1. A definition

The protection offered under a system of deposit insurance is a guarantee that all or a limited amount of the principal and the interest accrued on protected accounts will be paid. The guarantee may be explicitly given in law or regulation, as in the United States and many other countries, offered privately without government backing but under enabling legislation, as in Germany, or may be inferred implicitly from the verbal promises and/or the past actions of the authorities. A recent survey (Kyei, 1995) of deposit protection practices around the world reveals that there are more implicit schemes (55) than explicit ones (47), although there has been a tendency over the past few years for countries to adopt explicit systems because of the problems inherent in implicit schemes. Moreover, countries feel compelled to intervene in their banking systems in times of crisis; consequently, most of the remaining countries, including even those that explicitly disclaim deposit insurance in any form, can be treated as though they have it.

This paper argues that there are advantages to having an explicitly defined deposit insurance scheme (DIS). In fact, the EU promulgated a Directive in May 1994, which required each member country to adopt, by mid-1995, an explicit system of deposit insurance that meets at least some minimal standards. The Directive states in Articles 3 and 7:

“Each Member State shall ensure that within its territory one or more deposit-guarantee schemes are introduced and officially recognized.… No credit institution…may take deposits unless it is a member of such a scheme.… Deposit guarantee schemes shall stipulate that the aggregate deposits of each depositor must be covered up to ECU 20,000 in the event of deposits’ being unavailable.” ^2/ (Official Journal of the European Communities, Directive 94/19/EU, pp. 8-9).

2. An outline of the paper

Section II of the paper describes how deposit insurance differs from other forms of insurance and presents of model of deposit insurance that stresses the trade-off between financial stability (the principal benefit) and moral hazard (the main pitfall) of a formal DIS. ^1/ Section III contrasts two opposing philosophies that govern the configuration of the financial system in different countries, asks why warranties, insurance and guarantees are given, and inquires what characteristics of deposits and their suppliers necessitate a lender of last resort (LOLR) to protect bank liquidity and a guarantor in the event of insolvency. Section IV lists and evaluates potential objectives for deposit protection. Section V examines the pitfalls of deposit insurance and suggests some measures to avoid them. The design of an explicit scheme that avoids the major drawbacks in normal times and during crises is analyzed systematically in Section VI. Section VII summarizes the paper’s conclusions. The Appendix briefly describes systems of preference for depositors and their insurer that give them legal priority over the assets of a failed institution.

1. How deposit insurance differs from other forms of insurance

Although commonly used, the term deposit insurance is really a misnomer, because the depositor does not purchase insurance in the sense that a parent buys life insurance to provide for his/her family in the event of his death or a home-owner obtains protection against fire or burglary (Bartholomew, 1990). While deposit insurance exhibits some superficial similarities to life/health and property/casualty insurance (“regular insurance”), it differs from them in four fundamental respects.

First, bank failures are not the independent events that other forms of insurance typically cover. Rather, failures tend to occur in waves, partly in response to a severe recession or some other macroeconomic shock, partly because the legal/regulatory/ supervisory structure is inadequate, and partly because bank failures can be contagious when the failure of one bank brings down its counterparties. ^1/

Second, “regular” insurance aims to protect against “acts of God” or against the acts of other men over which the insured party has no control. Thus, life insurance policies do not cover suicide. There is always a danger (moral hazard) that the insured may reduce his efforts to minimize the incidence of the insured event. For example, depositors may monitor the condition of their bank less closely once they obtain deposit protection. Yet, the incidence of moral hazard is much greater for deposit insurance and extends beyond negligence on the part of insured depositors. As Benston and Kaufman (1986, 1994) point out, bank failures are often self-inflicted wounds arising from mismanagement. Moreover, the losses from such failures can be substantially reduced if the supervisor closes the bank before it becomes insolvent, whereas delayed resolution tends to exacerbate those losses. Thus, moral hazard can infect the actions not just of insured depositors, but of other several parties (including uninsured creditors, owners, managers, supervisors and politicians), as is discussed below in the “pitfalls” section.

Third, while deposit insurance is aimed principally at protecting deposits and their owners, it also serves to ensure the continued existence of individual banks by providing them with continued funding even in times of stress. That protection for the bank occurs because, unlike the bilateral nature of a property and casualty insurance contract, deposit insurance directly involves three parties—the guarantor, the depositor, and the depository institution. Thus, Kane (1995) likens deposit insurance to a bonding or surety contract. ^2/ As Kane points out, the immediate benefits of the surety’s guarantee are divided between the bank’s depositors, whose funds are protected, and the institution that receives credit enhancement to lower its funding costs and some protection against runs that could threaten its continued existence. But other interested parties (such as supervisors and politicians) can also take advantage of the availability of insurance to pursue their own interests, oftentimes to the disadvantage of the DIS.

Fourth, to make the guarantee effective in avoiding the danger of financial panic and other negative externalities, the government typically stands behind the deposit insurer in times of widespread failures, either explicitly or implicitly. Other forms of insurance are less likely to be government-backed.

These nonstandard features of deposit insurance dominate the answers to the questions addressed below whether deposit insurance is useful or not and, if it is, how much should be provided and how it should be delivered.

2. A model of financial instability and moral hazard

A government may sponsor a system of deposit insurance for a variety of reasons. Section IV below lists 10 of the most common possible objectives. In addition, some governments may have additional reasons that are not discussed. The principal rationale is, typically, a desire to promote financial stability by reducing the incidence of bank runs. This observation points out the existence of a trade-off between financial stability and moral hazard that is illustrated in Figure I.

This trade-off, depicted in Figure 1 by the line AA’, occurs because the government can reduce the incidence of runs today by increasing the proportion of bank liabilities that are covered by insurance, say from point A to A’. The intercept of the line is determined, inter alia, by the public’s expectations regarding the extent of the insurance coverage that will be provided. If the DIS does increase coverage from A to A’, the curve is likely to ratchet outward in the future from the origin to the line BB’, which represents a deterioration in the terms of the trade-off. The shift occurs because raising coverage unexpectedly increases moral hazard. Conversely, if the DIS provides less coverage than expected, the line may shift inward toward the origin as more market discipline comes into play to keep banks sound.

View Full Size

Moral Hazard and Financial Instability

Citation: IMF Working Papers 1996, 083; 10.5089/9781451850499.001.A001

• Download Figure
• Download figure as PowerPoint slide

Greater moral hazard raises the probability of instability in the future. Thus, when the government is facing a crisis and increases coverage to reduce the number of failures, it gains an immediate advantage that reduces the insurer’s resolution outlays and its need to back-stop the insurer’s financial resources. But if, as seems likely, the cost of financial instability to the economy is a monotonically increasing function of the extent of bank runs and panic, increasing coverage today builds trouble that raises the probability and cost of instability in the future.

1. Contrasting models of the financial system

This section introduces two polar models of the financial system that help to analyze the dynamics of deposit protection. Under an unregulated banking model, entry is uncontrolled; regulation and supervision are minimal; competition is abundant; there is no deposit insurance (given enough information, bank customers are expected to look after their own interests with the help of an efficient legal system); the government has made a credible commitment not to bail out failed banks; and, indeed, individual bank failures are expected and tolerated. The central bank provides liquidity to the banking system via open market operations and, in some variants of the deregulated model, also assists individual banks by last resort lending through the discount window.

It is intended that such a system be kept functional through market discipline: when a bank undertakes too much risk, its funding costs rise at first and later its creditors may become unwilling to lend to it at any price. If the troubled bank is solvent but illiquid, collateralized lender-of-last-resort lending avoids its “fire sale failure;” but if insolvent, it is sold or goes into receivership/liquidation. Its shareholders lose their capital and its depositors and other creditors suffer losses, at least to some extent. Occurring in isolation, such a development serves to warn other bankers to avoid the same fate and so keeps the system sound.

This model is conceptual; there is probably no country today that follows the unregulated banking model in its full extreme. ^1/ The reason may be that even those countries that are prepared to tolerate an occasional bank failure may intervene to avoid what they fear may be a systemic collapse. Indeed, since World War II, most countries that did not have either implicit or explicit insurance have rescued depositors when faced by widespread bank failures. This situation tends to be the worst of all possible worlds because the absence of a well-designed DIS creates an uncertainty on the part of depositors that can aggravate bank runs, ultimately necessitates greater coverage than would have been offered otherwise, and leaves the government with no alternative but to foot the bill. Nevertheless, the regulatory philosophies in New Zealand and South Africa are attracted to this model and follow it to some extent. ^2/

At the other end of the spectrum is a regulated banking system with controlled entry that limits competition so that banks’ charters have significant value (a rent); extensive regulation, strict supervision and enforcement; a system of explicit depositor protection (hopefully designed to preserve market-based incentives); less emphasis on market discipline; some concern over even individual failures; and firm control over the exit process. Most industrialized countries follow this approach.

The fact that design and certainty are important favors explicit DIS over implicit protection. In fact, there are several disadvantages to reliance on implicit guarantees. For example, there may be a substantial degree of uncertainty regarding the existence, timing and extent of implicit coverage, which often protects the largest and most politically powerful banks that are perceived to be “too big to fail,” whereas a formal system of deposit protection can (partially) redress the balance to favor smaller banks. Moreover, the cost of implicit systems falls heavily on the government, whereas explicit DIS normally require banks to cover the cost except in times of economic emergency.

2. Products that receive guarantees or warranties

A substantial number of products and services (summarized in Table 1) receive warranties or guarantees. For the purchase of many goods and services, the reputation of the seller is important. ^1/ But it may not be sufficient. A strong reputation takes time to build but can be destroyed overnight (by an airline crash, a fatal error by a hospital, or a death resulting from a contaminated medicine). So, the seller may signal his belief in the quality of his product by offering a warranty.

Table 1.

Types of Protection Available

Table 1.

Types of Protection Available

NO GOVERNMENT INVOLVEMENT
●	No recourse
	Inexpensive, short-lived products without consequences, such as flowers.
●	Replaceable or refundable
	Not expensive, longer-lasting products where safety is not a concern, produced or sold in competitive or oligopolistic markets, such as clothes.
●	Producer’s or retailer’s reputation
	Longer-lasting products, such as jewelry, and services where safety may or may not be a consideration—over-the-counter drugs and fast food.
●	Manufacturer’s reputation and warranty
	More expensive, longer-lasting products where reliability and safety may be important—automobiles, household appliances.
●	Mutual fund backup
	The fund’s parent company provides additional resources to avoid “breaking the buck” in order to retain the fund’s reputation for financial prowess.
●	Private evaluation
	Consumer organizations evaluate and publish their findings on the relative merits of different competitors in the market-various, especially appliances and cars.
●	Consumer insurance
	Products where performance and durability are factors, the consumer upgrades the warranty by purchasing additional insurance—electronics.
●	Life, health, property and casualty insurance
	The consumer purchases protection against catastrophic losses.
●	Credit life insurance
	Borrowers insure their heirs or themselves against loss from death or unemployment—mortgages.
●	Licensing and producer bonding/insurance
	Long-lasting products and services that have severe consequences from non-performance—doctors, home improvement contractors.
GOVERNMENT INVOLVEMENT
●	Government inspection
	Common for products and services where the consumer cannot inspect directly and safety is important—hospitals, airlines and depository institutions.
●	Investor protection
	Investors in the EU and the United States are guaranteed a minimum level of compensation in the event that their broker defaults. They are not covered against loss of principal if the market declines.
●	Guarantee funds for insurance companies
	Purchasers of life and property/casualty insurance in the United States are covered by “state” insurance in all 50 states and D.C. Levies on members companies are made ex post.
●	Deposit insurance
	A financial guarantees for depositors against loss of principal and interest on their (small) deposit. While often privately funded and run, it is typically backed by the Government.

View Table

Table 1.

Types of Protection Available

NO GOVERNMENT INVOLVEMENT
●	No recourse
	Inexpensive, short-lived products without consequences, such as flowers.
●	Replaceable or refundable
	Not expensive, longer-lasting products where safety is not a concern, produced or sold in competitive or oligopolistic markets, such as clothes.
●	Producer’s or retailer’s reputation
	Longer-lasting products, such as jewelry, and services where safety may or may not be a consideration—over-the-counter drugs and fast food.
●	Manufacturer’s reputation and warranty
	More expensive, longer-lasting products where reliability and safety may be important—automobiles, household appliances.
●	Mutual fund backup
	The fund’s parent company provides additional resources to avoid “breaking the buck” in order to retain the fund’s reputation for financial prowess.
●	Private evaluation
	Consumer organizations evaluate and publish their findings on the relative merits of different competitors in the market-various, especially appliances and cars.
●	Consumer insurance
	Products where performance and durability are factors, the consumer upgrades the warranty by purchasing additional insurance—electronics.
●	Life, health, property and casualty insurance
	The consumer purchases protection against catastrophic losses.
●	Credit life insurance
	Borrowers insure their heirs or themselves against loss from death or unemployment—mortgages.
●	Licensing and producer bonding/insurance
	Long-lasting products and services that have severe consequences from non-performance—doctors, home improvement contractors.
GOVERNMENT INVOLVEMENT
●	Government inspection
	Common for products and services where the consumer cannot inspect directly and safety is important—hospitals, airlines and depository institutions.
●	Investor protection
	Investors in the EU and the United States are guaranteed a minimum level of compensation in the event that their broker defaults. They are not covered against loss of principal if the market declines.
●	Guarantee funds for insurance companies
	Purchasers of life and property/casualty insurance in the United States are covered by “state” insurance in all 50 states and D.C. Levies on members companies are made ex post.
●	Deposit insurance
	A financial guarantees for depositors against loss of principal and interest on their (small) deposit. While often privately funded and run, it is typically backed by the Government.

Warranties are typically offered to consumer goods (such as automobiles and electronic equipment) that are relatively expensive, obtained infrequently because they are expected to have an extended life, and are technologically sophisticated so that it is difficult for the consumer to evaluate the merits of a product in comparison to those of its competitors. In short, the manufacturer has asymmetrically more information than the consumer. In this case, it is more effective for the manufacturer to advertize its product and offer a warranty to back its claims than to expect individual consumers to conduct extensive and costly comparative searches or to purchase consumer reports or agency ratings. ^2/

Deposit insurance provides protection against a similar informational deficiency. Bankers have asymmetrically more information about the condition of their asset portfolios, particularly their loans, than their depositors. While it is reasonable to expect investors to monitor banks, it would be too expensive and time consuming for regular customers to examine the relative condition of competing banks. ^1/ It is more cost-effective for banks themselves to offer a mutual guarantee of deposits in the system. In addition, lenders (and governments) sometimes require weaker mortgage and consumer credit applicants to buy “credit life insurance” from a third party surety, who guarantees that the lender will be paid if the borrower dies or defaults. This insurance also protects families who have lost their bread-winner from having their property repossessed and it insulates politicians from the public outcry that would accompany a bank’s exercise of its collateral rights in these circumstances.

Individuals and groups frequently protect themselves against potentially catastrophic losses resulting from “acts of God and man” by obtaining life, health, property and casualty insurance. Governments sometimes offer such protection where private parties find it uneconomic.

In cases where the public’s life, health, and wealth are at stake, the seller’s reputation and a private warranty or guarantee may not be enough, especially where it is easy for suppliers to signal good quality misleadingly. In many countries, the government or a government-backed private agency will issue an initial license and continue to regulate and inspect the product- or service-provider to protect the public from man-made disasters from unduly cost-conscious airlines, careless hospitals, faulty drugs, and failed banks.

A bank’s reputation will be important to its depositors’ confidence in the safekeeping of their money. It is built over a long period, but it can also be destroyed rapidly; then its depositors may run and bring down the bank. A run can become contagious because the losses from a bank default are felt instantly through the interbank market and the payments system and can destroy or weaken other banks. Governments fear that the public may not be able to distinguish bad banks in the economy from good banks, because the quality of bank assets is hard to evaluate (Mishkin, 1991, 1994; Stiglitz, 1992), so bank failures may spread to good banks. ^2/ Thus, most intervene to prevent collapse by regulating and supervising banks to protect the public from loss and the financial system from disruption. Because the public tends to regard bank regulation and supervision as government certification of an individual bank’s safety and the stability of the banking system, many governments stand behind even privately owned banks and provide systems of deposit insurance, which they often backstop even when they are privately funded.

3. Do deposits and their suppliers need a guarantee?

Banks play vital roles in the economy. They operate the payments system; intermediate between lenders and borrowers; transmit monetary policy to the economy; obtain, process and signal information; and discipline borrowers to ensure the efficient allocation of financial resources in the economy. Banks typically dominate the payments system and Goodfriend (1991) argues that it is efficient for them to do so. While some of these roles could be performed by other financial institutions, banks are often the only source of external finance for borrowers who cannot access the stock and bond markets, because banks have a unique role with regard to information. They overcome the free-rider problem associated with monitoring borrowers by keeping information about their borrowers confidential. In short, in most countries and particularly in developing and transitional economies, banks are important because their roles are difficult for other financial institutions to replicate. Thus, while it is counter-productive to guarantee the survival of individual banks, authorities seek to protect the system as a whole.

a. The nature of deposits compared to other financial instruments

While banks are important, the unique juxtaposition of their assets and liabilities makes them vulnerable to illiquidity and insolvency. Commercial banks fund their illiquid and opaque assets with relatively small amounts of equity and large quantities of debt, making them prone to insolvency. Bank debt typically consists of non-deposit obligations and short-maturity deposits that include: demand, savings, time (both regular and those represented by negotiable certificates of deposit or “CDs”), interbank, and government deposits. Their assets are often of longer-term maturity exposing them to illiquidity. Deposits are usually denominated in domestic currency, but can be denominated in a foreign currency, in which case the bank faces foreign exchange risk. Table 2 lists the characteristics of deposits and competing financial instruments and the quintessential assets of the institutions that offer them.

Table 2.

Characteristics of Financial Instruments and the Institutions that Offer Them

^1/While a mutual fund’s contract does not provide for fixed principal, many fund operators have taken steps to ensure it.

^2/Money market mutual funds in the United States invest in safe, liquid assets that reduce credit and interest-rate risk.

^3/Investing in safe liquid assets makes the fund less opaque. In addition, there are strict disclosure requirements in the United States.

Table 2.

Characteristics of Financial Instruments and the Institutions that Offer Them

Instrument	Characteristic of the Instrument					Typical Asset of the Institution offering the Instrument
Commercial Bank Deposits	Fixed Principal	Payment on Demand	Sequential Servicing	No Secondary Market		Credit Risk	Long Term	Opaque
Time deposit regular	x	only with a penalty	only with a penalty	x		x	x	x
Negotiable CD	x					x	x	x
Savings deposit	x	notice is often required	x, after notice, if required	x		x	x	x
Demand deposit	x	x	x	x		x	x	x
Interbank deposits	x	often overnight	x	x		x	x	x
Government deposits	x	x often	x			x	x	x
Narrow Bank Deposits
Time deposit regular	x	only with a penalty	with a penalty	x
Negotiable CD	x
Savings deposit	x	notice may be required	x, after notice, if required	x
Demand deposit	x	x	x	x
Nonbank Instruments
Equity					x	x	x
Bond	x				x	x	x
Mutual Fund closed end					x	x	x
Mutual Fund open end		x	x	x	x	x	x
Money Market Mutual Fund	1/	x	x	x	less 2/		less 3/

^1/While a mutual fund’s contract does not provide for fixed principal, many fund operators have taken steps to ensure it.

^2/Money market mutual funds in the United States invest in safe, liquid assets that reduce credit and interest-rate risk.

^3/Investing in safe liquid assets makes the fund less opaque. In addition, there are strict disclosure requirements in the United States.

View Table

Table 2.

Characteristics of Financial Instruments and the Institutions that Offer Them

Instrument	Characteristic of the Instrument					Typical Asset of the Institution offering the Instrument
Commercial Bank Deposits	Fixed Principal	Payment on Demand	Sequential Servicing	No Secondary Market		Credit Risk	Long Term	Opaque
Time deposit regular	x	only with a penalty	only with a penalty	x		x	x	x
Negotiable CD	x					x	x	x
Savings deposit	x	notice is often required	x, after notice, if required	x		x	x	x
Demand deposit	x	x	x	x		x	x	x
Interbank deposits	x	often overnight	x	x		x	x	x
Government deposits	x	x often	x			x	x	x
Narrow Bank Deposits
Time deposit regular	x	only with a penalty	with a penalty	x
Negotiable CD	x
Savings deposit	x	notice may be required	x, after notice, if required	x
Demand deposit	x	x	x	x
Nonbank Instruments
Equity					x	x	x
Bond	x				x	x	x
Mutual Fund closed end					x	x	x
Mutual Fund open end		x	x	x	x	x	x
Money Market Mutual Fund	1/	x	x	x	less 2/		less 3/

^1/While a mutual fund’s contract does not provide for fixed principal, many fund operators have taken steps to ensure it.

^2/Money market mutual funds in the United States invest in safe, liquid assets that reduce credit and interest-rate risk.

^3/Investing in safe liquid assets makes the fund less opaque. In addition, there are strict disclosure requirements in the United States.

(1) The definition of a deposit

While there is some ambiguity as to what constitutes a deposit, its defining characteristic is, in a noninflationary environment, the promise to repay principal and interest that are fixed in nominal terms. ^1/ Under inflation, unless interest rates rise correspondingly, such arrangements impose losses on depositors so the contract may be adjusted to promise to repay an amount that is fixed in real terms by, for example, linking the principal to a price index.

(2) The characteristics of deposits

The bank has to meet a request for repayment from a current account or demand deposit immediately. ^2/ In principle, a bank has time to adjust to the demand for repayment of a savings deposit, although the requirement for notice of withdrawal is frequently waived in practice. While the date for repayment of both regular time deposits and negotiable CDs is fixed by contract, the depositor can sometimes gain earlier access to his funds by paying an early withdrawal penalty. ^1/ There is no secondary market for deposits (except negotiable CDs) and requests for deposit withdrawals are handled sequentially (Diamond and Dybvig, 1983). Moreover, the law in some countries prevents banks from pledging particular assets (except to the central bank) in order to borrow funds. Thus, depositors have an incentive to run if they doubt the solvency and liquidity of their bank.

Thus, demand deposits have a challenging configuration (fixed principal, immediate withdrawal, sequential servicing, backed by nontransparent, risky assets that often lack secondary markets). This configuration is not replicated by any other instrument, although interbank deposits almost do so. Owners of open-end and money market mutual funds (MMMFs) share some of depositors’ incentives to run, however. If bad news breaks, mutual fund owners may want to be first to sell their shares back to their fund, before the share price falls. For this reason, MMMFs and their parent companies go to considerable lengths to avoid “breaking the buck,” i.e., allowing the redemption value of shares in their fund to decline. ^2/

(3) The risks attendant on deposits

While banks always face numerous economic risks, in particular, credit, interest rate, liquidity, and market risks, ^3/ they also face risks from high leverage, operational disruption, changes in exchange rates and in the legal and administrative environment. Credit risk in bank assets affects demand, savings, time, government, interbank and foreign exchange deposits equally. While idiosyncratic credit risk can be tempered by diversification across different regions of the country and types of assets, diversification will not protect against systematic risk. ^4/ Interest rate risk varies with the maturity/duration of a fixed-rate deposit, but can be handled in several ways including offering variable deposit rates.

Liquidity risk arises from instantly withdrawable liabilities that fund longer-term assets, especially where the assets are opaque and have no secondary market. It differs substantially among the deposit types, being greatest for demand and interbank deposits. Banks typically learn to forecast and manage their inflows and outflows of funds in normal times, but they may be unable to meet requests for withdrawals during a banking crisis if liquidity problems develop systemic proportions. Market risk, arising from changes in the prices of bank assets and liabilities, is particularly strong if banks speculate rather than hedge in the derivatives markets.

Leverage risk, which is typically higher for banks than for other financial institutions and enterprises, applies to all liabilities equally. Operational risks that occur particularly in the payments system can make a bank illiquid. ^1/ Banks also face risks that changes in exchange rates will harm the values of their assets or liabilities or the ability of their borrowers to repay their loans. They may also be harmed by changes in the legal, political, or regulatory framework in which they operate. On the other hand, MMMFs face fewer risks in both their assets and liabilities. Narrow banks, which invest only in short-term, safe (e.g., government) securities have the same liability structure but have more transparent assets that largely avoid credit and interest rate risks. ^2/ Other nonbank institutions offer instruments that face similar risks to banks on the asset side but they have less demanding liabilities.

(4) Ensuring the safety of deposits

In general, the public’s nondeposit financial assets are not guaranteed against market risks. ^3/ Even debt contracts that share many characteristics with deposits are not usually formally guaranteed. ^4/ Some institutions (some Islamic banks, bond and equity funds, and even MMMFs) that lack a lender of last resort or system of deposit insurance, meet the liquidity challenge arising from the right to immediate withdrawal by not promising to repay a fixed amount of principal. Nevertheless, the public values the certainty and the convenience of fixed principal, so that banks that offer deposits and MMMFs try to avoid “breaking the buck.” A fixed-principal contract is particularly important for transaction balances. Keeping track of funds available for meeting payments obligations would be difficult absent fixed principal. One could envision the cascading repercussions of a payor issuing a payment instruction at a time when the value of his account was sufficient, only to default because that value had declined before settlement for reasons beyond his control. Consequently, even mutual funds that offer payments services settle their customers’ obligations through the deposits they hold at commercial banks.

The “narrow bank” concept in economies that have a deep secondary market enables a bank to offer demand deposits, while limiting credit, interest rate, and liquidity risk. (See Table 2.) A narrow bank has less need for insurance and should experience fewer failures and impose smaller demands on any insurance fund. The premium that a narrow bank would need to pay, should therefore, be lower. However, as a narrow bank forgoes many of the efficiencies of bank intermediation, a second, less radical suggestion is to require a commercial bank to protect its transaction deposits by fully or partially collateralizing them with safe, liquid securities.

(5) The systemic implications of deposit losses

Losses on any of the financial instruments listed in Table 2 (except the government’s deposits) can give rise to negative externalities where the spillover from the demise of one institution causes others to default. Spillover comes in two forms: domino and contagion effects that present a greater problem for banks than for other financial institutions.

Domino effects occur when those who have suffered losses in the initial failure, default on their personal or business obligations, which leads to a chain reaction in which other institutions fail and cause their creditors to default. While losses on any financial instrument (unless owned by foreigners or the government) have wealth effects that could contribute to a recession, domino effects are particularly problematic in banking where losses on current accounts and interbank deposits can immediately cause others to default. (Losses on MMMFs are also problematic because these are often readily transferable to current accounts for transactions purposes.)

If the public cannot distinguish good banks from bad banks it may avoid all banks, or all similar and adjacent banks indiscriminately when one bank fails. The practical incidence of such contagion is disputed, however. While Kaufman (1994) argues that fears of contagion are exaggerated, research by Pozdena (1991) has shown, at least in the United States, that banks are more prone to contagion than nonbanks; contagion affects large and adjacent banks in particular; and increases in times of uncertainty. Pozdena hypothesizes, but does not demonstrate, that the danger of contagion is greater in the absence of deposit protection.

While the authorities in many countries may be over-cautious with regard to banks’ negative externalities, their statements and actions suggest that they are concerned that spillover will lead to contagious bank failures and recession and that they believe deposit protection can alleviate these dangers. For these reasons, banks typically obtain assurances while other institutions do not.

4. Summary

Banks are essential to the economy, but their deposit contract and the opaqueness of their assets make them particularly vulnerable to insolvency and illiquidity. No other financial institution offers this configuration of characteristics—narrow banks forego risky assets, mutual funds eschew fixed repayment, MMMFs avoid both, equities do not promise fixed returns; neither equities or bonds are payable on demand and both are marketable. Consequently, there is more reason to guarantee bank deposits, particularly demand deposits and to protect the institutions that supply them. ^1/

1. Avoiding/resolving a crisis

Governments often adopt a DIS to avoid or resolve a crisis, because deposit protection provides banks with a core of stable small deposits. Nevertheless, the DIS’s efficacy in this regard is ultimately limited. Further, while the government has a legitimate interest in the safety and soundness of the banking system, there are costs associated with maintaining or restoring bank soundness, particularly systemic soundness, so the costs have to be weighed against the benefits. The EU apparently has made such an assessment and in the preamble to its 1994 Directive has declared:

“Whereas the cost to credit institutions of participating in a guarantee scheme bears no relation to the cost that would result from a massive withdrawal of bank deposits not only from a credit institution in difficulties but also from healthy institutions following a loss of depositor confidence in the soundness of the banking system.” (EU, 1994, No. 1, 135/5.)

This statement acknowledges a trade-off between the cost of providing protection and the losses that would be experienced during a banking panic. The EU asserts that the costs of deposit protection are small in relation to the costs of financial crisis and therefore are worthwhile.

The costs of deposit insurance are of two kinds. As discussed further in the pitfalls section below, they involve both the direct expenditure of funds and the indirect but adverse effects that deposit guarantees have on the incentives that bankers, their creditors and others face. The outcome of the discussion below is that a pending crisis cannot be entirely avoided or an existing one resolved without offering a full guarantee of deposits but that the fiscal and incentive costs of such a full guarantee can be prohibitive. ^1/

2. Protecting small depositors’ funds

Perhaps the most common objective for deposit insurance that is achievable is consumer protection. Given the opaqueness of bank assets, it is difficult for bank customers (especially “widows and orphans” as opposed to “investors”) to monitor the condition of their bank and it may be an inefficient use of their resources to expect them to do so. Thus, governments frequently protect small transactions and savings deposits. Bartholomew and Vanderhoff (1991), Kyei (1995), Lindgren and Garcia (1996), and McCarthy (1980) show that explicit deposit insurance schemes typically place limits on the amount of deposit that are covered. ^2/ These limits range from only $18 in Lebanon to $100,000 in the United States and even higher in Italy. Limited guarantees serve to protect the retail payments system and reduce the negative wealth effects of bank crises for consumers. Some countries that have faced banking crises recently (Finland, Japan, Kuwait, Mexico, Sweden and Turkey) have guaranteed deposits in full, at least for a period of time. Sweden ended full coverage at the beginning of 1996.

The analysis in Section III.2 above suggests that savings and time deposits are conceptually easier for banks to manage than demand deposits, so there would appear to be less need to guarantee savings deposits than transactions balances. Nevertheless, Kyei (1995) shows that all the countries surveyed at least partially guarantee savings deposits. Doing so provides a degree of consumer protection, encourages a positive attitude toward saving, changes social behavior and helps to create a middle class and, as discussed later in this section, raises the national savings rate.

3. Reducing the impact of recession

Authorities recognize that losses incurred in the financial sector reduce wealth and expenditure; if widespread, they can cause a recession. ^1/ It is wishful thinking to believe that a DIS will avoid the occurrence of recession, however, partly because bank failures typically occur after a recession has started. But it can attenuate the severity of recession by providing insured depositors with speedy access to their funds. In addition, the guarantees would need to be comprehensive to avoid wealth effects entirely and comprehensive guarantees are contra-indicated. Nevertheless, even partial protection mitigates recessionary effects.

4. Guarding the payments system

It has been argued above that the availability of a deposit contract is important to facilitate making payments. First, even in a cash-based economy, bank deposits are useful for replenishing cash balances on hand and for storing funds in excess of immediate needs. Second, demand deposits are essential if the payments system is to evolve beyond a system of settlement by cash. Checks, credit cards and electronic transfers are all settled by adjustments to the payor’s and payee’s current account balances, but in many countries transactions accounts, by law, can only be offered by banks. As discussed above, the obligation to service and honor demand deposits contracts is an onerous one, so a LOLR and DIS are valuable in protecting the integrity of the payments system. In fact, Chile protects current accounts in full, but coinsures a limited amount of other deposits. ^2/

Nevertheless, a DIS with limited coverage will not guarantee the wholesale payments system (Lindgren and Garcia, 1996). In fact, 24 countries exclude interbank deposits completely from the system of deposit protection (Kyei, 1995; Lindgren and Garcia, 1996).

5. Enhancing competition among banks

In the absence of an explicit system of deposit protection, depositors are likely to prefer using large banks that are perceived to be less risky because they: (1) have more diversified portfolios than small banks or (2) are protected by an implicit government guarantee because they are “too big to fail (TBTF).” Explicit insurance will help small banks to prosper in these circumstances and it will also facilitate the creation of de novo banks, which tend to be small. The authorities hope that additional competition in the banking industry will improve financial efficiency, lower interest rates and spreads and foster investment and economic development.

A system of limited protection will also serve to protect the franchise value of those banks that serve mainly retail customers by reducing their depositors’ incentives to run. Such protection will help them to avoid default and increase the likelihood that someone will purchase the franchise if they do fail.

Insurance will help private banks in their competition with state-owned banks that have explicit or implicit guarantees from the government. It will also assist domestic banks that are competing with foreign banks that import insurance coverage from their home country. In addition, it may allow domestic banks that carry their domestic insurance with them to gain a foothold in a foreign country. Such possibilities can lead to competition through deposit insurance escalation. Consequently, Article 4 of the EU Directive includes a specific requirement aimed at preventing this occurrence. It states that an EU member shall protect deposits in branches of banks set up in other member countries but that, through 1999, neither the level nor the scope of coverage offered shall exceed that offered by the host country’s scheme. But these rules do not apply outside the EU.

6. Providing a formal mechanism for resolving failed banks

Where a country lacks laws that are designed explicitly to govern the demise of failed banks, it may use the establishment of a formal DIS as an immediate opportunity to put the missing laws into place to hasten the disposal of failed banks. ^1/ Banks cannot operate under bankruptcy protection, so speed is necessary in handling failed banks. Because insolvent banks have incentives to gamble for recovery, they often deteriorate rapidly. Consequently, bank closure laws need to be separate from the laws that govern the bankruptcy of other firms so that they can be dealt with more quickly. In addition, both theory and experience have shown that delay in closing insolvent banks raises resolution costs; that regulators tend to favor delay; and that some politicians encourage it. Consequently, the bank closure statutes should be constructed to limit regulatory discretion and to mandate closure and prompt corrective action (Carnell, 1993).

7. Reducing the government’s obligations and getting banks to contribute

In countries that have no formal DIS or where the guarantee is implicit, the cost of any attempt to protect deposits falls on the government directly or indirectly through the central bank. In this situation, instituting an explicit DIS can reduce the government’s obligation by limiting coverage and providing a formal mechanism that forces banks to meet at least the major part of the cost of deposit protection. If the scheme is well-designed, if macro-economic policy successfully protects the economy from shocks, and if failed banks exit promptly, losses to the system should be light enough for the banking system to absorb them completely. Only in unforeseen emergencies should the government need to supplement bank resources to resolve failed banks. ^1/

8. Helping to justify bank legislation, regulation, and supervision

For a variety of reasons, legislators are often unwilling to enact necessary banking laws, and supervisors are hesitant to impose and enforce prudential regulations and to supervise banks’ condition (Kane, 1989). However, the prospect of the government’s having to cover the costs of an underfunded insurance scheme and face the consequent taxpayer anger may provide the impetus for legislators and supervisors to pay greater attention to serving the taxpayers’ interests by keeping the banking system sound.

9. Promoting economic growth

Deposit protection can encourage saving and reduce the risk premium that depositors require on their funds, which enables banks to charge lower loan rates, and so stimulates investment. The greater competition that a DIS facilitates also encourages banks to reduce the spread between their deposit and loan rates. Thus, a DIS can promote saving, intermediation and investment, reduce both interest rate levels and spreads and foster economic growth. ^2/

10. Subsidizing preferred industries

Some nations utilize under-priced deposit protection as a means to eliminate or reduce the risk premium paid to depositors and rates charged for loans by banks, often development banks, that are devoted to serving preferred industries. The savings and loan industry in the United States, for example, was initially designated to provide housing finance and still has its own separate system of deposit insurance. Its DIS was under-priced for many years, failed, and had to be back-stopped by the United States Government.

11. Summary: on consistency and feasibility

The objectives that countries select will most probably influence the design of the DIS they adopt. Nevertheless, some objectives are more readily achieved than others. Countries may, with justification and consistency, adopt a DIS to protect bank customers’ funds, help to stabilize the financial system, reduce some of the adverse consequences of recession by giving depositors quick and certain access to their insured funds, protect the retail payments system, enhance bank competition, provide a formal mechanism for resolving failed banks, and draw on bank resources to defray the cost of bank failures. These favorable results may stimulate economic growth. Subsidizing special industries is, however, likely to conflict with other objectives. Moreover, where stability is at stake, countries may extend the guarantees beyond small deposits and so establish expectations of repeated rescues that reduce the incentives for owners, managers and depositors to keep the system sound. Thus, as discussed in Section II, while a comprehensive guarantee may bring stability temporarily, it can weaken the banking system and actually increase instability in the long run.

1. Incentives

In the life, health, property, and casualty branches of the insurance industry, the policy-holder knows his exposure better than the insurer. This can lead to adverse selection and heavy losses. ^1/ The insurance company reacts by taking steps to limit its costs by choosing the risks and the risk-takers that it will guarantee and charging risk-adjusted premiums to deter adverse selection. The information asymmetry can also promote moral hazard that the insurer discourages by giving the insured incentives to avoid taking risks. In addition, the insurer changes its premiums over time so that it can cover its expected outlays, build its capital reserves and pay dividends. It also transfers some of its remaining exposure to other parties through reinsurance and coinsurance.

Asymmetric information is also a problem in banking, where the guarantee gives rise to particularly strong problems of adverse selection and moral hazard. Moreover, since deposit insurance is similar to a put option that involves three parties—depositors, their bank and their guarantor—it can alter the behavior of bank owners, managers and depositors. It also provides opportunities for other interested parties (such as supervisors and politicians) to take advantage of its protection.

While the deposit guarantor can take some of the protective measures adopted by the insurance industry, other safeguards are not available to it. For example, while it can set prices to provide constructive incentives, the deposit insurer often has no choice over membership and may also be subject to political interference.

b. Deposit insurance as a put option

While the intuitive idea is that deposit insurance protects depositors, recognizing that deposit insurance involves three parties suggest the ways in which bank owners and managers benefit from its operations. ^3/ When a bank in an uninsured banking system gets into difficulties, its owners have to decide whether to continue to operate the bank in a weak condition and risk a run, insolvency and default; recapitalize it; voluntarily wind it up; or terminate its operations by defaulting on its obligations.

Deposit protection gives the owners a fifth alternative—one that can disrupt healthy internal governance. By obtaining a guarantee, the owners purchase the right to surrender the remaining assets and the charter of the bank to the insurance agency in exchange for payments to insured depositors, which is equivalent to purchasing a put option. ^1/ To enhance their chances of recovery, informed but uninsured creditors normally force the closure of a bank when it becomes market-value insolvent. Regulators typically do not close it until later when it is book-value insolvent. An insured bank’s owners would not choose to exercise their option until an even later point, when the option has become “in the money” for them. That point occurs when the market value of the asset portfolio plus the value of the insurance option and any residual value of the bank’s charter have fallen below the face value of the liabilities. (See Figure 2 below.)

View Full Size

Ratio of the Market Value of Assets to Liabilities

Citation: IMF Working Papers 1996, 083; 10.5089/9781451850499.001.A001

• Download Figure
• Download figure as PowerPoint slide

Option pricing models have been widely used to estimate the value of deposit insurance to a bank owner (Marcus and Shaked, 1984; Merton, 1977 and 1978; McCulloch, 1981; Pennachi 1987; Pyle, 1984 and 1986; Ronn and Verna, 1986; and Thomson, 1987 a and b). The owners purchase the put by paying the premium that the insurance agency charges for writing the option. The value of this option has been shown to rise with: (1) the risk-free interest rate, (2) the amount of risk undertaken by the bank; (3) the extent to which the option is in the money, that is, the less capital the bank holds; and (4) the time until the expiration of the option (Saunders, 1994). ^2/

2. Mispriced deposit insurance

Systematically under-priced insurance can have adverse macroeconomic implications. Banks obtain arbitrage profits by increasing their deposits (whose risk and interest rates are reduced by the guarantee) and expanding their loans beyond the socially optimal level. When the overexpansion can no longer be sustained, the subsequent recession and decline in asset prices will be sharper than would have occurred without the deposit guarantee. This problem is hard to limit by regulation and supervision, especially where promoting savings and stimulating the economy are objectives of deposit insurance.

3. The cost of insurance

A DIS that is poorly designed can be expensive. By insuring the worst banks, encouraging risky behavior, postponing the need for effective regulation, supervision and information dissemination, and delaying exit, a DIS can cause the condition of the banking system to deteriorate and the cost of its ultimate restoration to escalate.

There are numerous examples of poorly designed DIS that have allowed systemic deterioration in the banking system. Perhaps the best known example is the S&L debacle in the United States, where depositors and other creditors were so completely protected that they exerted no discipline; the industry’s top supervisors guarded their careers by passing problems unsolved to their successors; owners and managers of insolvent institutions looted or gambled and lost; poor credit risks obtained funding (for windmill farms, exotic resorts, and “see-through” office buildings) to the detriment of more socially useful projects; the industry’s special lender of last resort subsidized insolvent institutions, ^1/ politicians interfered to delay needed remedial actions; and the taxpayer ultimately expended approximately $130 billion to resolve the situation. ^2/

4. Conclusion

Many—healthy banks, depositors, other creditors, the government, central bank, suppliers, consumers, and taxpayers—stand to lose from a banking system that is operating without regulatory or market discipline. Deposit protection that weakens internal governance, market discipline and official oversight can distort the financial system. It may have contributed to the severity and cost of crises in the United States and several other countries where the supervisor or the insurer did not exercise its call option to close the bank in a timely manner.

With many parties exploiting the deposit guarantee, it is possible that the demand on a privately funded guarantor will exceed its resources and it will fail. Facing insolvency, it may have no choice but to forbear, adding to the ultimate cost of resolution until it can be recapitalized by government and resume closing failed banks. Thus, it is important, for the insurer to manage its operations, in ways described in the following section, to avoid this catastrophe and for the government to back up the fund when a systemic crisis places unanticipated demands on the fund.

1. An implicit or explicit scheme?

Establishing a formal, legally binding DIS, allows the authorities a structure to avoid many or most of the pitfalls, while obtaining the benefits. Implicit systems—arising in public expectations based on the authorities’ public statements or past actions in protecting state-owned banks, guaranteeing their loans, and intervening elsewhere to protect depositors—are low on benefits and high on pitfalls.

2. Pre-conditions for successful operation

Establishing a successful DIS involves considering the structure of the banking system, ensuring that certain essential legal and supervisory prerequisites are in place and allocating financial responsibility for it.

3. Starting up the system

Getting a protection scheme up and running requires initial funding and skilled staff. (Regular funding is discussed below.) There are essentially four ways to obtain the initial resources for a funded scheme: (1) place a start-up levy solely on the banks; (2) share the levy jointly among the commercial banks, the central bank and the Treasury; (3) hold the government alone responsible for meeting the initial financial needs; or (4) start the system without accumulated funds but grant authority for the scheme to borrow to meet its needs and repay the borrowing through an emergency ex post levy if a bank fails.

A decision also has to be made whether official funding will be permanent or is to be repaid by the banks over time. Starting with funds that the public perceives to be insufficient will not win confidence and risks the fund becoming insolvent. Insolvent funds are prone to forbearance, costly forms of resolution, and crises (as was the Federal Savings and Loan Insurance Corporation in the United States in the 1980s).

At first, staff may be borrowed from the central bank which is likely to have a pool of financial sector experience to draw upon. Later, the insurer can train its own staff and they will mature with experience over time.

4. Running the system in normal times

Utilizing the insurance industry’s techniques suggests that the scheme underwrite, control, and transfer part of its risk exposure.

b. Controlling risk

The deposit protection scheme can control the risks it faces by offering inducements to those insured to limit their risk-taking. For example, it can restrict portfolio risk by adopting the requiring banks to hold safe assets as does a narrow bank, or it can limit its obligations by instituting a system of depositor or insurer preference, demanding that insured banks meet high capital standards, levying risk-based premiums, promulgating and enforcing strong regulations to constrain risk exposure, imposing effective penalties for non-compliance with regulations, mandating prompt corrective action and closure, supervising and monitoring banks’ condition, publicly revealing as much information as possible, and imposing a “haircut,” or small levy, on all depositors and creditors if a bank fails. ^2/

These actions result in incentives for market discipline through owners, managers, depositors, other creditors, supervisors and politicians. Such discipline requires imposing losses without fail on those (honest owners, boards of directors, senior management, as well as those acting illegally) that are directly responsible for bank failures and ensuring that shareholders are first in line to lose their claims completely before losses are imposed on other claimants. Members of the board of directors and senior managers should lose their jobs if they were at fault in the failure. Those involved in criminal activities should be punished by fine and/or imprisonment and barred from the industry in future. Systems of accountability need to be established to keep supervisors and politicians devoted to serving the public interest.

(1) Constraining or collateralizing risks

Laws and regulations can restrict the types of assets and liabilities that insured banks hold and prescribe or restrain the activities that banks undertake, but doing so may reduce competition in the financial services industry in a way that becomes increasingly counter-productive to enforce when technological change is rapidly breaking down old barriers between different segments of the industry. Supervisors can monitor compliance with the restrictions and banks’ condition. ^1/ As discussed in the conclusion, however, there is a trade-off between constraining banks’ portfolio composition for safety and permitting them to operate innovatively and efficiently.

Alternatively, a commercial bank can be required to collateralize its insured deposits with riskless securities in a form of narrow banking. ^2/ A third form of restriction would be to impose a 100 percent marginal reserve requirement on under-capitalized but solvent banks to restrain their growth and gambling behavior while they recapitalize and to help lower failure costs if they fail to do so. A fourth option is to limit the interest rates that under-capitalized banks can pay to prevent them from growing fast and gambling. For example, Argentina’s DIS does not cover deposits that pay rates that exceed a reference rate plus a certain percentage.

(2) Limiting exposure through depositor and insurer preference

Systems of priority where there is no DIS are discussed before priorities under systems that provide deposit protection.

(a) Priorities when there is no deposit insurance

A country without a DIS can provide its depositors with some protection from loss by giving them priority over the assets that are recovered from a failed bank. As shown in Table 3, many countries give collateralized claims the first priority in law (to the extent of the value of the collateral) and liquidation/receivership expenses have second priority. Often, all creditors, including both depositors and other general creditors, share the next priority equally. Thus, where the central bank’s claims are collateralized, it has a good chance of recovering the funds it has lent. Subordinated debtors stand fourth in line after these other claims have been met in full. In the unlikely event that anything is left after that, it belongs to the equity holders. This system of priorities is laid out in column 1 of Table 3. A system of depositor preference, in column 2 of the table, protects depositors by allowing them to recover their claims in full before the remaining claimants obtain anything. Uncollateralized claims by employees, the central bank, and the tax authorities receive different priorities in different countries and are therefore omitted from Table 3. ^1/

Table 3.

The Order of Priorities Over Receivership Assets Without Deposit Insurance

Table 3.

The Order of Priorities Over Receivership Assets Without Deposit Insurance

No Preference for Depositors			Depositor Preference
1.	Collateralized claims		1.	Collateralized claims
2.	Receivership expenses		2.	Receivership expenses
3.	Senior claims, equally:		3.	Depositor claims in full
		deposit claims and
		general-creditor claims
			4.	General creditor claims
4.	Subordinated claims		5.	Subordinated claims
5.	Equity		6.	Equity

View Table

Table 3.

The Order of Priorities Over Receivership Assets Without Deposit Insurance

No Preference for Depositors			Depositor Preference
1.	Collateralized claims		1.	Collateralized claims
2.	Receivership expenses		2.	Receivership expenses
3.	Senior claims, equally:		3.	Depositor claims in full
		deposit claims and
		general-creditor claims
			4.	General creditor claims
4.	Subordinated claims		5.	Subordinated claims
5.	Equity		6.	Equity

(b) Priorities when there is deposit insurance

Priorities can be established in law to reduce the demands that are placed on a deposit insurance fund. Systems of depositor preference, insurer preference, or a combination of both can encourage market discipline and limit fund outlays. After collateralized claims and receivership expenses, a country might allocate pro rata shares of the recovery pool to all types of creditors without preference as in column 1 of Table 4. ^2/

Table 4.

The Order of Priorities Over Receivership Assets With Deposit Insurance ^1/

Table 4.

The Order of Priorities Over Receivership Assets With Deposit Insurance ^1/

No Preference			Depositor Preference			Insurer Preference
1.	Senior claims, pro rata:		1.	Depositor claims, pro rata:		1.	Insured deposits
		insured deposits,			insured deposits and
		uninsured deposits and			uninsured deposits
		general creditor claims
			2.	General creditor claims		2.	Other senior claims by:
								insured depositors and
								general creditors
2.	Subordinated claims		3.	Subordinated claims		3.	Subordinated claims
3.	Equity		4.	Equity		4.	Equity

View Table

Table 4.

The Order of Priorities Over Receivership Assets With Deposit Insurance ^1/

No Preference			Depositor Preference			Insurer Preference
1.	Senior claims, pro rata:		1.	Depositor claims, pro rata:		1.	Insured deposits
		insured deposits,			insured deposits and
		uninsured deposits and			uninsured deposits
		general creditor claims
			2.	General creditor claims		2.	Other senior claims by:
								insured depositors and
								general creditors
2.	Subordinated claims		3.	Subordinated claims		3.	Subordinated claims
3.	Equity		4.	Equity		4.	Equity

The second column shows how priorities change when depositors (both insured and uninsured) receive preference over other claimants. The third column illustrates the effects of adopting insurer preference, which ensures that the insurance agency recovers its outlays to insured depositors before noninsured claimants are paid from the remaining pool. The Appendix develops numerical examples that show that a system with depositor preference imposes fewer financial demands on the fund than a system without it and that insurer preference is less costly still to the fund. Naturally, nondeposit creditors lose correspondingly more in each case.

(3) Using risk-based premiums

Adopting risk-based insurance pricing in sophisticated financial systems attempts to make bankers pay the social cost of excessive risk taking and compensate the fund for the expected outlays they require from it. Argentina, Bulgaria, and the United States levy risk-responsive premiums. Hungary levies supplementary premiums on risky banks and Sweden plans to do so.

A risk-based premium will reduce but not eliminate adverse selection or moral hazard. Adverse selection will remain as long as the premium cannot be sufficiently fine-tuned to eliminate any subsidy to riskier banks. Moral hazard remains when the bank is able to increase its risk exposure after the premium is set. But as the premium is paid on a regular basis, the bank can be monitored frequently so that its premiums adjust promptly when its risk exposure changes to reduce moral hazard. In addition, it can be discouraged by an ex post settling up, so that any bank that increases its risk within the period will have to pay an additional premium at the end of the period (Berlin, Saunders and Udell, 1991). An alternative possibility, is for the bank to post a surety bond to guarantee its risk exposure (Osterley and Thomson, 1991). Market value accounting, risk-based schedules that cause bad banks to be examined on-site sooner and more frequently than good banks and periodic relicensing have all been suggested as ways to reduce the residual moral hazard of deposit insurance even when risk-based premiums are in place.

There are, however, practical difficulties in establishing risk-based premiums. On the one hand, a risk-seeking bank may still be able to disguise its risk-taking. On the other hand, fine tuning premiums gives regulators opportunities to behave arbitrarily, so that the criteria on which decisions are made have to be publicly available and acceptable, which means that they will have to be approximate. ^1/

(4) Setting high capital requirements

As discussed above, deposit insurance, if not appropriately priced to reflect risk, encourages and enables bankers to reduce capital and increase the probability and cost of failure. This incentive needs to be countered by requiring banks to hold substantial amounts of capital, which acts like “a deductible” in an insurance contract and by restricting dividend and other payments to discourage looting and give owners and managers incentives not to exploit the guarantee by taking excessive risks.

(5) Adopting incentive-compatible forms of failure resolution

The method chosen to resolve a failed bank can alter the risk exposure of a failed bank’s stakeholders. For example, if a bank is put into receivership and liquidated, the rules limiting insurance coverage and imposing losses on owners, managers and uninsured depositors can readily be observed. However, if a failed bank is merged (voluntarily or forcibly), the rules may be bypassed, so that these parties lose less than in a liquidation and discipline is thereby reduced. ^2/

The legislature should require the insurer to use the least costly form of resolution to avoid protecting owners, managers and uninsured creditors. ^3/ That cost should be as assessed by the insurer at the time of closure.

(6) Releasing reliable information

It behooves the deposit insurance agency to ensure that the public has access to reliable information, provided publicly or privately, on the condition of individual banks and of the insurance system. Reliable and timely information facilitates market discipline, prevents bank insolvencies and forces the timely resolution of those that do occur, which is essential to a strong exit policy. Moreover, lacking data, depositors may run unnecessarily.

(7) Taking prompt corrective actions

A supervisory system that discovers and remedies bank deficiencies at an early stage can prevent many bank failures and will promptly address those that do occur. Such policies will keep the banking system strong and competitive and will reduce demands on the fund. This is the rationale behind the prompt corrective action provisions of the FDIC Improvement Act of 1991 (Benston and Kaufman, 1994; Carnell, 1993; and the United States Congress, 1991). Delaying remedial action caused private DIS to fail in the United States (English, 1993).

5. The system in a crisis

An insurable risk is one that is rare and independently distributed—conditions that are not met in banking during a crisis when failures are frequent and related. For example, a system-wide crisis beyond the scope of most preventive measures can arise when investors flee a country whose currency is expected to depreciate. Realism suggests recognizing that privately funded deposit protection schemes are likely to founder in the face of such a systemic event. Calomiris (1993) argues that an underfunded DIS will be ineffective in preventing runs and Brock (1996) demonstrates that such a DIS encourages an escalation of risky behavior and a continuation of high interest rates.

Where the Government has an adequate tax base, its assistance can be effective in avoiding a meltdown of the financial system (Lindgren and Garcia, 1996). ^1/ Consequently, arrangements should be made for supplementary funding when the deposit protection scheme is incapacitated. Two types of incapacity are relevant: fund illiquidity and fund insolvency.

1. Market discipline and financial stability

The DIS needs to be designed to temper those elements of market discipline that are destabilizing and associated with costly financial panics. Yet, at the same time it needs to mimic the markets’ risk-restraining influences on the behavior of bank owners and managers by providing appropriate incentives.

The configuration of such a scheme was discussed in Section VI. It should be explicitly laid out in the law, be compulsory, and be accompanied by strong civil and commercial laws, and well-established systems of accounting, loan valuation, prudential regulation, and supervision. The supervisor needs authority to take prompt actions to correct deficiencies before failure occurs and to deal effectively with insolvent banks. The DIS should be established only after unsound banks have been restructured and should treat large, small, private and state-owned banks equally. The supervisor needs good information to keep the banking system sound and should release as much as possible to the public to enhance market discipline. The DIS should feature limited coverage of all types of deposits, including foreign currency deposits, provide for the prompt repayment of a failed bank’s insured deposits in domestic currency.

The DIS can be privately or publicly run, but in the latter case the administering agency should be politically independent and not have a board of directors dominated by bankers. The agency may or may not be part of the central bank and or supervisory agency.

The DIS can be funded by ex post assessments on banks or by premiums that build into a fund. The size of the fund should be large enough to deal with the failures that occur in normal or stressful times, but may prove insufficient in periods of financial emergency. Where the fund faces insolvency, it should be back-stopped by the government, not the central bank. The central bank might provide temporary liquidity, however, to an illiquid fund that is sure to be able to levy additional premiums on surviving banks to cover its deficiency in the near term. Premiums should ideally be risk-based and for administrative simplicity be assessed on all deposits. Although risk-adjustment is difficult to accomplish fairly, several countries are already adopting risk-based systems to provide banks with incentives to operate soundly. The DIS should invest its resources safely and consider diversifying into safe foreign assets to limit exchange risk.

The DIS can limit its exposure and increase market discipline from sophisticated creditors by utilizing systems of depositor and/or insurer preference over the assets of the failed bank. It can transfer more risk by encouraging banks to mutually guarantee and oversee their competitors, use coinsurance for larger depositors, and purchase reinsurance where it is credibly available.

Many countries find discipline from nondeposit creditors to be a viable form of market discipline but that discipline from depositors, especially from small depositors, is counter-productive because it may cause runs and such runs may spread. If the authorities decide to rely on market discipline, however, it can be introduced by sharply limiting the maximum amount and types of liabilities that are covered, having large depositors proportionately share losses with the fund under coinsurance, and strictly enforcing these conditions when a banks fails and the guarantee is effected. Such a scheme needs to be reinforced by strong systems of law, accounting, regulation and supervision.

2. Regulatory discipline and financial innovation

Strong systems of supervision and regulation are often used to counter insurance’s encouragement to risk taking. There is a disadvantage to heavy reliance on regulatory discipline, however. Over-reliance on regulatory and supervisory discipline stifles innovation in banking and prejudices economic growth. There is a challenge to the authorities, therefore, to find ways to find the right balance between regulatory and market discipline, and to exert the regulatory discipline that is necessary to maintain financial stability while allowing financial innovation to promote economic growth. Approaches for achieving this objective, lie beyond the scope of this paper, however.

Regulatory disciple also requires effective accounting, reporting, regulation, supervision, and enforcement. Adequate capital is an essential component because it provides incentives for owners to restrain risk taking, reassures depositors of the safety of their funds, and protects the insurance fund from loss. Regulatory discipline can be assisted by imposing stringent conditions on the quality of applicants for bank charters (both new and transferred) and considering how to temper the adverse effects of limited liability for owners when their bank fails. ^1/ It also requires that laws, banking practices, and institutions are in place to deal expeditiously and efficiently with the exit of failed banks from the system.

In summary, both market and regulatory discipline are needed to reinforce internal governance, which is the main guardian of stability in the financial system.