countries, the failure of several banks or a large bank could result in the premature recall of outstanding loans. Also, loan demands would not be met until other lenders expanded or entered the market. This situation probably would impose financial costs on the failed banks’ former borrowers, their customers and employees. These costs could result in their bankruptcy. This is an externality that should be mitigated, as long as the cost of the “cure” does not exceed the benefits therefrom.

To look ahead, costly disruption of borrowing facilities can be avoided efficiently by implementing policies that reduce (indeed, virtually eliminate) the possibility that banks will become insolvent and have to be closed and by reducing barriers to entry of additional institutions that could offer alternative sources of loans.


4.1 Arguments against deposit insurance

Good arguments can be made against government-provided deposit insurance. Similar to the situation facing creditors generally, depositors could assess the risk to which their funds are subject and place their funds in prudently managed and adequately capitalized banks and/or demand compensation for the risk that they might incur. Depositors’ concerns, in turn, provide bankers with strong incentives to maintain their depositors’ confidence, thus avoiding both the need to pay depositors high rates of interest and the possibility of substantial and rapid depositors’ withdrawals (runs). Thus, the moral hazard cost of deposit insurance could be avoided. Furthermore, the central bank could use open market operations to offset runs to currency that otherwise would result in a multiple decrease of the nation’s money supply and bank credit. Although runs that result in bank failures could disrupt the payments system, the system could be protected by rules requiring collateral coverage of liabilities, payments only for good funds, and netting of obligations.

4.2 Arguments for deposit insurance

Nevertheless, three important reasons support the provision of government guarantees to depositors: efficiency—government’s comparative advantage in providing people with a safe and convenient payments and saving medium; the political reality that elected governments will not permit depositors to absorb losses; and enhancement of competition by giving privately owned banks the same advantage as government-owned banks. Whether the deposit insurance should cover all or a portion of deposits is considered in the last part of this section. I conclude that 100 percent coverage of deposits is best.

4.3 Efficiency—government’s comparative advantage in providing “safe” depositories

A country’s economy is enhanced when the means of effecting claims over resources at a point in time (via a circulating medium) and over time (via savings) is efficient. In the past, notes issued by banks were the principal means of making payments. At present, demand deposits largely serve this function. Checks are less costly for people to use and accept, when they do not have to be concerned about the solvency of the bank on which the checks are drawn. Consequently, people would be more likely to use checks, if they did not have to worry about the solvency of banks.

Banks also provide a convenient means for people to save for investment and future consumption. This vehicle is particularly valuable for people with relatively small amounts of resources, because the transaction costs of depositing and withdrawing savings is much less than that incurred for most other investments, such as stocks and bonds. However, people might under-use banks and, hence, engage in less saving than is optimal, if they have to incur the cost of determining and monitoring the extent to which their funds might be lost should a bank fail. Rather than each individual having to make this assessment, a government could offer deposit insurance. In general, then, it would be socially beneficial for a government to insure deposits, if the costs of this insurance (including direct costs imposed on banks and the cost of monitoring and administering the scheme) were less than the costs that individual depositors would have incurred.

4.4 Political reality—deposit insurance will be provided

Even if, on the average, the benefits of government-provided deposit insurance did not exceed its cost, experience indicates that elected governments almost never allow depositors to incur losses as a result of bank failures. In the United States, for example, from the enactment of federal-government-provided deposit insurance in 1933 through 1991 (when the Federal Deposit Insurance Corporation Protection Act—FDICIA—was enacted), almost all depositors in banks, savings and loan associations, and credit unions were protected from loss of their funds, often by the assumption of their deposits by another bank or thrift. FDICIA now emphasizes limiting coverage, in fact as well as in law, to $100,000 per account. Banks and thrifts that failed in Canada, Australia, Japan, and Western Europe have been merged into solvent banks, usually at the request of governments, thereby protecting depositors. Governments (e.g., Mexico) often have protected depositors by nationalizing banks. Other governments have exchanged bad loans for government bonds, to avoid having to close insolvent banks. Only a few exceptions stand out. Australia and some other countries have imposed some costs on depositors by requiring them to take government bonds in payment for their deposit balances. The United Kingdom limits deposit insurance to seventy-five percent of £20,000, a limit which has been maintained in two failures. The United Kingdom recently increased coverage to ninety percent. Argentina has allowed some depositors to take losses; however, most depositors were protected by mergers and by central bank assistance supported by pension fund purchases of obligations, and Argentina has now adopted deposit insurance. The European Union has even mandated deposit insurance for countries (e.g., Denmark) that have not had not needed deposit insurance.

Governments protect depositors from loss for several reasons. First, the banking systems in most countries have many depositors. These people comprise a large voting block that might incur specific personal losses compared to taxpayers generally, who rarely understand that they will bear the cost of paying off depositors in failed banks. Second, large depositors often exercise strong political pressure to be bailed out. Examples outside of Asia include Chile under General Pinochet (despite prior assurances to the contrary), the United States prior to enactment of FDICIA, and the possibility that the United Kingdom would have protected depositors in Barings had the bank’s liabilities not be taken over by a Netherlands bank. Third, large banks and groups of institutions often exercise strong political pressure to prevent being closed. Examples include the U.S. savings and loans associations and farm-state banks. Fourth, the public and legislators fear bank runs and economic collapse. Although, central banks could take offsetting actions, people who want to be protected from loss can exploit this fear.

4.5 Private vs. government-owned banks

Although a government could provide banking facilities directly, via government-run banks or postal savings, the cost to the economy is likely to be greater than if these facilities were provided privately. Government enterprises tend to be inefficient, and they are subject to misallocation of resources as the result of political pressures. Privately owned banks also have greater incentives to determine present and potential customer demands and to develop and adopt efficient innovations, processes, and services.

Nevertheless, several countries, including Asian countries, have government-owned banks and postal-savings systems. These institutions offer depositors 100 percent deposit insurance, because it is not conceivable that they would be permitted to fail with losses imposed on depositors. Hence, privately owned banks are at a comparative disadvantage. Government insurance of deposits for all domestically chartered depositories would, at least, tend to put privately operated banks on an equal footing.

4.6 Limited vs. full deposit insurance coverage

A good case can be made to limit deposit insurance to relatively small deposit amounts. The cost to holders of these accounts of individually assessing and monitoring the risk of loss tends to exceed the benefits they might achieve. Holders of large deposit balances, though, could make these judgements, thereby reducing the moral hazard that otherwise would plague deposit insurance.

However, rather than monitor their bank’s activities, holders of large deposit balances could arrange to transfer their funds almost instantly to another bank, if it appears that their bank might be or might become insolvent. These depositors could accept higher interest payments for deposits placed in a risk-prone bank and incur the cost of monitoring the bank sufficiently to decide when to run. It is likely that they could transfer their balances before the bank is closed. Thus, if the bank does get into financial trouble, the cost will be borne primarily by depositors who are neither fully insured nor sufficiently aware of the bank’s condition. Many people would consider this to be “unfair,” which might lead the government to extend insurance protection, ex post. Furthermore, from the experience of many countries, runs by large depositors tend to result in government intervention and an extension of deposit insurance. For example, in 1984, when it was rumored that the authorities might have to close the Continental Illinois Bank, the Federal Deposit Insurance Corporation extended protection to all creditors of the bank and its holding company to forestall runs both at Continental and at apparently similar large banks. Although FDICIA (1991) imposes conditions designed to avert such “too big to fail” extensions of deposit insurance coverage, there has been no occasion, as yet, to put these provisions to the test.

A more important reason for my conclusion that Asian and other governments should provide 100 percent deposit insurance coverage is that subordinated debt with remaining maturity of at least two years can be more effective than legally uninsured deposits for reducing the moral hazard cost of deposit insurance (as is explained below). However, it is vital that the insurance is limited to genuine deposits. This can be done by defining deposits as liabilities that can be withdrawn only at face value (par) and that bear interest at no more than the market rate on government obligations of the same maturity. The later requirement is suggested to constrain risk-prone banks from readily obtaining funds.

In addition, deposit insurance should be available only to institutions that meet the capital, reporting, and prudential requirements presented below. These requirements, if implemented, should limit the direct and indirect costs of deposit insurance to a very small amount.


It is a truism of economics that entry with few constraints usually is necessary and sufficient for competition to be effective. Consumers benefit when new suppliers enter a market and vie for their trade. New entrants usually must offer better products and services to businesses and consumers to overcome the cost to these potential customers of shifting their accounts from their present bank. Thus, it is likely that new entrants will have developed preferable (to consumers) alternatives that established suppliers either cannot provide or may not want to provide. Collaterally, the mere possibility that other suppliers may enter their market gives established suppliers strong incentives to please their customers and to innovate effectively.

However, completely unregulated entry of banks when their deposits will be covered by government-provided deposit insurance is likely to be costly. Opportunistic or dishonest people might establish banks that invest in excessively risky assets or engage in self-dealing or fraud, because they do not have to be concerned about monitoring by depositors to whom risk-reflecting interest would have to be paid. In addition, existing banks that cannot successfully compete might fail and their depositors would have to be bailed out, with the cost borne by taxpayers or other banks and their customers. If it is likely that the exit of existing banks will impose costs on taxpayers or other banks, it also is likely that government officials will heed existing bankers’ self-interested demands for restraints on entry. Hence, the essential issue, to which I now turn, is how inadequate or unfortunate banks can exit without imposing costs on the implicit or explicit deposit insurance fund or taxpayers and without greatly disrupting their borrowers and other customers.

The regulatory structure now described both permits the banking authorities to allow essentially unrestrained entry and 100 percent deposit insurance. Concomitantly, this structure substantially reduces (almost to zero) the probability that existing banks would exit in a manner that imposes substantial costs on borrowers and depositors.


The proposed system consists of four elements. One is a capital requirement, which effectively deals with the moral hazard problem. The second is a structured early intervention and resolution (SEIR) rule for determining when a bank’s capital is inadequate and what steps first may and then must be taken by the banking authorities, which would effectively deal with the banking authority’s agency problem. The third is limited prudential reporting and examination that enables the authorities to determine that the capital requirement is being met. The fourth is permitting, indeed, encouraging, the entry of foreign banks that are chartered and supervised in countries with strong capital and prudential requirements.

6.1 Capital requirement

The proposed capital requirement, if adopted and implemented, would ensure that institutions offering insured deposits have strong incentives not to take excessive risks and sufficient resources to absorb losses that might be incurred. The banking authorities also will have both the incentive and the means to ensure that the capital requirement is being met.

The proposed capital requirement has two elements: (1) the amount of capital relative to assets (both on and off the balance sheet) should be equivalent to the ratio that banks would hold if their deposits were not covered by government-provided insurance; and (2) subordinated (explicitly uninsured) debt that cannot be redeemed until the authorities can act should be both counted as part of capital and be required. Of course, the effectiveness of any capital requirement depends on the extent to which economic capital can be measured meaningfully. I conclude this subsection by outlining the measurement problems and suggesting solutions to those problems.

6.1.1 Capital-to-assets ratio

A required minimum ratio of capital to assets is necessary to overcome the moral hazard incentive of bankers to operate with low capital, given government insurance of deposits. Without such a requirement, opportunistic bank owners may be tempted to take excessive risks—risks they would not take if they paid the full cost of decisions that turn out badly. Even though some (perhaps most) bankers would not act opportunistically, the capital requirement is necessary for two additional reasons. One is that bankers may misperceive the extent to which they might incur loan losses and other costs, perhaps because economic conditions and their recent experience have been favorable. The other is that capital provides a cushion to absorb losses, whether expected or not.

Capital is the total claim by equity holders and creditors on a bank’s resources that is not insured by the government and, hence, is at risk. It should be sufficient to absorb almost all of the losses that a bank might incur, so that these losses are not imposed on the deposit insurance fund or taxpayers. “Almost all” is specified, because there could be instances of massive fraud and severe economic downturns that deplete an unusually large proportion of a bank’s assets. These risks are insurable. The amount of capital, though, would not be risk weighted, as is the procedure employed for the Basle international capital standard. My objection to the Basle risk weights is that they necessarily are crude (the risk categories are very wide—all commercial loans are weighted equally), the measurement is incomplete (covariances of cash flows are not accounted for), and the weights and categories are subject to political pressures (all residential mortgages are given lower weights than other consumer loans, even though fixed-interest mortgages are subject to greater interest-rate risk). Rather, the required ratio should be sufficiently high to discourage bankers from taking excessive risks and to cover the costs of the risks they do take. Furthermore, as is discussed later, the interest that banks must pay to private investors in the banks’ subordinated debt is the equivalent of a risk-adjusted deposit-insurance premium.

Some bankers might object to a higher-than-present capital requirement, because capital is more costly than deposits. Indeed, capital is more costly, but for only two reasons. One is that deposits are government-insured and the insurance is under-priced. The proposed capital requirement is designed to eliminate this under-pricing. The other results from the income tax statutes present in many countries that permit companies to deduct against taxable income payments to debt holders, but not payments to equity holders. Consequently, debt (which includes deposits) is less costly than equity. Avoidance of this situation is one reason for permitting banks to meet their capital requirement with subordinated debt. In effect, they would simply be substituting a portion of their insured deposits with uninsured debt, both of which offer the same tax advantages.

6.1.2 Subordinated debt

Subordinated debt is, by definition, explicitly uninsured. It must be issued in large denominations, so that people will not confuse it with certificates of deposit and claim they thought it was government insured. To be included in capital, subordinated debt also must not be directly or indirectly repurchased or redeemed by a bank and must have a remaining maturity of at least two years, to allow authorities to act before it may be redeemed. These requirements are imposed to prevent holders of subordinated debt from “running” to avoid taking losses that the bank might incur.

In addition to being available to absorb losses, subordinated debt offers several advantages over equity. This source of capital serves as a means of imposing risk-determined deposit insurance premiums on banks, because, similar to the situation faced by corporations whose debt is not government insured, the risks perceived by debt holders are reflected in the interest rates that the bank must pay to get subordinated debt holders’ funds. Unlike equity holders, subordinated debt holders do not benefit when the risks result in high net profits, while they might lose if the cost of the risks exceeds the equity. The interest rates required on new debt and reported on traded debt, therefore, provide the banking authorities with early warning signals of the risks taken by banks as perceived by investors. The difficulty or ease experienced by banks in refunding maturing debt also provides an early warning. Consequently, there is reason to require banks to have a substantial portion of their capital provided by subordinated debt, preferably debt that must be refunded continuously.

The authorities should recognize a serious possible limitation of subordinated debt. The benefits it provides from risk-reflecting interest rates and early warning signals would be limited or lost if equity holders were permitted also to hold subordinated debt, directly or indirectly. Furthermore, should a bank have incurred losses that absorbed its equity capital, its subordinated debt holders would be, in effect, equity holders. They then would have the same incentives as equity holders to take excessive risks.

Some bank owners, particularly those whose banks are small, might argue that it is difficult for them to find purchasers of subordinated debt. However, they could obtain funds from insurance companies, pension funds, and investors who otherwise would purchase the bonds of other corporations. Bankers who cannot find investors who would be willing to put their funds at risk should not be permitted to operate institutions that hold government—guaranteed deposits.

6.1.3 Measurement of capital—problems and solutions

Mismeasurement of Capital Generally

Capital should be measured as the difference between the economic market values of a bank’s assets and liabilities (other than subordinated debt that serves as capital). Because these values are rarely readily available or even precisely measurable, the amount of capital to which a capital requirement usually applies is accounting (book) capital. (An exception is securities and similar financial obligations that, under US generally accepted accounting principles, must be reported at market or current values, although the balance sheet numbers are not changed from historical-costs if the securities will be held to maturity.) Traditional accounting procedures tend to both under- and over-state capital (relative to economic values). Understatements of capital result from inflation; reported fixed asset values usually are not increased to account for changed price-levels and long-term liabilities that were sold before an inflation was expected are not reduced to their present values. In addition, intangible assets, such as business development, employee training, goodwill, and charter value, are not capitalized and recorded as assets. Since these understatements result in greater economic- than book-value capital, there is a regulatory advantage.

Accounting (book-value) capital tends to be overstated as a result of changes in interest rates that decrease the present value of assets more than the present value of liabilities (duration imbalances) and when loans that might not be repaid as contracted are not adequately written down. The overstatement resulting from interest rate changes could be corrected by requiring banks to use current interest rates to determine and record the present values of restructured loans and fixed-interest obligations.

The overstatement of capital that results from understating allowances for bad loans is more important and more difficult to solve. Understatement of loan loss allowances (reserves) has been found at banks in many countries, both developed and underdeveloped, as documented by Goldstein (1997, pp. 16–18.) Three factors appear to be responsible. First, bankers may deliberately understate the allowance to avoid reducing their reported capital. They can do this by not recognizing seriously delinquent loans as uncollectible until the borrower has formally been declared bankrupt. This practice can be reduced by constructing explicit rules that specify when loans must be reported as nonperforming (e.g., when they are sixty or ninety days past due) and when they must be reserved against and written off. Even then, bankers can adopt an often-costly procedure of lending more funds to delinquent borrowers so that they can keep interest and principal payments current (a practice called “ever-greening"). Only bank examinations or audits by independent public accountants can detect this subterfuge. Second, the banking authorities often permit or even direct bankers to forbear from writing down bad loans. They do this to avoid having to recognize publicly banks’ weaknesses and insolvencies, which would require the authorities to take politically difficult actions, including requiring weak banks to obtain more capital and resolving insolvent banks. Third, the loan loss allowance may have been reasonably recorded, ex ante, but is inadequate, ex post, when business conditions unexpectedly changed. The proposal outlined below is designed to deal with all three situations.

6.1.4 Economically meaningless capital at closely held or controlled banks

The owners of closely held or controlled banks, which are common in less developed and small countries, have both the incentive and opportunity of evading the capital requirement by fulfilling it with economically misstated assets. They can do this by “paying” for their bank stock with substantially overvalued assets (such as undeveloped land and shares in companies that are not publicly traded or that are controlled) or with loans from their banks or related banks. Consequently, purchasers of banks stocks should be required to pay for their shares in cash. They should not be allowed to borrow from their own bank, directly or indirectly, at any time. Nor should they be allowed to borrow from another bank, if their bank has deposited funds with that bank.

Direct loans to stock holders in banks are easy to detect and, in most countries, are not permitted. However, bank owners can avoid a self-lending prohibition with reciprocal borrowing, wherein two unrelated parties (A and B) mutually fund their capital contributions to banks that they control with loans from the other’s bank. Assume that party A becomes bankrupt and cannot pay his loan at all to bank B. That bank would become insolvent, if its capital were less than the amount of the loan. With bank B insolvent, party B might not be able to repay her loan to bank A. This, in turn, could result in Bank A becoming insolvent.5 A similar situation is described by the Governor of the Central Bank of Ecuador, Augusto de la Torre. He explains:

We in Ecuador have found that business groups that own industrial companies and banks have actively used offshore jurisdictions to generate fictitious capital increases. Domestic deposits have been transformed into accounting capital (as opposed to real capital) by circulating them around the various Caribbean jurisdictions and others where similar practices prevail. A striking example is a bank in which the Ecuadoran authorities intervened. [F]rom the end of 1994 until the beginning of 1996 [t]he owners [of this very rapidly growing bank] had created fictitious capital by transferring deposits to offshore centers and then using those deposits to lend to the bank’s shareholders, who in turn had used those funds to increase the bank’s capital, (de la Torre, 1997, p. 79)

Rules can and should be established to prohibit evasion of the requirement that bank owners provide capital represented by meaningfully valued assets. Bank examiners and independent external auditors can uncover the mismeasurement of bank capital, and they should be specifically charged with this task. For example, they should closely examine all large loans to determine the recipients and the reason for and (if possible) disposition of deposits in other banks. These examinations should be preceded by an analysis of the holdings and relationships among a bank’s capital holders (equity and subordinated debt). Closely held and rapidly growing banks should be examined particularly carefully.

Therefore, I conclude that capital can be measured meaningfully, although probably imperfectly. That is one reason that a higher ratio of capital to assets is recommended. Before discussing the ways in which bank reporting and prudential examination can be effective for assuring that the measurements are likely to be accurate, I describe the proposed rules that specify when and how the banking authorities should act to ensure that the capital requirement is maintained.

6.2 Structured early intervention and resolution (SEIR)

SEIR was proposed in 1988 by Benston and Kaufman and was substantially adopted by the United States in 1991 as part of the Federal Deposit Insurance Corporation Improvement Act (FDICIA). It provides incentives for and imposes requirements on the banking authorities to act expeditiously and responsibly. Together with the capital requirement, the result should be almost no depositor bailouts.

Four capital zones or trip wires (or traunches) are established that define first when the authorities, at their discretion, may act and when they must act. The ratios suggested here are higher than those specified in FDICIA, which are too low, even for the United States. The ratios should be higher for less-developed countries for the reasons enumerated by Rojas-Suárez and Weisbrod (1995), derived from their studies of Central and South American countries and by Goldstein (1997), who reviews bank failures in many developing countries:6

  • 1. Financial data often are not reliable;
  • 2. Arms-length relationships between borrowers and banks often are not present;
  • 3. The legal environment makes it difficult for banks to gain possession of collateral in the event of default; and
  • 4. National economic policies often have been destabilizing, resulting in sharp downturns in economic conditions and losses in the values of bank assets.

The four capital-to-assets-ratio zones and the pre-specified actions by the authorities are as follows.

6.2.1 Adequately capitalized banks

These are banks with capital-to-assets ratios approximately equal to those of firms without government-provided deposit insurance (perhaps twenty percent); they would be subject to minimum supervision that is limited to determining that the bank was reporting its financial numbers correctly and was not being managed fraudulently or recklessly.

6.2.2 First level of supervisory concern

Includes banks with capital-to-asset ratios below, say, twenty percent, but above, say, twelve percent; they would be subject to increased regulatory supervision and more frequent monitoring. The authorities would require a business plan for quick recapitalization and, could, at their discretion, impose such sanctions as restricting growth, prohibiting dividend payments, and restraining payments for services provided by related companies.

6.2.3 Second level of supervisory concern (inadequate capitalization)

If the bank does not bring its capital-to-assets ratio back into compliance, the authorities must impose additional and harsher sanctions, including prohibition of dividend payments and interest payments to subordinated debt holders and restrictions on growth and on transfers of funds to related entities.

6.2.4 Resolution

Should a bank’s capital-to-assets ratio falls below a predetermined point (say, five percent), the authorities must resolve the bank quickly through sale, merger, or liquidation. Rather than permit this to happen, an economically solvent bank most likely would voluntarily raise its capital-to-assets ratio into compliance, liquidate, or sell out to or merge with another institution.

6.3 Prudential reporting and examination

6.3.1 Reporting requirements

The banking authorities should regularly monitor the activities of institutions with government-insured deposits to determine whether or not the capital requirement is being complied with and to provide an early warning of possible problems that warrant closer examination and may require supervisory intervention. The requirement outlined should impose low costs on the institutions, because they surely would be maintaining and reviewing most of the information for purposes of internal management and oversight by their boards of directors.

Deposit-insured institutions should submit monthly and annual reports to the banking authorities. The annual reports should be audited and attested to by independent external auditors (certified public accountants) who are approved by the banking authorities. All directors of the bank should sign the reports. To the extent feasible, assets and liabilities should be stated at current (present or market) values. Delinquent and nonperforming loans should be clearly defined (e.g., payments over due by more than thirty or sixty days or where additional loans are made to enable payments to be met). Loans to related parties should be identified as such, as should the total amount of loans made to an associated group.

In addition, the authorities should conduct regular field audits to check that the reports are correct. These audits should be directed towards and limited to concerns about the measurement and adequacy of capital. Additional examinations should be made where the monthly reports or other information indicates substantial growth or possible exceptionally risky operations. Banks owned or controlled by a few persons or groups should be more closely monitored and examined. The direct cost of these examinations should be charged to the banks examined.

6.3.2 Prudential regulations

Self-dealing must be prohibited. This includes all loans to stockholders, subordinated bondholders, and bank managers and to parties related to them, whether personal or business. An exception, however, could be made for loans made to parties who have but a small interest in the bank (e.g., ownership of less than five percent of total capital). Loans to one or a related group of borrowers should be limited to a percentage of capital (perhaps fifteen percent). Additional, so-called prudential regulations that limit the activities in which banks can engage, are not required and are likely to be detrimental both to banks and consumers, because they tend to limit competition rather than actually reduce the risks in which banks might engage.

6.4 Foreign banks

Gavin and Hausmann (1997) point out that, to achieve a competitive market for banking services, less-developed countries must meet several conditions. These include a stable macroeconomic environment, effective bank regulation conducted by experienced, well-trained people, and judicial enforcement of contracts. In addition, in some countries relatively few domestic investors may be willing to meet the capital requirements that should be imposed on banks with government-insured deposits. Gavin and Hausmann propose opening domestic markets to foreign banks. These banks can bring an additional source of funds and services to consumers. Not only will they tend to offer better products to consumers, as would a domestic new entrant, but they often bring with them improved procedures and products that were developed in their home countries. Foreign banks generally find it desirable to hire local people. Hence, they also can provide training for people who might leave to work for domestic banks or to establish their own banks. In addition, because foreign banks are diversified geographically, they are less subject than are domestic banks to domestic macroeconomic changes.

Consequently, I suggest that banks owned by foreigners that are chartered and supervised in countries with strong capital and prudential requirements should be encouraged to open branch offices in less-developed countries. These banks should be required to insure their deposits either from their home country funds or from other sources, such as collateral or insurance policies written by reputable and secure companies. The local banking authorities, therefore, would only have to determine that these banks could fully meet deposit withdrawals by domestic depositors. Otherwise, there should be no restraints on the repatriation of funds by foreign banks to their home countries. Domestically chartered subsidiaries of foreign banks should be treated in the same manner as are other domestically chartered banks.


Goodhart et. al (1998, Chapter 7) review banking crises in the Nordic countries, the United States, Japan, Argentina, Chile, and Mexico, from which they draw “three basic principals” with which, they say, “good banking crisis management must begin.” The system of bank regulation and supervision proposed above fulfils their principals:

  • “Ensure that parties responsible for the crisis bear most of the costs of restructuring” the capital requirement together with the reporting requirement should prevent owners of banks from taking excessive risks that imposed deposit insurance costs on taxpayers or on other banks;
  • “Prevent problem banks from expanding credit to delinquent borrowers” the structured early intervention and resolution (SEIR) procedures together with the reporting and prudential requirements should prevent this from happening;
  • Avoid financing the programme with inflation by making the restructuring programme a high priority” the capital requirement that includes subordinated debt and SEIR should prevent restructuring from becoming necessary.

Better yet (I believe), the solution, if adopted, would largely prevent crises from occurring, would enhancing competition by permitting relatively free entry, and would not impose costs on banking customers and taxpayers as a result of bank failures.

The proposed system can be implemented without disrupting present banking systems or banks that are well capitalized. Undercapitalized banks will be affected, but this is as it should be. Owners of some undercapitalized banks might prefer to merge with or be acquired by other banks, thereby relieving the banking authorities from later having to intervene and possibly resolve them. With entry unrestrained (except for capital and reporting requirements), consumers will not be disadvantaged from mergers and acquisitions that reduce the number of competitors.


See Benston and Kaufman (1995) for a comprehensive review of studies of banking fragility, which concludes that although banks are fragile, banking systems are not unstable.


See Goldstein (1997, pp. 3-21), Lindgren, Garcia, and Saal (1996), and Goodhart et. al (1998) for a comprehensive review of studies on the causes and effects of recent banking crises in many countries.


This situation might be dealt with by legislation that must be adopted by a country as a requirement for it to be eligible for an IMF loan. The legislation would require a mandatory reduction (“haircut”) in the principal amount of loans made to domestic banks that are denominated in a foreign currency, if these loans are withdrawn after IMF assistance is announced. This proposal is detailed in Shadow Financial Regulatory Committee’s Statement 145 (May 4, 1998) and outlined in Robert Litan’s comments in this voume.


An earlier version of much of the following materials was presented at an InterAmerican Development Bank conference in Buenos Aires, Argentina in August 1997 and is published in Spanish (Benston, 1998.)


This illustration is derived from Rojas-Suárez and Weisbrod (1997, pp. 46-49), who point to several instances in South American where this was found to have occurred.


Goldstein (1997) describes similar problems present in less developed countries.


    Benston, George J.1998, “Ingreso y Retire de Bancos in América Latina y el Caribe: Preocupaciones por la Polítaca Publica y Propuesta de Solución,” Superbancaria, Superintendencia Barcaria de Colombia, Marzo, 25–36

    • Search Google Scholar
    • Export Citation

    Benston, George J., and George G.Kaufman. 1988, “Risk and Solvency Regulation of Depository Institutions: Past Policies and Current Options,” Monograph Series in Finar and Economics, Salomon Center, New York University Graduate School of Busini Administration, monograph -1.

    • Search Google Scholar
    • Export Citation

    Benston, George J., and George G.Kaufman.1995, “Is the Banking and Payments System Fragile?,” Journal of Financial Services Research, 9:209–240.

    • Search Google Scholar
    • Export Citation

    de la Torre, Augusto.1997, “Comment,” in Safe and Sound Financial Systems: What Works for Latin America, Liliana Rojas-Suárez (ed), Washington, DC: Inter-American Development Bank, 79–81.

    • Search Google Scholar
    • Export Citation

    Gavin, Michael, and RicardoHausmann.1997, “Make or Buy? Approaches to Financial Market Integration,” in Liliana Rojas-Suárez (ed)., Safe and Sound Financial Systems: What Works for Latin America, Washington, DC: Inter-American Development Bank, 133–162.

    • Search Google Scholar
    • Export Citation

    Goldstein, Morris.1997, The Case for an International Banking Standard, Washington, DC: Institute for International Economics, 1997.

    Goodhart, Charles, PhilippHartmann, DavidLlewellyn, Liliana RojasSuárez, and StevenWeisbrod. 1998, Financial Regulation: Why, how and where now?, London and New York: Routledge (in association with the Bank of England).

    • Search Google Scholar
    • Export Citation

    Lingren, C.–J., GillianGarcia, and M.I.Saal. 1996, Bank Soundness and Economic Policy, Washington, DC: International Monetary Fund.

    Rojas Suárez, Liliana, and Steven R.Weisbrod. 1995, Banking Crisesin Latin America: Experiences and Issues, Washington DC: Inter-American Development Bank, September.

    • Search Google Scholar
    • Export Citation

    Rojas Suárez, Liliana and Steven R.Weisbrod. 1997, “Towards an Effective Financial Regulatory and Supervisory Framework for Latin America: Dealing with the Transition,” in Safe and Sound Financial Systems: What Works for Latin America, LilianaRojas-Suárez, (ed.), Washington, DC: Inter-American Development Bank, 35–74.

    • Search Google Scholar
    • Export Citation

    Shadow Financial Regulatory Committee. 1998.

    Other Resources Citing This Publication