_____ ____ ____ _____       ____       ____W                  PS   24+C 3
POLIcY RESEARCH WORKING PAPER   2453
Designing Financial Safety                                              For optimal regulation, one
size does not fit all.
Nets to  Fit Country                                                    Differences in countries'
Circum stances                                                         informational and contracting
environments-in the
transparency of information,
EdwardJ. Kane                                                          in protections for
counterparties, and in political
accountability-influence the
design of their financial safety
nets and their strategies for
managing the breakdown of
those nets.
The World Bank
Development Research Group
Finance
September 2000



I POLICY RESEARCH WORKING PAPER 2453
Summary findings
Kane explains how differences in the informational and        contracts.
contracting environments of countries affect the optimal        The weaker a country's informational, ethical, and
design of their financial safety nets and their optimal       corporate governance environment, the more a wholly
strategies for managing financial crises.                     governmental system of explicit deposit guarantees is apt
He explains how to design and operate safety nets at        to undermine bank safety and stability. How a country's
minimum cost to taxpayers and well-managed banks in           safety net evolves depends on the ability of the private
countries whose informational and contracting                 and public sectors to value banks, discipline risk-taking,
technologies differ. His basic premise is that optimal        and resolve financial difficulties promptly. And political
regulation is not a one-size-fits-all proposition.            accountability is essential if the public part of these tasks
A country's safety net should be transparent, deterrent    is to evolve effectively and efficiently.
to too much risk-taking, and accountable, but Kane              As a rule of thumb, safety-net managers should avoid
shows large differences across countries in the               either subsidizing or taxing bank risk-taking, says Kane.
transparency and deterrence banks afford their                Even if analysts could formulate a beneficial tax or
depositors, highlighting why the design of safety nets        subsidy rule, it is unlikely that channeling the effect
must allow for differences in the enforceability of private    through a government-run deposit insurance system that
fails to account publicly for the size of taxpayers' stake
could improve upon more straightforward arrangements.
This paper-a product of Finance, Development Research Group-is part of a larger effort in the group to design financial
safety nets for developing countries. The study was funded by the Bank's Research Support Budget under the research
project "Deposit Insurance" (RPO 682-90). Copies of this paper are available free from the World Bank, 1818 H Street
NW, Washington, DC 20433. Please contact Kari Labrie, room MC3-456, telephone 202-473-1001, fax 202-522-1155,
email address klabrie@worldbank.org. Policy Research Working Papers are also posted on the Web at www.worldbank.org/
research/workingpapers. The author may be contacted at edward.kane@bc.edu. September 2000. (64 pages)
The Policy Research Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about
development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The
papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this
paper are entirely those of the authors. They do not necessarily represent the view of the World Bank, its Executive Directors, or the
countries they represent.
Produced by the Policy Research Dissemination Center



DESIGNING FINANCIAL SAFETY NETS TO FIT
COUNTRY CIRCUMSTANCES*
Edward J. Kane
Boston Colleze
The author wishes to thank Tolga Sobact for sterling research assistance and Richard Aspinwall, Philip Brock,
Charles Calomiris, Aslt Demirguc-Kunt, Christopher Lamoreux, James Moser, C. Charles Okeahalamn, Joseph
Sinkey and James Thomson for valuable comments on a previous draft of this paper.






Optimal financial regulation is not a one-size-fits-all proposition. This paper seeks to
gather evidence on how differences in the informational and contracting environments of
individual countries have influenced the design of their financial safety nets and their strategies
for managing safety-net breakdowns.
Its larger purpose is to apply the concept of revealed preference to evidence on safety-net
design to frame a menu of principled policy advice that takes reasonable account of differences
in the informational transparency, counterparty protections, and political accountability found in
individual countries.
I. The Safety Net As Metaphor
A net is a mesh that has been rigged to serve a set of particular purposes. In erecting a
literal "safety net," a circus manager has in mind at least two purposes. The direct purpose is to
protect falling acrobats by having the net catch them before they can harm themselves and
traumatize the audience by splattering against the ground. However, the net's larger, indirect
purpose is to enhance audience enjoyment in a positive way. The net does this by making it
economically rational for circus acrobats to undertake difficult, but do-able stunts in which the
danger of a spectacular fall seems very real.
The word do-able is emphasized to clarify that the net is meant to encourage prudent risk-
taking. The net's entertainment function is frustrated unless the acrobats almost always complete
their stunts successfully. Carrying out a simple trick is typically more engaging than failing to
accomplish a complicated one. Trapeze and tightwire performers that fall repeatedly into the net
earn boos and catcalls that sensible circus managers want to avoid.
A circus manager must also make sure that the costs of building and maintaining the net
do not exceed the benefits it produces. Managers must balance a series of subtle tradeoffs
between original costs, monitoring expense, safety, visibility, and audience response. There is
no general reason to suppose that governments should direct net managers either to tax or to
subsidize institutional risk-taking. In general, widespread and sturdy meshes are safer, but they
are more costly and -by being more visible- lessen the sense of risk sharing that lies at the heart
of audience enjoyment. Sturdy filament materials exist whose thinness can make the net less
visible, but these materials may make the net more costly to build and maintain. Finally,
2



whenever the net a manager erects proves unable to stop an unfolding disaster, emergency
medical treatment must be accessed optimally and the manager must expect a storm of
condemnation to rain down on his head from all sides: from performers, from circus owners, and
from the audience.
Whether or not the elements of the net are defined by explicit statutes, authorities in
every country establish a de facto safety net for banks and support the net by incurring
monitoring costs and penalizing to some degree fraudulent and corrupt behavior in financial and
governmental transactions. This paper seeks to explain how to design and operate safety nets at
minimum cost to taxpayers and well-managed banks in countries whose informational and
contracting environments differ in stylized ways.
A country's financial safety net shares four metaphorical entailments with its circus
counterpart. First, like acrobats, financial institutions may choose to engage in activities so risky
that critical mistakes can quickly cripple them or even end their existence. In the metaphor,
borrowers and depositors play the dual role of onlookers and manufacturers of trapezes and
tightwire platforms, while taxpayer audience members also own shares in the circus. Financial
safety nets serve not just to protect borrowers, depositors, and taxpayers from being harmed by
financial-institution mistakes, but also to encourage individual institutions to accept the risks
associated with funding economically productive activities. In countries where reliable public
information about business performance is in short supply, banks tend to dominate the flow of
institutional finance. In such circumstances, a country's financial safety net reduces to its
arrangements for protecting bank customers and for monitoring, assisting, and controlling banks.
Second, just as circus managers do, regulators must guard against excessive risk-taking
and must manage the several costs and benefits the safety net produces. Safety nets for banks
may be instructively conceived as a nexus of contracts that help them cope with runs and other
economic shocks (Diamond and Dybvig, 1983; Kane, 1995; Brock, 1999). Each net is a
multidimensional policy scheme that is alleged to balance the costs and benefits generated by:
1. protecting bank customers from being blindsided by bank insolvencies;
2. limiting aggressive risk-taking by banks;
3. preventing and controlling damage from bank runs;
4. detecting and resolving insolvent banks; and
5. allocating across society whatever losses occur when an insolvent bank is closed.
3



Third, a country's financial safety net may be analyzed as a figurative mesh whose
filaments tie regulatory officials and bank stakeholders into a web of mutually reinforcing
contracts. Officials must be made accountable for any design flaws or breakdowns in stringing
the web that compromise the net's overall effectiveness. For strategies of crisis prevention to be
maximally successful, the web of contracts must hold top regulators accountable for measuring
and managing the social costs and risk-taking incentives generated by their decisions about the
net's various design features. Cross-country variation in the legal and private-informnation
environment should influence these decisions. Ideally, the design of the regulatory portion of net
should tie securely into the characteristics of the particular financial system and economy in
which it is embedded. To fashion a weave of the right size and strength, decisions bearing on the
cost and effectiveness of net components should be observable enough to be disciplined by
appropriate market and budgetary tests.
Finally, authorities must understand that on occasion even the best-designed safety net
will fracture or prove too small. This means authorities must develop and regularly review
strategic plans for managing financial crises and train their staff in the use of crisis-management
protocols. Paradoxically, unless the safety net is backed up by solid crisis planning,
improvements in the safety net may result in less frequent but more devastating crises. On
balance, the more effective a nation's safety net becomes, the less likely it is that regulatory
personnel will have prior hands-on experience in coping with severe crisis pressures.
Examining how and where the circus metaphor breaks down is also instructive. The rub
is that --unlike the splattering of an unlucky, incompetent, or well-connected acrobat--
breakdowns in financial safety nets are not imunediately visible to the naked eye. This
indiscemibility creates opportunities for incumbent regulators to conceal and sugarcoat
infornation about difficulties that occur during their particular terms in office. Regulators can
strongly influence the character of information that may and may not be entered on bank balance
sheets and income statements. Precisely because top regulators do not want to see their
professional reputation besmirched by being held responsible for a financial-institution disaster,
their ability to influence reporting protocols can be a mixed blessing. Defects in transparency
support an incentive to delay insolvency resolution, so as to allow individual banks whose
insolvency is not yet widely recognized an opportumity to gamble for resurrection at taxpayer and
competitor expense.
4



To the extent that regulators can block the flow of adverse information, their sensitivity
to public criticism tempts managers and owners of insolvent institutions to reinforce this
dangerous incentive conflict. They offer subtle and unsubtle benefits to top regulators who are
willing to conceal an incipient crisis and to accept the social costs of postponing the strong
public actions that may be needed to restore banking solvency. By helping to hide developing
insolvencies, top regulators may reasonably hope to achieve a reputationally "clean" getaway.
Most of the blame for allowing a crisis to develop can be sidestepped by patching whatever signs
of crisis surface on their watch and shifting responsibility for recapitalizing the damaged banking
system to succeeding generations of regulatory officials. This hope supports efforts to downplay
the importance of crisis planning and to cornmit regulators to work through every incipient
breakdown as if it were a unique event that they had to handle in an ad hoc manner.
II. Contracting Difficulties that Make a Safety Net Socially Beneficial
Modem finance theory emphasizes that, even in a corruption-resistant society, bank
depositors must worry about controlling incentives for opportunistic behavior by managers,
owners, and borrowers (Jensen and Meckling, 1976; Diamond, 1984). Besides the difficulties
depositors face in coordinating collective action, thesc incentives have two intertwined roots:
1. difficulties a depositor faces in obtaining reliable information about unfavorable
developments and observing adverse actions by bank managers, including
recklessness, negligence, incompetence, fraud, and self-dealing (monitoring costs);
2. difficuities a depositor faces in adequately analyzing and responding to whatever
information their monitoring activity turns up (policing costs).
The tools of regulatory loss control are rulemaking and enforcement. To understand the
role played by a country's financial safety net, it is helpful to imagine a world in which
depositors' monitoring and policing costs would be uniformly zero. In this world, each deposit
contract would be self-enforcing. Establishing a team of centralized monitors and enforcers to
thwart misconduct by bank insiders would offer no benefit either to banks or to their depositors.
In such a world, changes in a bank's condition and risk exposure would be transparent to
depositors and depositors would possess sufficient expertise and sanctions to deter bank insiders
from trying to take advantage of them. Maximal transparency describes a framework of
disclosure that would perfectly and costlessly inform depositors about changes in bank
performance and risk-taking activities. To provide a pair of parallel rhyming words, we use
5



maximal deterrency to describe a situation in which depositors would immediately understand
the implications of information flows perfectly and would be able to protect themselves
completely and costlessly from whatever threat to their wealth this information might reveal.
The more closely an economy comes to offering creditors maximal transparency (MT)
and maximal deterrency (MD), the less incremental value that banks and safety-net managers can
create for depositors. In an MTMD economy, cash in advance and credit could substitute
perfectly for each other in every payment context. Similarly, direct and indirect finance would
provide equally economical ways of mobilizing savings, of choosing which real investment
projects savers ought to support and of deciding how to price project risk. As envisaged in the
Capital Asset Pricing Model, corporate and government securities could be offered in
denominations small enough to allow virtually every individual saver to invest directly in a
diversified portfolio of stocks, bonds, and derivative securities.
The MTMD thought experiment clarifies that safety nets owe their existence to
difficulties of contract enforcement: blockages in information flows; differences in monitoring
costs; variation in financial transaction costs; delays in appreciating and processing relevant
information; and the costliness and inadequacy of the deterrent remedies that individual
depositors have available to them. It also clarifies that a safety net entails a five-party contract.
The net imposes mutual rights and duties on: bankers, borrowers, depositors, safety-net
managers, and safety-net owners (principally healthy banks and taxpayers). The touchstone by
which to judge the performance of safety-net managers is the fairness with which they treat each
of their counterparties and the efficiency with which they manage the diverse social costs of
coping with divergences from MT and MD conditions.
All real-world economies establish a framework of centralized bank monitoring and
deterrent response. Centralizing these functions aims at increasing depositor confidence while
solving three coordination problems: avoiding redundant monitoring expense; standardizing
contracting protocols; and timing and calibrating disciplinary action. In principle, a centralized
monitor-enforcer makes it unprofitable for banks to misrepresent their economic condition to
depositors and to pursue profit-making opportunities that might exploit depositors' informational
disadvantage. It is not enough for safety-net managers to aim at blocking corrupt and unwise
flows of institutional credit and avoiding depositor runs. They must seek also to minimize the
social damage caused by temporary bank illiquidity and by lasting bank insolvencies. In
administering lender-of-last-resort facilities, safety-net managers are expected to perform the
6



financial tiage function of shielding solvent, but illiquid institutions from having to sell assets
into momentarily disorderly markets.
In practice, a safety-net manager must have the expertise to wield six categories of
regulatory instruments fairly and efficiently:
1. record-keeping and disclosure requirements;
2. activity limitations;
3. capital, loss-reserving, and other position limits;
4. takeover rights and other enforcement powers;
5. lines of credit; and
6. performance guarantees.
The first four categories define the net managers' authority to regulate the bank; the last two
categories provide credible ways for regulators to bond themselves to exercise their supervisory
authority in the interests of depositors and other creditors. To complete the web of contract
enforcement, taxpayers must be able to observe and discipline the economic value of their stake
in the rulemaking and enforcement activities that regulators undertake. Ideally, taxpayers must
impose reporting requirements and establish deterrent rights sufficient to persuade net managers
to deploy their examination, supervisory, and lending powers at minimum economic cost to
society as a whole. These costs must be defined comprehensively and include both the costs of
operating the net and the costs of managing its occasional breakdown. Taxpayer-regulator
contracting is important because the practical politics of financial regulation tend to make
regulatory authorities responsive only to immediate bank and depositor concerns.
A nation's safety net is a multiparty web of contractual duties and obligations whose
most palpable features are deposit guarantees and lender-of-last-resort credit facilities. The ideal
or optimal safety net is one that efficiently mitigates the particular monitoring and policing
difficulties that present themselves to banks, depositors, and taxpayers in the informational,
ethical, legal, and economic environment of a particular country at a particular time. This means
that the optimal design and operation of a country's safety net must adapt promptly to changes in
the market, legal, bureaucratic, and ethical/cultural problems the net is intended to alleviate.
For the web to establish incentives for bank and regulators that are compatible with the
interests of all other parties, net design must be environment-specific. As a rule of thumb,
information systems and supervisory technology for monitoring bank capital and risk exposures
7



should be made transparent at least to outside experts and regulatory discipline should mimic
market procedures.
Evidence of Variation in Informational Transparency
Depositors want to be sure that deposit interest rates fairly compensate them for the risk.
exposures that bank loans and investments pass through to them. The "information" needed to
benchmark this compensation consists of valid facts and projections that would help a well-
trained financial analyst to calculate the market value of bank net worth as the difference
between present discounted values of bank assets and liabilities.
When a nation's financial markets inaccurately identify and price risks, they misdirect
savings and investment. Such misdirection undermines a nation's economic growth and well-
being. It is helpful to think of bank disclosures as ore and information as a mineral that
depositors and regulators can, with effort and only imperfectly, extract from this ore. Extraction
is imperfect for two reasons: because banks have a legitimate interest in reserving proprietary
information for their own use and because they may want to conceal potentially damaging
information from other parties.
Bank regulators are supposed to identify and promptly correct material misinformation.
The less effectively the ethical norm of "fair dealing" constrains the business dealings of
corporate and government officials in a given country, the more thoroughly safety-net managers
ought to doublecheck data provided by banks and bank borrowers. However, as a practical
matter, strong incentives may push regulators in the reverse direction. The less effectively
ethical norms and investigative journalism constrain government officials, the more likely it
becomes that safety-net managers may be enlisted to use their instruments to help banks and at
least some bank customers to exploit taxpayers.
In financially sophisticated environments, the reliability of disclosures about bank values
is tested and disciplined --albeit imperfectly-- by an array of outside parties. Rules governing
bank disclosures come both from statutory and administrative law. Statutes are shaped in
legislatures. Regulations governing how to value and itemize sources and uses of funds are
established by administrative agencies and self-regulatory organizations. Enforcement by
rulemaking entities is subject to due-process and constitutional review by a nation's judiciary
system.
Dishonest corporate and government reporting is additionally deterred by the knowledge
that information flows will also be reviewed informally by private "watchdog institutions:"
8



professional accountants, credit bureaus, credit-rating agencies, an independent financial press,
investment advisors, and even academic researchers. However, the information-verification
mission of these watchdogs often conflicts with their other economic interests. Hence, even in
high-income countries, interinstitutional competition may be weak, reporting standards may be
relatively uninformative, and validity checks on bank and borrower disclosures may allow a
great many informational impurities to survive the smelting process.
Recapping the discussion, across countries informational transparency (T) varies with
accounting integrity (Al), ethical norms (EN), press freedom (PF), and the quality and credibility
of compensating restraints regulators place on financial transactors (R). In symbols:
T=T(AI,EN,PF;R).                                    (1)
Several research institutions rate in different ways the quality of information available to
depositors and taxpayers in different countries. Table 1 reproduces measures of the relative
informnativeness of a country's accounting standards, the degree of corruption observed in
government or business transactions, and the extent of press freedom. The table shows that the
quality of relevant information varies greatly across countries. The table also indicates that what
we may call accounting and ethical "integrity" correlate positively with press freedom and each
other and also with the level of a country's per capita income. Across the 41 countries for which
the spottier accounting index exists, accounting integrity and 1990-1995 average real per capita
GDP show a correlation of .59, while the correlation coefficient for accounting integrity with
ethical integrity and press freedom in this subset of countries is .63 and .40, respectively. The
first principal component of the three information variables explains 73.4 percent of their joint
variance. For the 66 countries for which the corruption index has been constructed, the first
principal component of the press-freedom and corruption indexes explains 80.1 percent of these
variable's joint variance and the correlation of ethical integrity with per capita GDP is .80. The
index of press freedom (which is actually available for 73 countries) shows an r=.67 with per
capita GDP.
These correlations suggest not that the level of development determines the level of
informational transparency or vice versa, but that both variables are simultaneously determined
by omitted variables. These omitted variables may be interpreted as a culture's shared beliefs
about what is tolerable and intolerable deal-making behavior. Using this perspective frames per
capita GDP as an imperfect control for evolving social and cultural attitudes that strengthen the
enforceability of financial contracts.
9



For a safety net to operate fairly and efficiently in environments where informational and
ethical integrity are low, the policy-making process of selecting design features must be open
enough to establish accountability between regulators and taxpayers. Political Accountability
increases with the freedom accorded a nation's press and with the political and economic
freedoms it grants its citizens to challenge government policies. However, the correlation of
measures of ethical integrity with the Freedom House Index of press freedom and Heritage
Foundation indexes of economic freedoms suggests that accountability is often weak in the
particular countries where it most needs to be strong.
Where accounting integrity meets a threshold standard of informativeness, independent
credit-rating agencies can consolidate a bank's accounting disclosures into a broad measure of
the risk the bank passes through to its depositors. Using data only on banks whose deposit
quality has been rated by Moody's, Table 2 reports annual values for a system-wide "bank
credit-rating proxy" from 1987 to 1997. The data show many gaps and cover at most about 50
countries. Still, the number of countries for which it is possible to calculate this proxy is
expanding.
For the period 1987 to 1997 and for the much larger sample of countries studied in Table
1, Table 3 displays the percentage of a country's total assets that was held in banks whose
deposit quality was rated by Moody's. This table shows that in many low-income countries, a
sizeable portion of the banking industry is yet to be rated. In these countries, transparency is so
weak that the outside monitoring embodied in the credit-rating proxy provides only an
incomplete doublecheck of the reliability of self-reported banking information.
Evidence of Variation in Depositors' Deterrent Capacity
Given a country's level of informational transparency, an individual depositor's ability to
protect itself from looming bank or borrower defaults is limited by the deterrent rights and
enforcement powers conveyed to contracting parties by its country's legal system. A depositor
may be regarded as holding a contingent claim on the stock of its bank. Similarly, a portfolio of
stock options written on the bank's corporate borrowers is imbedded in the value of bank's loans.
Both sets of stock options come into the money when banks and their borrowers choose to
default.
All defects in counterparty rights, in their enforceability, or in judicial and bureaucratic
efficiency leave financial markets less complete and banks and bank depositors more vulnerable
10



to default. Deterrency (D) depends on a country's systems for policing corporate governance
(CG) and property rights (PR):
D = D(CG, PR; T)                               (2)
Weaknesses in D disadvantage banks as lenders, their depositors as creditors, and
taxpayer-owners of the financial safety net. Where counterparty rights are poorly protected, a
rational saver will always be reluctant to trust its funds to unrelated parties. Table 4 summarizes
cross-country differences in corporate governance, while Table 5 reports measures of broader
property-rights protection. These tables show that legal constraints on opportunities for a
"'controlling-insider" to exploit creditors and minority shareholders differ greatly around the
globe.
In each table, countries are grouped by per capita income. Much as we found for the
informational-integrity variables arrayed in Table 1, alternative measures of deterrent protections
prove highly correlated with one another and the level of deterrency increases on average as per
capita income rises. For the 41 countries where the relevant indexes of accounting and ethical
integrity both exist, each correlates positively with the various indexes of counterparty
protections. The correlation between accounting integrity and the five property-rights indexes in
Table 5 ranges from .53 to .67. The correlations the property-rights indexes show with per capita
GDP and ethical integrity run even higher, ranging from .77 to .89.
The corporate-governance variables are less collinear than the property-rights indexes.
Seven eigenvalues of the corporate-governance covariance matrix exceed the random-correlation
benchmark of unity, while only the first eigenvalue of the property-rights matrix does and the
corresponding eigenvector is able to explain 84.8 percent of the joint variance. This collinearity
in property-rights protections reinforces our contention that unmeasured socio-cultural norms
and freedoms drive these variables and encourages us to focus further analysis on the three
property-rights variables that are available for most countries.
Although bank credit-rating agencies seek to overcome weaknesses in informational
transparency, the profitability of doing this varies across countries with the quality and
usefulness of financial information. As a result, the average proxy value (CR) reported for a
particular country in Table 2 only covers its principal banks. As measured by CR, reported
weakness correlates strongly and inversely with accounting standards (r = -.58), indexes of
counterparty deterrent rights (the highest individual r = -.91), ethical integrity (r = -.74), and
press and economic freedoms (r = .63 and .73, respectively). Because deterrent rights, press and
11



economic freedoms, and ethical and accounting integrity are so highly correlated, the marginal
influence of individual variables cannot be established by their statistical significance in a
multiple-regression framework (Belsley, Kuh, and Welsch, 1980). The confluence of
opportunistic and disciplinary forces is so strong that at five percent significance, corporate-
governance and counterparty protections do not significantly improve our ability to predict the
bank credit-rating proxy. Using only the corruption index and per capita GDP as regressors to
explain CR for the 51 countries for which CR exists produces satisfactory t-values and an
adjusted multiple correlation coefficient R of .81. As a proxy for informational transparency, the
significance of the corruption index supports two hypotheses: that informational reliability
affects bank ratings and that banking fragility increases as it becomes harder for depositors and
taxpayers to monitor banks and bank regulators.
Evidence of Cross-Country Variation in Accountability
No existing data set specifically documents cross-country differences in top officials'
accountability for safety-net performance. However, measures of central-bank independence
have been compiled for 56 countries by Cukierman, Webb, and Neyapti (1992) and inverse
indexes of press and economic freedoms are compiled by Freedom House and the Heritage
Foundation, respectively. Each of these indexes proxies to some degree the accountability
taxpayers impose on economic policymakers in general. The freer is a country's press, the more
readily taxpayers can observe and respond to government policymaking decisions. Similarly, the
less coercive are a country's economic policies, the easier it is to observe circumventive behavior
that both limits and underscores the potential damage that inefficient or unfair policies might
otherwise generate. Finally, for central-bank officials, complete legal independence is the
opposite of political subservience. Cukierman et al. (1992, pp. 380-381) produce a measure of
the extent to which central-bank officials have the authority and autonomy to pursue the goal of
price stability even when this goal conflicts with other government objectives. By extension, the
more politically independent is a country's central bank, the more readily taxpayer can hold its
top officials responsible for the macroeconomic effects of the supervisory policies the bank
adopts.
Like most other variables, central-bank independence and press and economic freedoms
prove significantly correlated with GDP, with correlation coefficients running as high as .69.
We interpret the marginal explanatory power of cross-country declines in central-bank
12



independence or increases in economic repression as evidence of taxpayers' inability to hold
safety-net managers accountable for the costs of the policies they follow.
Evidence of Cross-Country Variation in Banking Market Structure
This section considers how the influence of transparency, deterrency, and accountability
might be affected by the level of banking concentration. In countries where actual competitors
are few and entry protections are effective, the net interest margin between loan and deposit
interest rates is apt to be high. Given an opportunity to extract monopoly rents from their
customers, incumbent banks would be apt to charge interest rates on loans that lie above their
marginal funding costs and to pay interest rates on bank deposits that lie below their marginal
value to the bank.
In the 72 countries for which both banking-market indexes exist, the three-firm
concentration ratio and net interest margin are negatively related to GDP (r = -.29 and -.22,
respectively). However, the banking concentration ratio does not reliably correlate with the net
interest margin. The correlation between the net interest margin and the three-firm concentration
ratio is an insignificant .14. The best predictor of a country's net interest margin in our data set
is the Heritage Foundation's inverse measure of the average expected inflation rate (r=.47). At
5-percent significance, neither banking concentration nor the net interest margin is significantly
correlated with press freedom or accounting and ethical integrity and both variables show few
reliable correlations with deterrent rights. However, the concentration ratio is correlated with
economic repression (r ; .30 for five of the Heritage Foundation's indexes). Observing a
correlation between banking-market concentration and economic repression suggests that a
country's banking market structure is flattened by events that increase customer opportunities to
transact more freely.
To persuade government officials to resist entry requires local bankers to devise a
politically workable way of sharing government-generated rents with friendly politicians and
regulators. Deducting the side payments a bank makes to support entry protections produces
what we may call its net regulatory rent.
In principle, the capitalized value of the projected future net regulatory rents an
incumbent bank can earn is an intangible asset that influences a bank's earnings, stock price,
market capitalization, and credit rating. However, technological change continually undermines
the effectiveness of the entry barriers a country erects by expanding avoidance opportunities for
customers and outside institutions. Monopoly rents create incentives for customers and foreign
13



banks to use technological advances as a way of innovating around govermnent restrictions. In a
world of rapid technological change, enlisting the aid of government officials can only
temporarily slow the rate of outside entry. Thanks to the communications and information
revolutions, globalization is driving the capitalized net value of entry-restriction rents to banks in
high-barrier countries toward negative values. The benefits that entry restrictions deliver to local
banks are eventually forced below the costs of cultivating the political influence necessary to
support them. Once the profitability of local banks becomes dependent on the favors of
government officials, schemes for compensating friendly officials impose tax-like influence-
peddling costs that --as banking market structures meld globally-are eventually bound to
exceed the benefits that entry restrictions can reasonably produce.
Over time, entry tends to benefit host-country customers by pushing loan and deposit
interest rates in their favor. This kind of path dependence makes it impossible to sign a priori the
effect that banking market structure has on safety-net design. The more important efforts to limit
banking competition to sustain bank profits have been in the past, the more rapidly financial
innovation can enable foreign and domestic nonbank competitors to wean high-margin business
away from local banks today. Where a country's market structure is in transition, banking
concentration is more apt to be a sign of large banks' local political clout and vulnerability to
outside competition than of their continuing capacity to earn monopoly profits.
III. Cross-Country Differences in Safety-Net Design Features
In the absence of MT and MD, depositors must watch for harm from two directions:
1. from past losses that bank insiders have managed (possibly with regulatory
connivance) to conceal from public view;
2. from hidden exposures to future losses from illiquidity, bad luck, incompetence,
negligence, fraud, corruption, or zombieness.
For the safety net to efficiently and fairly protect depositors from these dangers, the net's
managers must incorporate design features into the mesh that counter the particular weaknesses
in transparency and deterrency that characterize financial transactions in their country. For best
results, the reasoning leading to particular design decisions should be made transparent to
taxpayers, so that outside analysts can challenge and deter decisions that threaten to harm the
public interest.
14



In most countries, the major element of the safety net is deposit insurance. Table 6
presents data on the major ways that deposit-insurance coverage, management, pricing, and
funding differ across nations. Like most of the data presented in this paper, all but one of the
variables (the coverage limit) are classificatory in character and lack a natural scale. Since the
foreign-exchange value of each local currency limit fluctuates over time, even the coverage limit
cannot be converted to a time-consistent cross-country scale.
The most important difference is whether the guarantees provided to depositors are made
partially explicit or left completely implicit. Guarantees are explicit when they are embodied in
enforceable obligations that may be collected from the insurer's assets as a matter of law.
Explicit systems are usually funded from ex ante premiums or ex post assessments imposed on
eligible institutions. Although implicit deposit insurance is by nature unfunded, it is important
and exists always and everywhere that banks are formally chartered by a specific government.
Guarantees are implicit when their enforceability depends on public confidence in the strength of
recognized political incentives for a country's leaders to bail out or rescue stakeholders in banks
that become economically insolvent. Even in an explicit system, a degree of implicit insurance
comes from the discretion authorities have to treat troubled institutions mercifully. An incipient
banking crisis creates political incentives for incumbent officials in any government with an
explicit system to extend regulatory forbearances, subsidized loans, and unfunded de facto
coverages that exceed the formal limits specified in the nation's laws and regulations. Also, in
many countries, one or more banks are state-owned. For such banks, implicit deposit insurance
is widely perceived to be absolute.
During the 20th Century, socio-cultural expectations and cross-country pressures have
added deposit insurance to the mix of baseline governmental responsibilities in many countries.
Because implicit insurance always exists, whatever explicit limits a government places on the
insurance it writes matter less than it might formally appear. Politicians reserve options to
extend coverages beyond formal limits at taxpayer expense when and if that serves politicians'
collective interests. The probability is far from zero that extracontractual coverage will be
provided when market-mimicking regulatory discipline would better promote taxpayer interests.
Explicit deposit insurance can easily constitute an entry-deterring barrier to exit. This
occurs when deposit insurers allow troubled and inefficient deposit institutions to survive beyond
what we could call their "natural death." Instead of assuring the prompt exit of firms that make
crippling mistakes or allow themselves to become insolvent for any reason, deposit insurance can
15



narrow the industry's "exit drainpipe." When hopelessly insolvent "zombie" institutions are
supported by government guarantees, they bid down profit margins in the industry to
unsustainably low levels that render competing firrns unprofitable. This is most likely to occur
when politicians are unwilling to allow these institutions' contribution to politically inspired
credit-allocation programs to be disrupted.
In the United States, pressure that built up in the exit drainpipe starting in the mid-1960s
was released in the 1980s by an explosion of belated deposit-institution exits. Many now-
departed firms were living-dead institutions whose insolvency could and should have been
resolved long ago. Some of the others were marginal institutions that might have survived had
their competitors' profit margins and risk-taking incentives not been allowed to become badly
distorted.
It is important that society keep in place economically reasonable exit pressure on poorly
performing deposit institutions. When deposit insurance retards exit, new entry is discouraged
and healthy competitors and taxpayers are routinely required to help ruined firms to keep
themselves in play. Such a policy encourages the overexpansion of very high-risk enterprises
and assigns taxpayers a badly structured option on the stock of failing banks. Taxpayers are
required to pay off future losses but receive very little opportunity to participate in gains. The
most that taxpayers can receive from a troubled institution is relief from the loss exposure to
which bailout arrangements have committed them.
Table 6 documents that, during the 1980s and 1990s, countries in many regions adopted
explicit deposit insurance. In Asia and Europe, explicit deposit-insurance guarantees often
extend even to accounts whose value is denominated entirely in foreign currencies.
Figure 1 indicates that explicit insurance is part of the "best practices" policy standards
promulgated in IMF policy reports (Lindgren, Garcia, and Saal, 1996 and Garcia, 1999).
Whether these standards are in fact wise advice is the principal policy question addressed in this
study. Introducing explicit deposit insurance imposes costs as well as benefits on any society
that adopts it. The major cost is to diminish depositor discipline on bank risk-taking by
strengthening the implicit guarantees that government officials otherwise convey. Explicit
arrangements reinforce implicit guarantees by providing bureaucratic and political mechanisms
for patching weak banks.
Where transparency, deterrency, and acountability are weak, installing explicit insurance
may be a great mistake. In contradistinction to the IMF, we hypothesize that the ambiguous and
16



unfunded nature of purely implicit deposit insurance leads depositors in poor informational and
contracting environments to demand a risk premium that is broadly commensurate with the risk-
taking capacity of their bank.
The more completely and more reliably government or private insurers cover depositors
against loss, the less incentive individual depositors have to police the risks their banks can or do
take. An unintended perversity of credible deposit insurance is that it undernines a depositor's
interest in ethical responsibility to look out for itself by gathering information about an
institution's financial condition and by reacting to bad news about this condition as soon as it is
received. This anesthetization of depositor concern permits minor bank insolvencies to fester
and grow into deep insolvencies if the incentive system under which regulators labor leads them
to duck rather than confront supervisory problems.
Whenever a country's banking system is visited with overwhelming losses, bank
stakeholders may be expected to twist the controls of representative government to extract
taxpayer assistance. Introducing explicit deposit insurance is attractive to stakeholders of
zombie banks as a device that can end a crisis and temporarily rescue a deeply troubled banking
system without requiring the embarrassment of explicitly recognizing bank losses or imposing
new taxes. When explicit deposit insurance is introduced in crisis circumstances, the immediate
benefits of the banking-system rescue inevitably come at the expense of longer-run deterioration
in supervisory and bank risk-taking incentives. Because explicit insurance reduces depositor
pressure for transparency and deterrency, opportunities for engaging in unsound and corrupt
banking practices will expand unless government banking supervision strengthens apace.
Typically, adopting explicit insurance as an emergency measure threatens to enhance the
danger of deeper future crises. In crisis circumstances, insurance authorities seldom receive
sufficient monitoring and policing authority to compensate for the depositor discipline their
deterrent activity is bound to displace. Moreover, even in cases where the insurer's deterrent
powers are sufficient in principle, safety-net managers are not made adequately accountable for
using these powers in the interests of society as a whole. When these critical design features are
compromised, explicit deposit insurance encourages a nation's banks to direct a considerable
amount of credit to imprudent longshot investment projects that promise to waste a nation's
scarce savings and reduce the present discounted value of its aggregate stock of real capital.
Confirming these concerns, Demirguc-Kunt and Detragiache (1998 and 1999b) find that,
when they control statistically for the impact of exogenous crisis-generating forces, the
17



likelihood of undergoing a banking crisis is higher in countries that have adopted an explicit
deposit-insurance system than in countries in which guarantees of bank deposits are entirely
implicit in character. A companion paper (1999a) by these same authors shows that open
banking crises are likely to follow the lifting of binding interest-rate ceilings on deposits and that
the likelihood of a crisis is higher in countries where "the rule of law is weak, corruption is
widespread, the bureaucracy is inefficient, and contract enforcement mechanisms are
ineffective." In these environments, capital-impaired institutions are not identified and
disciplined quickly enough to avoid massive losses to insuring agencies and their taxpayer-
owners. The combination of virtually complete coverage and resolution delay encourages
depositors to allow weak institutions to increase risky positions until the aggregate losses
become too large for the insurance system to credibly support.
In principle, several deposit-insurance design features can constrain banks' ability to
exploit weaknesses in transparency and supervisory deterrency. Market discipline can be
generated by assigning private parties a clear margin of responsibility for absorbing at least some
of the losses an insolvent bank accrues. The value to society of incorporating such privatizing
features turns on the credibility of the expectation that government officials will force private
parties to live up to their contractual responsibilities and the presumption that loss-sharing
private parties will not let government procrastination expose them to increasing risks.
One way to privatize bank loss exposures is to make private parties underwrite and
manage some or all of the deposit-insurance system. In Table 6, when the column labelled
"Management" shows a "2," the deposit insurance system is jointly managed by private and
governmental entities; when it shows a "3," responsibility for insurance is formally private.
Another way to constrain bank risk shifting is to insist that formal insurance coverage be
truly incomplete. Most countries specify an upper limit to the size of deposit balance that is
explicitly protected. Relatively few countries extend formal coverage to interbank deposits or
accounts denominated in foreign currency. However, the Long-Term Capital Management
rescue clarifies how easily and unaccountably coverage limits can be breached.
The modem literature on deposit-insurance reform stresses the social benefits of private
coinsurance as a mechanism for disciplining and uncovering regulatory mistakes (e.g.,
Calomiris, 1998; Kane, 1992). Coinsurance means that depositors are contractually required to
bear a share of their bank's accrued losses when their bank fails. This share may be defined as a
18



combination of a nonindennifiable loss-sharing percentage and a fixed amount that the insurer
deducts from each depositor's reimbursable insurance claim.
Although benefits of coinsurance cannot be realized without assured enforcement, they
can be realized without tuming each and every depositor into a loss-bearer. What matters is to
assign to a designated class of private monitors the information and incentives they need to
control bank risk-taking. About 15 countries make at least some depositors coinsure bank losses.
In practice, putative loss bearers are either very large depositors, bonding companies, or
subordinated debtholders.
Presumptions of politically enforced implicit coverages are particularly strong among
depositors of state-owned banks. To investigate the effect of this presumption, we use the La
Porta, Lopez-de-Silanes, and Shleifer (1999) cross-country index of the relative importance of
state-owned banks (GB, for Government Banking presence). GB measures the percentage of
aggregate assets in a country's ten largest banks that were controlled in 1995 by state-owned
institutions. The index runs from precisely zero in about eight countries to precisely 100 percent
in three others. The index is particularly high in socialist and ex-socialist countries. The median
percentage is about 40 percent in Middle Eastem, Asian, and Latin American countries, and is
notably lower for the so-called industrialized countries of Europe.
The GB variable is significantly and negatively correlated with GDP (-.34), all of the
property-rights and information measures (the median r=-.40), economic freedom (r=-.58), and
three corporate-governance variables. These correlations support the hypothesis that government
banking presence is likely to be larger in environments where informational integrity, deterrent
rights, and accountability are weak.
IV. Correlating Deposit-Insurance Design Features with Individual-Country Characteristics
Contracting theory emphasizes that counterparties face strong incentives to minimize the
costs of agency. Black, Miller, and Posner (1978) conceive of a country's deposit insurers as
"stepping into the shoes of individual depositors." This conception clarifies that, absent outside
pressure from international institutions, conscientious officials in individual countries would
design their portion of the safety net to cope with the particular deficiencies in transparency and
deterrency that depositors face in their country's financial and economic environment.
This section builds on cross-country data on deposit-insurance characteristics first
assembled by other researchers (e.g., Talley and Mas, 1990; Kyei, 1995; Goldstein and Turner,
19



1996; Lindgren, Garcia and Saal, 1996; Garcia, 1999; and Demirguc-Kunt and Sobact, 2000).
The analysis seeks to show that observable characteristics of a country's deposit-insurance
system correlate significantly (at the five-percent level) with some of the proxy measures for
transparency, deterrency, and accountability we have identified. Confirming that sensible
bivariate correlations exist supports the hypothesis that cross-country differences in
transparency, deterrency, and accountability matter. The policy implication of this finding is that
IMF and World Bank personnel should recommend changes in the structure of a country's
existing safety net only after carefully analyzing the impact each proposed structural change
promises to have on transparency, deterrency, and accountability.
Explicit vs. Implicit Coverage
We define "expliciticity" (E) as the binary variable which codes a country that offers
depositors explicit guarantees as one and which assigns a zero to countries that offer only
implicit deposit insurance. Expliciticity is insignificantly correlated with government banking
presence (r=-.08). However, expliciticity does correlate significantly and sensibly with other
potential determinants: with per capita GDP, the corruption index, the three accountability
indexes, the rating proxy, the concentration index, the first two property-rights indexes, and nine
indexes of corporate governance. Countries with substantial press and economic freedom, low
banking concentration, central-bank independence, good property-rights scores, good rule-of-law
scores, German or Scandinavian legal origins, and high per capita GDP are apt to have explicit
deposit insurance. Countries that restrict managers' ability to block shareholders from voting
and to disadvantage creditors in various ways are significantly more apt to restrict themselves to
implicit deposit insurance. The explicit-insurance dummy variable shows a strong correlation
with the fiscal-capacity variable, per capita GDP (r=.42). Among the corporate-governance
variables, the highest correlations occur for the management-does-not-stay dummy (r = -.63)
and the index of creditor rights (r = -.55). Unless a creditor has access to reliable public
information and the capacity to win and exercise deterrent rights, it is unlikely that a country's
insurer could be relied upon to wield taxpayer rights effectively either.
Three corporate-governance indexes are always associated with explicit deposit
insurance: German legal origins; no automatic stay on creditors' right to the secured assets of a
reorganizing firm; and creditors' ability to restrain managerial efforts to throw their firm into a
court-protected reorganization. As a predictor of expliciticity, a probit equation that uses the
management-does-not-stay dummy as its sole explanatory variable achieves a pseudo-R2 of .36
20



in the 47 countries for which both variables are coded. Introducing per capita GDP or indexes of
accountability in stepwise fashion cannot significantly improve upon the benchmark
performance of this straightforward creditor-rights explanation.
An alternative way to generate a benchmark equation for E is to consider as predictors
the first few eigenvectors of the 17x17 covariance matrix of corporate-governance variables.
Such eigenvectors are frequently called "principal components." The first three principal
components of the larger set show correlation coefficients of -.49, +.23, and .25, respectively.
When all three are inserted into a probit model, only the first component is significant and the
pseudo-R2 is .285.
Privatization Features
Conventional wisdom maintains that private-sector funding and involvement in deposit-
insurance management enhances regulatory incentives to monitor and discipline inappropriate
bank risk-taking. Private loss-bearers are expected to press for risk-control systems that
substitute economic efficiency for political expediency.
Private participation in deposit-insurance management (PM) is significantly and
positively correlated with all five property-rights indexes, with confiscation risk and rule of law
showing the highest coefficients (.42 and .44, respectively). Private management participation
also correlates significantly with per capita GDP (r=.5 1), ethical integrity (r = .50), press
freedom (r=.43), and most measures of economic freedom (r : .34).
A probit model that recodes private management as unity and joint or government
management as zero and uses confiscation risk and press restriction as explanatory variables
achieves respective z-values of 2.54 and 2.1 1 and a pseudo-R2 of .46. This equation supports the
hypothesis that reliable press reports and rule-of-law deterrent rights are necessary before
taxpayers and depositors can be persuaded to put much confidence in a private fund. Introducing
per capita GDP and bureaucratic quality in stepwise fashion does not significantly improve the
model's predictive performance. An alternative benchmark probit equation using the first
principal components of the property-rights and information variables achieves a pseudo-R2 of
only .25.
Whether a country explicitly funds its deposit-insurance obligations from bank or
government sources correlates significantly with per capita GDP (r=-.35), the focal three
property-rights variables (r--.37), and preemptive rights (rE-.45). The negative sign of these
21



correlations broadly supports the hypothesis that as a device for creating public confidence
explicit funding can partly offset weaknesses in deterrent rights and government fiscal capacity.
Binary variables for other privatization features are not significantly correlated with per
capita GDP, central-bank independence, nor with any of the informational-integrity variables.
However, some other privatization features do correlate with one or another of the property-
rights and corporate-governance variables.
Figure 2 depicts the cross-country trend in adopting explicit insurance over time. For 66
of the 68 adopting countries, we found data on 1995 GDP at market prices. In this sample, the
date of a country's adoption correlates negatively with GDP and positively with press freedom
and ethical integrity. The following regression equation uses YR to represent the year of
adoption and GDP to represent 1995 GDP in billions of U.S. dollars:
YR =  1988.3  -.00067GDP,  R2 =.32, N = 66
(t- 1404.6)  (tP-5.67)
Many recent adoptees show extremely low per capita GDP and a contracting environment
that is so lacking in credibility that it doesn't even register on some of the radar screens from
which we have constructed proxies for transparency, deterrency and accountability. The 40
adopting countries in our principal sample all had explicit insurance by 1998. For these 40
countries, the binary variable designating whether coinsurance exists (CI) correlates significantly
only with confiscation risk (r=.39) and the condition of being funded partly or exclusively from
government sources (r = -.42). Nevertheless, all 13 countries in our sample which give
shareholders a preemptive right to buy shares in new issues formally incorporate coinsurance
into their deposit-insurance system. Using per capita GDP and the first principal components of
information, corporate-governance, property-rights, and accountability variables fails to establish
a satisfactory benchmark probit model for CI. No individual coefficient proves statistically
significant and the pseudo-R2 is only .14.
V. Principal-Components Analysis of Design Features
For the sample of countries for which data on design features could be assembled in early
1999 Table 6 reports on eight focal dimensions of individual-country safety nets. These include
E, CI, PM, whether and how guarantees are funded, the existence of nominal coverage limits,
involuntariness of membership, and whether foreign-denominated and interbank deposits are
also insured. The collinearity we can observe in the design features suggests that we might use
22



principal-components analysis to fashion a few summary measures of the net's character
(Demirguc-Kunt and Detragiache, 1999b). Principal-components analysis estimates linear
combinations of features ("factors") that reproduce the covariance observed in individual
elements. Eigenvectors of the covariance matrix whose corresponding eigenvalue is less than
unity are presumed to express random covariance.
Table 7 shows that only the first three eigenvalues of the covariance matrix of the eight
major design features exceed unity. Cumulatively, their corresponding factors explain 61.6
percent of the variance in design features.
The coefficient loadings that each principal component assigns to individual features
sometimes allow us to interpret them economically. The first component is a summary measure
of the degree to which the net is subject to private market discipline. This factor places heavy
weights on PM, net funding from private sources as opposed to government sources, and
coinsurance. The second component measures the breadth of the bank liabilities the net covers.
It assigns very high weights to the binary variables for compulsory membership and coverage of
foreign-denominated liabilities. The interpretation of the third component is more problematical.
However, to the extent that a system's nominal coverage limits lack credibility, it is fair to regard
limits as potentially disinformational devices for concealing growth in implicit government
guarantees. Granting the need to delete countries where accountability can be established on
other grounds, a case can be made for treating the third factor as a proxy for the extent of
taxpayer susceptibility to hidden risk shifting by banks. Table 8 reports the value of the three
factors for each of the 40 countries whose design features we analyze.
VI. Accountability for Implicit Coverages As a Design Feature
No other index specifically addresses cross-country differences in accountability for the
cost of implicit guarantees. We have emphasized that every country's safety-net managers are
pushed and pulled in contradictory directions. On the one hand, managers are expected to
minimize the risk of a banking disaster. On the other hand, they are expected to minimize the
cost of supporting troubled banks by subjecting banks to market-mimicking disciplines. Because
exercising market-mimicking discipline would help depositors to identify weak banks, a hard-
nosed net manager may expect to be blamed after the fact if a disruptive bank run ensues. This
expectation imposes on every net manager a painful tradeoff between the immediate bureaucratic
and reputational benefits that can be reaped by being merciful to troubled banks and the
23



unmeasured long-run costs that an insolvent bank is apt to shift onto taxpayers when its
insolvency is not resolved promptly.
Society must recognize the problem that opportunistically covering up evidence of
banking trouble and engaging in costly regulatory forbearance is a rational managerial response
when safety-net officials derive reputational and personal benefits from the strength of their
political support. This incentive conflict is not easily resolved. Even a privately managed and
funded deposit-insurance scheme enjoys implicit catastrophic taxpayer back-up. This means that
formal privatization efforts are never complete. The taxpayer remains a silent partner whose
stake in implicit guarantees is both unfunded and unlikely to be formally acknowledged by the
fund's managers. Unless regulatory decisions take place in a MTMD environment for taxpayers,
no practical way exists to make safety-net managers fully accountable in a timely manner for
managing taxpayers' economic stake in the safety net.
Taxpayers' stake consists of the value of the support they provide by explicitly and
implicitly backing up the obligations of whatever government guarantees exist. In the absence of
taxpayer back-up, private and government deposit-insurance managers would have to expend
additional resources each year to convince their fund's counterparties that the managers can be
relied upon to fulfill their contractual commitments (Merton and Perold, 1993). The capitalized
value of this incremental reduction in expenses may be defined as the "risk capital" taxpayers
contribute to the deposit-insurance system. Unless taxpayer-contributed risk capital earns a fair
market return, deposit-insurance schemes end up subsidizing bank risk-taking.
Whenever the informational environment makes it practicable, it is desirable to make
specific officials responsible for measuring the aggregate losses to which the safety net exposes
taxpayers and to price and manage this exposure appropriately. However, especially in
environments where reliable information is scant and corruption is rampant, adopting explicit
deposit insurance with nominal coverage limits may expand implicit guarantees and short-circuit
imperfect, but socially beneficial depositor discipline on bank risk-taking. In the long run, such
societies are apt to pay a high price for substituting unaccountable government supervision for
value-driven private supervision.
Three Ways to Improve Accountability
To guard against unhappy results, political independence for safety-net officials is not
enough. A country moving to explicit deposit insurance would be well-advised to incorporate
design features that promise to generate helpful private discipline on safety-net managers and
24



bankers alike. Broadly speaking, economists have identified three such design features.
Although each of these features can improve supervisory incentives at the margin, none of them
offers taxpayers enough transparency or deterrency to make safety-net managers fully
accountable for the consequences of their disciplinary strategies and procedures. For this reason,
safety-net designers should plan to deploy all three approaches in tandem.
1. Improved Public-Service Contracting
The first design feature is both more straightforward and less practicable than the other
two. It focuses on strengthening the contractual incentives of government officials to intervene
in the affairs of every troubled bank before the bank can impose a loss on the insurance fund.
Unless informational and ethical integrity have evolved to a critical level, this is unlikely to be a
feasible policy strategy. Until managerial rewards for good performance and penalties for bad
performance can be tied to reliable and observable indexes of the ongoing quality of safety-net
management, it is hard to see how to formulate an enforceable incentive-based contract.
This underscores once more the central role transparency plays in safety-net design. It
tells us that a good starting point for improving public-service contracts for bank regulators is to
insist that authorities take steps to make bank risk more visible to outsiders. In particular, a
positive obligation might be placed on banks to compile marked-to-market estimates of bank
capital positions and to report promptly all material changes in individual-bank risk exposures
and capital to regulators. Alternatively, banks that do not make mark-to-market could be
required to hold substantially higher levels of accounting capital and to be subject to particularly
severe penalties for failing to disclose material adverse information.
Once such information became available, government regulators could be obliged to
calculate in a reproducible manner how events affect the opportunity-cost value of deposit-
insurance risk exposures in individual banks and to intervene promptly in the affairs of
overextended banks to resolve their capital shortage in a market-mimicking manner.
Accountability for these activities could be established by having regulators' calculations closely
audited by a multinational private accounting firm and by offering top regulators' in each
country deferred compensation that is expressly tied to annual movements in the net unreserved
value of implicit and explicit deposit-insurance exposures. This compensation would be
forfeited by the members of any management team whose calculations are discredited by
subsequent events. Fostering an information framework that makes government officials more
25



accountable would have the additional value of making strategies for privatizing some of the
safety net work more effectively.
2. Extended Liability
The second design feature seeks to curtail the benefits that stockholders of insolvent
banks and other financial institutions can receive from engaging in go-for-broke risk-taking.
This can be done by extending stockholder liability for liquidation losses beyond the level of the
paid-in capital accumulated at the corporate level. Historically, several now-industrialized
countries (including the U.S.) imposed extended liability on bank shares when their banking
systems were still rudimentary.
For deposit insurers, extended liability has the advantage of increasing transparency,
deterrency, and accountability at the same time. It increases transparency by transforming
movements in the stock price of publicly held banks into a clearer signal of developing strength
or weakness. As compared to a limited-liability shareholding, the deterrency comes from
stockholders' duty to pony up additional funds if (but only if) managers and regulators allow the
bank to become so insolvent that it passes into liquidation. For this reason, an extended-liability
obligation would barely affect the stock price of a well-managed and adequately capitalized
bank. However, with extended liability, the insurer's right to liquidate an insolvent bank carries
a put option that lets it collect a designated amnount of personal assets from every bank
stockholder. Stock markets would imbed the value of this option into the price of each bank's
shares. The value of the option would be negligible for banks that were performing well and
adequately supporting their risk with enterprise capital. However, the put would become
increasingly valuable when and as a bank took poorly supported risks or fell into serious trouble.
Moreover, in environments where the chance that stock-market participants might over-react to
bad news is strong, the deterrent effect of extended liability would be particularly forceful. By
increasing the sensitivity of bank stock prices to changes in bank earning power and earnings
volatility, extended liability would encourage information-revealing stockholder runs on troubled
banks well in advance of their deterioration to zombie condition. These stock sell-offs would
protect depositors by helping safety-net managers to identify institutions that deserve increased
supervisory attention long before the enterprise-contributed capital of these institutions could be
exhausted.
Extended liability also increases accountability. This is because adverse stock price
movements put public pressure on regulators to investigate and take remedial action. They also
26



raise the news value to the financial press of reporting regularly upon regulatory actions and their
effects. To make sure that extended-liability assessments can be collected promptly from failed-
bank stockholders, it would be reasonable for authorities to require stockholders to bond their
extended-liability obligation by depositing a pool of relatively liquid earning assets in a collateral
account at the central bank. Where appropriate, this account might impose some kind of
currency hedge and even be held in safekeeping at the insured institution. It is important to see
that neither transparency nor deterrency requires that the particular assets in the collateral pool be
frozen. Stockholders could be free to move individual assets into and out of the pool over time
as long as settlement float is avoided and the aggregate market value of the collateral account can
be continuously verified. Just as in an ordinary margin account, if the total value of the pool falls
below a specified threshold value, it must be promptly replenished. If it is not, the custodian
must be empowered to force the sale of the securities involved.
For any economy that ranks low in informational and ethical integrity, extended liability
is an elegant solution to the problem of deposit-insurance risk-shifting. This is precisely why it
promises to be difficult to build an effective political constituency for it. Bank stockholders will
resist the efforts to narrow their risk-shifting option and in many countries, regulators will be
reluctant to acknowledge the relevance of the side payments that limited liability generates for
them. Still, extended liability provides a way of driving an observable market-driven wedge
between the economic interests of ethical and unethical bankers. It can be framed most
persuasively as an alternative to more intrusive government supervision of bank activity and as a
strategy that would have little impact on banks that manage to stay healthy.
3. Selective Privatization
The third design feature is to privatize "appropriate" parts of the insurance fund's
aggregate loss exposure in three potentially complementary ways. Doing this would
simultaneously privatize the monitoring and loss-control responsibilities that attach to these
particular positions.
a. Private-Public Insurance Partnerships
The first approach is to establish in countries that choose explicit deposit insurance a
private-public partnership in loss control. The partnership would seek to assign the government
only the catastrophic risk and the particular loss control functions that are associated with
bearing catastrophic risk. An immediate problem in implementing this scheme is to find ways to
separate catastrophic risks from the privately insurable risks that private parties in a given society
27



can credibly insure. A practicable screening criterion would be to insist that plan sponsors
establish just how satisfactorily the insurance functions they propose to transfer could be
managed in the private sector in a crisis environment.
The private deposit insurers would technically function as "sureties." They would act as
third-party guarantors that bond particular deposit obligations that an insured bank offered its
customers. The deposit-insurance law would insist that banks contract with a private surety to
accept responsibility for absorbing the first "X" thousand dollars of losses that accrue to any
deposit account if the bank were liquidated. This responsibility would lead private sureties to
take over the first line of examination and loss control at every insured deposit institution.
Government officials would become managers of what may be characterized as an "overline"
insurer. This layering would leave officials responsible for reviewing and ratifying the
capitalization and loss-control strategies of each authorized private surety. The result would be
that, except during a phase-in period and to test the monitoring performance of individual
sureties, govermnent authorities wouldn't regularly examine individual banks. The private
sureties would be examined and supervised by the government insurer to contain two
catastrophic risks: (1) that one or more private sureties could not cover their obligations to
depositors and (2) that losses in insured institutions could become large enough to exceed the
maximum value of private coverage.
The privatization plan put forward in Kane (1992) has six basic elements. First, insured
deposit institutions would be required to carry a minimum layer of private insurance. Second,
insured institutions could contract for amounts above the minimum, but would be required to
publicize this fact. Third, an insured institution could offer any and all services for which an
authorized carrier was willing to underwrite the pass-through risk during the bonding contract's
specified period of coverage. Fourth, if at the expiration of the surety bond the carrier was
unwilling to renew at least the minimum coverage, the deposit institution would have only three
choices: to improve its risk profile to re-qualify itself for coverage; to find another carrier; or to
be declared in default and taken over by authorities who would be responsible for liquidating or
reprivatizing the bank in a cost-effective manner. Fifth, although individual sureties might
operate in many countries, to assure transparency, authorized carriers would have to be
organized as publicly traded single-line entities and required to issue an array of publicly traded
bills and bonds whose maturity structure would encourage market participants to assess their risk
exposures and loss-control systems at regular intervals. Finally, each private surety would itself
28



be subject to carefully targeted government oversight. Estimates of the loss exposure in a
surety's operations would have to be communicated regularly to the government overseer in each
country in which the surety operates. If authorities judge that the surety's reserves have fallen to
an inadequate level, prompt recapitalization would be required for it to continue operating as a
deposit surety in that country.
To reduce the social costs of deposit insurance, the private surety must be designed to
produce stronger incentives to monitor and deter inappropriate bank risk-taking than a
government surety would face in the same informational and contracting environment.
Privatizing monitoring and disciplinary functions transforms -but does not eliminate-the
incentive conflicts that lead government officials to cover up silent runs and to underdiscipline
troubled banks (Aspinwall, 1992). Irrespective of whether a surety is publicly or privately
owned, its managers are partners in banks' risk-taking outcomes. Surety managers are going to
worry that revealing or openly responding to adverse information about client banks may trigger
depositor runs and crises for which the surety will have to absorb blame. This gives each surety
an incentive to overlook and hide deficiencies at large institutions.
However, a critical difference exists in the strength of this incentive at private and
government entities. Without access to unpriced taxpayer risk capital, private managers must
fund at least some of their losses explicitly. Managers of a government insurance fund can better
conceal the costs of offering inappropriate forbearances to insolvent clients and even to its
partner sureties. Except in open crisis, it is virtually impossible for outsiders to associate
incremental changes in the cost of financing government debt with changes in the performance
of a particular government enterprise. To complete the scheme, the back-up government
catastrophic-risk insurer should be required to have outstanding stock and bonds (Kane, 1992;
Wall, 1997). The market value of these securities would be sensitive to outside estimates of
changes in the firm's operating costs and financial condition. As Kane (1992, pp. 71-72)
explains:
[W]henever poor or self-interested management allows a private surety's
loss exposure to begin to overwhelm its gross reserves, two corrective forces
come into play. First, the surety's guarantees become increasingly less credible.
This causes the demand for its services to fall. It also increases the intensity of
the regulatory scrutiny it receives from private rating agencies and state insurance
commissioners. Second, since a fall in demand makes the surety's earnings
decline, the market value of the surety's stock and outstanding guarantees
decrease also. The decline in stock price creates pressure for the corporation's
stockholders and creditors to press for better management and for outside parties
29



to take over the firm either to liquidate it or to manage its affairs more effectively.
It is important to see that, in informationally efficient financial markets, these
forces would exert themselves irrespective of accounting tricks that permit the
managers of a private surety to keep their firm's financial weakness from being
formally reported in its balance sheets and income statements.
These market forces cannot operate dependably in information-poor environments. They
require liquid capital markets in which traders can reap large rewards from researching each
private surety's financial condition, contracting protocols, loss-control policies, and denials of
coverage. Implementing this design feature begins with identifying particular firms that might
be willing to invest sufficient amounts of capital to operate a surety efficiently. Although
bonding activity has been a traditional province of insurance companies, Ely (1986) has long
argued that large banks are in a good position to sponsor sureties. However, in most low-income
and middle-income countries, financial markets are not highly competitive. In such
circumstances, if explicit deposit insurance exists, it makes considerable sense for surety bonding
to be underwritten largely by reputable foreign financial firms. This would allow each surety to
diversify individual-country risks and would constrain the ability of politicians in particular
countries to interfere covertly with the surety's monitoring and disciplinary efforts.
Still, no one should suppose that a public-private deposit-insurance partnership can be
made solid enough to eliminate the threat of financial crises. In a world in which technology and
science regularly devalue old ways of doing things, a succession of previously profitable
enterprises are bound to become inefficient and have to be shut down. When the plug is pulled,
institutions that have lent money to these enterprises will be asked to absorb some of their
accumulated losses. The purpose of the partnership structure is to contain financial losses as far
as possible to people and institutions that volunteer explicitly to put themselves in harm's way.
Holding these volunteers accountable for engaging in loss-control activity should reduce the
chance that financial-industry losses will pass through to a country's taxpayers who never meant
-nor in any way wanted-to be saddled with responsibility for losses.
b. Subordinated Debt
Without formally introducing private sureties, it is still possible to shift some of the same
responsibility for monitoring and disciplining bank risk-taking to private parties. The central
goal of the second approach is to create a tier of bank debt that stands between depositors and
stockholders and to keep the tier thick enough to absorb whatever losses the bank might
experience. The protective tier of debt is said to be "subordinated" to deposits, because
30



whenever an insolvent bank falls into liquidation, the subordinated debtholders cannot receive
any funds until depositors have been paid off in full.
To protect its stake in the bank, a subordinated debtholder has a strong incentive to
negotiate covenants that give it the capacity to monitor the bank and to limit the extent to which
bank risk-taking exposes the debtholder to loss. Once a firm decides to use subordinated debt,
stockholders have an incentive to minimize interest cost by accepting appropriate covenants
(Smith and Warner, 1979).
Still, in many respects, reliance on subordinated debtholders and private sureties raise
similar economic and bureaucratic issues. In both cases, private securities are being created with
the intention of letting their changing market values publicly signal bank weakness in timely
fashion. In both cases, difficulties in renewing the private contract can force a bank to reduce its
risk-taking or be closed. In both cases, the likelihood of depositor runs is reduced by the
knowledge that private loss absorbers have incentives to close or recapitalize the bank before
losses can grow large enough to reach the depositors.
Both plans make similar demands on banks and government regulators. Banks have to
offer sufficient transparency and deterrency to persuade investors to put enough capital into these
plans to provide the protection depositors need. To make sure this happens, government
regulators have to set, monitor and enforce requirements on sureties and subordinated
debtholders. They must also avoid bailing out troubled banks and the designated loss absorbers.
The central difference between the two plans is that a debtholder invests in the bank at
the outset, while a surety does not have to pay out money unless and until an insured bank fails.
However, because a surety must accumulate enough assets to keep its promises credible, the
difference between the two approaches lies more in the concentration of monitoring
responsibilities and supporting investments in a single entity.
Promising subordinated-debt plans have been put forward by Wall (1988) and Calomiris
(1997). Both plans envision that banks be formally required to meet a government-set minimum
ratio of subordinated debt to either total or risk-weighted assets.
Wall's plan would allow subordinated debtholders to request redemption of their debt at
any time, but be forced to wait 90 days for reimbursement. To fend off closure, the bank would
have to use this 90-day window to replace the debt being redeemed or to raise new funds from
stockholders sufficient to pay the debt off. Banks choosing the second option would receive
31



another 90 days of grace in which to meet the government-imposed subordinated-debt
requirement.
Calomiris focuses his plan on making sure that the subordinated debtholders will remain
a reliable source of discipline when and as a country's banking system weakens. His proposal
follows Benston et al. (1986) in requiring a "laddering" of maturities to assure frequent market
testing of each bank's risk exposures. He would also make subordinated debt nontradable and
limit ownership to a group of pre-approved holders that a country's government would certify to
be both sufficiently well-capitalized and independent of the issuing bank. Looming large on the
roster of eligible debtholders would be global banks. Introducing a "prequalifying" process
makes the subordinated-debt proposal very closely resemble the surety arrangement. The
differentiating feature lies in linking government supervisory discipline to the observable yield at
which the debt is issued. Calomiris recommends that authorities deny banks the right to roll over
subordinated debt at a yield that would exceed an appropriately country-specific maximum
allowable yield spread (say, 5%) over Treasury instruments of comparable maturity. To meet
their subordinated-debt requirement, banks from whom higher yields are being demanded would
have to cut back their risk.
VII. Crisis Management
When a safety net fails in a circus, managers face a multidimensional disaster: at least
one dead or badly broken acrobat, a shocked and grieving staff, a panicked crowd of traumatized
spectators, and a mess that needs to be contained and cleaned up before it can escalate into a
catastrophe. Unless staffmembers have been carefully drilled in efficient techniques for
containing and cleaning up such a crisis, they are unlikely to prioritize and coordinate their
activities in the best interests of the circus or the audience.
When a country's financial safety net actually breaks down or threatens to break down,
equally pressing problems of priority and coordination arise. The first line of crisis management
is an unstinting program of preventive inspection, testing, and prompt repair. In bank
supervision, as in so many other areas of human endeavor, an ounce of protection is worth a
proverbial pound of cure. As we have emphasized, a program of breakdown prevention has two
parts. It begins with policies that extract reliable information on a bank's risk exposures and on
the ability of its stockholders and uninsured creditors to cover these exposures. It concludes with
protocols for regulatory response that commit policymakers to employ the information collected
32



to make sure that capital shortages and excessive risk exposures at individual banks are corrected
promptly and equitably.
It must be recognized that banks have strong incentives to circumvent the prevention
system. Whenever a country's prevention program fails to contain this circumvention, a number
of banks go splat. Crisis management means resolving the insolvency of damaged banks.
Insolvency resolution determines who loses what when the net worth of many banks is wiped out
at the same time. For authorities to work through a banking disaster efficiently, they must
establish a set of reasonable policy priorities in advance and commit themselves to pursuing
these priorities in the event. The first priorities are rescue and triage. Dead and injured
institutions must be sorted out immediately and cared for appropriately. The second priority is
crowd control. Evidence that triage is being handled efficiently should help to curtail panicky
audience runs, but specific staffinembers must take up the task of helping those who want to take
deposits out of their banks to do so in a reasoned and orderly manner. The third priority is to
clean up the mess so that the show may resume without an undue delay.
Rescue and Triage
Rather than extend the circus metaphor, it is better to compare the job of managing a
systemic banking crisis to the task of coping with the sudden collapse of a tall building that
produces an overwhelming number of dead and injured parties. In a systemic banking crisis, the
casualties are the stockholders, employees, depositors, and nondeposit creditors of a nation's
banks. Authorities cannot be expected to find and treat individual casualties efficiently unless
that have formulated an integrated disaster plan and drilled appropriate personnel in its
execution. The trick is to prepare regulatory staffs to react immediately to the first signs of crisis
without having to wait for specific directions from above.
Bank examiners must have access to the data and expertise to size the depth of emerging
insolvencies promptly. Supervisory personnel must be divided into teams that are trained to
determine for every individual institution that suffers an insolvency-revealing run the degree of
help that the institution's various stakeholders would require to make them whole. Without this
information, higher officials cannot evaluate the reasonableness of asking taxpayers to supply
that help.
Supervisors must be conditioned to regard as moribund stockholders' position in any
bank that is too far gone to have any prospects of repaying a taxpayer loan. It is easy to see that
it makes no sense in the midst of an emergency to divert limited surgical resources to sewing up
33



the wounds or resetting the broken bones of a moribund individual. Similarly, officials must be
trained to recognize that in dealing with hopelessly insolvent institutions it makes no sense to
open the public purse to preserve the positions of stockholders and subordinated creditors or to
keep paying top managers lofty salaries.
Insured depositors should be granted access to their funds as soon as it is administratively
possible and uninsured depositors should be accorded a fair degree of immediate fractional
access to their funds. Procedures for setting the transactable fraction of different deposit
accounts should be founded in conservative valuation techniques whose application is carefully
rehearsed in advance. Examiners should be trained to estimate the minimum percentage of
uninsured deposits that could be recovered in an orderly liquidation of the bank's tangible
portfolio. The rest of each depositor's balance should be set aside and unfrozen in stages as the
depth and intangible elements of each bank's insolvency can be more accurately sized. How
fully the positions of other uninsured creditors should be marked down (or "haircut") depends on
the depth of - and margin for error in-the loss assessments that the examination team is able to
assemble.
Just as in a building collapse, urgent decisions must be made quickly. However, this
means that authorities must decide ahead of time what decisions, though necessary, are best
postponed to a more convenient time. In any medical disaster, making decisions about the
urgency of treating different patients is called triage. Triage begins with assessing the extent of
each casualty's particular needs. The first objective of this assessment is to determine which
parties are and are not beyond help. The second objective is to set priorities for treating those
that can benefit from specific kinds of assistance. At each moment, available medical resources
must be allocated to the particular tasks that promise to do the most good. This means rushing
into surgery those whose wounds and injuries either are at a life-or-death stage or are apt to
worsen greatly with delay. Nonsurgical (i.e., less scarce) personnel should be assigned the task
of comforting moribund patients and to helping noncritical patients to keep their spirits up while
they wait for treatment.
It is important to understand that, in the midst of an emergency, such decisions cannot
wait for formal ratification by less-informed higher-ups. Although staffjudgments must be
reviewed and criticized later, during an emergency the autonomy of examination teams must be
respected and supported at all levels of the bureaucracy. The difficulty of switching from a
hierarchical structure of cover-your-ass decision-making to a decentralized task-oriented and
34



cooperative structure of immediate response helps to explain why government officials must not
presume that they can work out efficient schemes of disaster management on the fly.
Crowd Control
Using the words "panic" and "depositor runs" to describe the onset of financial crises
tends to build a subliminal case for authorities to mount a rescue without stopping to establish
priorities. However, one can leamn a great deal about how to manage a panicky depositor run by
analyzing rigorously the metaphors embodied in these emotive words.
The first lesson is to distinguish efforts to treat a fast-burning panic from the methods that
would be appropriate for addressing the longer-lasting phenomena of a financial crisis. The
word "panic" is rooted in classical mythology. Panic comes from the name of the Greek god
Pan. Pan combined a human torso with a goat's legs, horns, and ears. Because of his ugliness,
Pan was said to strike sudden and profound terror into the hearts of shepherds and other
wanderers unlucky enough to cross his path.
The triggering ugliness in a bank panic is the murky surfacing of adverse information that
destroys customer confidence in the repayment capacity of a group of banks. This loss of
confidence may be based either on general information about the consequences of unfolding
economic events or information specific to individual banks or to the particular assets they are
known to hold.
The essence of a banking panic is twofold. First, in a panic, runs are widespread enough
that affected banks cannot raise funds quickly by selling portfolio assets to other parties at fair
prices. Second, institutions not experiencing runs are reluctant to lend enough funds to affected
banks to allow them to maintain the convertibility of their deposits into cash.
For authorities, a systemic panic creates an urgent need to deal with a wave of spreading
bank illiquidity and financial dislocation. However, the urgency of stopping a panic must not be
allowed to over-ride the need to identify hopelessly insolvent zombie institutions and begin the
process of winding up their affairs. Issuing blanket government loans and guarantees to all
troubled banks implicitly shifts the burden of absorbing the losses imbedded in the portfolios of
zombie institutions to taxpayers and relieves managers, stockholders, and creditors from bearing
due responsibility for the loss-making decisions they had previously ratified contractually.
To end a panic efficiently, liquidity must be offered only to potentially solvent
institutions and control over zombie institutions must be put into new hands. This means that
authorities must stop to take the measure of a financial institution's wounds before greatly
35



expanding its access to liquid funds. It should be understood that, during any insolvency-
assessment timeout, would-be transactors have strong incentives to use standard and innovative
forms of credit to prevent transactions from grinding to a halt. Credit cards and checks can be
accepted, along perhaps with other documents, as evidence of personal indebtedness whose
collectable value may be supplemented from other sources if the issuing bank falls into
liquidation.
Periods in which the government temporarily dictates that banks shut their door and
suspend the convertibility of their deposits are called, with more than a little irony, "banking
holidays." The irony lies in the unwelcomeness of the interruption in routine and in how hard
banking and regulatory personnel have to work to bring the holiday to a successful end.
Although the runs metaphor is too vivid to discuss in full, visualizing a panicked banking
system as suffering from the runs underscores the same dilemma. The dilemma flows from the
idea that the suffering party has inhaled or swallowed some potentially contagious antigen and is
undergoing a lagged but rapid elimination of unhealthful material. In a crisis, the antigen is a
loss of wealth that must work itself out through the liability side of the balance sheets of troubled
banks. In designing a treatment plan, it would be a mistake to let the speed of elimination
become the sole consideration. At each moment, the long-run cost of interfering with the
transfer of bank losses to those contractually responsible for financing them must be weighed
against the benefits of bringing the panic to a quicker end.
Three alternative metaphors for understanding and dealing with an unfolding banking
panic can help us to explore the workout process:
1. a meltdown of financial value
2. a stampede
3. epidemic control.
1.    Meltdown: The meltdown metaphor conveys the idea of a sudden annihilation of bank
portfolio value. This metaphor accommodates the idea that the sudden emergence of adverse
infornation about borrower prospects or previously hidden bank frauds or losses might seem to
vaporize what once looked like solid asset value. If the information guiding the metaphorical
"value heat ray" is accurate, the ray ought to make slag only out of banks and assets that are truly
"junk." A well-aimed ray can clear the road for future economic growth by seeking out and
vaporizing society's garbage. As did the stampede metaphor, the meltdown conception identifies
36



the top public-policy priorities in crisis management as limiting the meltdown to junk and
limiting industry exits and takeovers to institutions that truly warrant failure.
2.    Stampede: Bank customers who are running a bank are trying to mitigate their individual
exposure to loss. In this respect, they resemble a herd of animals that has been stampeded by the
appearance of a pack of predators. In the animal kingdom, such stampedes (e.g., of moose
responding to the appearance of hungry wolves) help to cull the herd by letting the quick and the
nimble escape and offering up the slow and the lame for the predator's dinner. Hence, the
stampede metaphor introduces the important entailment that a bank run is a mechanism that
identifies and eliminates weak and infirm banks: a winnowing process that exemplifies the
Schumpeterian ideal of "creative destruction." The deadweight social cost of the winnowing
process is the damage customer credit relationships suffer at institutions whose affairs are
reorganized or liquidated.
Allowing ownership claims on insolvent banks to expire when they become valueless is
necessary for banking markets to function appropriately. Schumpeterian economic theory
emphasizes the practical wisdom -a wisdom that is celebrated in hunting and warrior cultures-
that life springs from death. Joseph Campbell (1972, pp. 176-178) clarifies that, in simpler times
than ours, myths and ceremonies taught generation after generation of their members that to
activate life one kills. The root idea is that life and value do not come into the world, they come
"out" of it. Dead things are transformed biologically into nutrients (and economically into object
lessons), which means that nothing and nobody dies in vain.
A depositor run creates incentives that reward customers and other financial firms for
helping to gather information that analysts can use to judge which banks are and are not strong
enough to deserve an inflow of liquidity-enhancing loans and deposits. Institutions and
individuals who make well-reasoned loan and deposit decisions will earn attractive returns; those
who make bad decisions should not be bailed out by blanket guarantees.
3.    Epidemic Control: The possibility that customer fear and confusion might result in
unwarranted value destruction brings us to the public-policy metaphor of "contagion control."
Contagion occurs when an infection is transmitted from a diseased individual to a previously
uninfected party (Schwartz, 1999). It differs radically from circumstances in which two parties
have ingested a common poison (i.e., bad loans).
Any epidemic spread of confusion per se can and should be countered by micro and
macro policies. Macro control may be effected by avoiding contractionary monetary policies
37



and by making government loans and guarantees available to certifiably healthy banks. Macro
policies may be likened to raising shields to blunt the destructive power of a heat ray, eliminating
predators, or using vaccination programs to block the spread of germs.
All three metaphors sound the theme that policy action should reinforce triage activity
and not counteract it. Micro policy measures should enlist nongovernmental parties in the tasks
of identifying likely survivors and of focusing liquidity assistance on them. The planning and
design question is to make sure that, in judging which banks and assets deserve to be shielded
from spreading harm, authorities establish ways to make use of the analytic capacities of
nongovermnental entities, customers, local clearinghouses, and foreign banks.
Keeping the Exits Open
The idea that regulators should keep some exits open for depositors is a metaphor for
creating a supervisory environment that minimizes the impact of bank crises on the nation's
aggregate money supply. The most straightforward way to preserve the liquidity of bank
depositors is to assure insured and uninsured customers that arrangements are being made to let
them directly or indirectly borrow central-bank funds against the collateral of their recoverable
net claim on each insolvent bank. Preparing examiners to calculate promptly the value at which
bank assets could be liquidated in an orderly manner is the key step in assuring that access to
depositor funds will be restored reasonably promptly.
Once examiners have made these quick-and-dirty calculations, depositors can be granted
fractional access to the funds in their accounts. At the same time, the government should
establish a formal claim on the equity of each insolvent bank either by completely extinguishing
the rights of former shareholders or by taking a warrant position large enough to compensate
taxpayers for the administrative and risk-bearing costs of overseeing the bank's recapitalization.
In either case, the aim would be to sell the government's equity claim to private parties as soon
as fully reliable information on asset values can be developed.
Banking regulators are unlikely to engage in disaster planning unless they acknowledge
that neither systemic banking crises nor banking holidays are unthinkable events. Banking crises
are part of a repeating cycle of bank lending to parties whose ability to repay cannot be fully
assessed in advance. A banking panic typically occurs when previously hidden weaknesses in
borrowers' repayment capacity surface suddenly. The depth of borrower weakness tells us that
large losses need to be allocated across those who hold the liabilities that now-troubled banks
38



issued in the process of funding what, with the benefit of hindsight, everyone now sees to have
been loss-making activities.
It is unreasonable to expect risky bank lending to generate an endless string of success
stories. It is equally unreasonable to suppose that having no alternative to keeping stakeholders
in deeply insolvent banks from paying for their mistakes is a desirable public policy. The
fundamental question in a crisis is what parts of society should wield the hammer and what parts
should be made to act as anvil.
Banks that inefficiently manage their resources destroy their ownership capital. This loss
of capital should unleash market forces that serve to transfer the valuable parts of such a bank's
franchise into better capitalized and potentially more skillful hands. In our softhearted modem
times, industry spin doctors try to convince taxpayers that closing an insolvent bank or assigning
its business to a new owner is a barbarously cruel thing to do. Unless creditors and investors
expect inefficient managers and undercapitalized firms to be promptly and appropriately
disciplined, the incentives that govern the evaluation and selection of risky investment projects
will break down. This is why society must find ways to reward banking regulators for
formulating and enforcing meaningful capital requirements and must pound regulators that dare
to keep zombie banks in play.
Regulators who choose not to develop a plan for allocating losses in a disaster must be
held accountable for inefficiencies that emerge in their inevitably hasty and ad hoc response to
crisis pressures. During the last 60 years, shortsighted regulators have prototypically acted first
and thought about it later. By and large, they have justified the costs of this ad hockery by
framing crisis situations as so unique and so dire that indiscriminate use of government
guarantees and bailout support serve the public interest not only in the short run, and also over
the longer haul. It is hard to understand why taxpayers accept the lack of accountability for the
projectable future budgetary costs and long-run incentive distortion that this characterization
conveys.
Crises are not unique events, nor are they isolated in time from the events that precede
and follow them. On average, the faster and more fairly an insolvent banking system's losses
can be allocated, the quicker socially desirable patterns of bank lending can resume. The more
efficiently and more fairly the process of loss resolution can be conducted, the smaller the bill in
economic resources and social demoralization that taxpayers will have to pay to make the
banking system whole again.
39



VIII. Summary
Banks are in the business of taking risk. These risks originate in the real economy. By
intermediating the flow of real savings and real investment, banks evaluate and monitor the
quality of individual borrowers. They also provide customers with ways to avoid some or all of
the ultimate responsibility for bearing the risks the banks approve.
Accumulating stockholder equity is a way for individual banks to bond the quality of
their monitoring and evaluating activity, but it is not the only way. In particular, because
governments have an interest both in the efficient allocation of national savings and in limiting
the damage bank risk-taking may visit on other parties, governments inevitably supply taxpayer
risk capital to their banking system by erecting and managing a financial safety net.
We take it as a rule of thumb that safety-net managers should avoid either subsidizing or
taxing bank risk-taking. Economic research has never established that taxing or subsidizing
private risk-taking is an optimal way to run a nation's economy. Even if analysts could
formulate a beneficial tax or subsidy rule, it is unlikely that channeling the effect through a
government-run deposit insurance system that fails to account publicly for the size of taxpayers'
stake could improve upon more straightforward arrangements.
A country's safety net is unlikely to be optimal unless it establishes or maintains
transparency, deterrency, and accountability. This paper shows that large differences exist
across countries in the transparency and deterrency that banks afford their depositors. It also
shows why safety-net design needs to take these differences into account.
The weaker is a country's informational, ethical, and corporate-governance environment,
the more a wholly governmental system of explicit deposit guarantees is apt to undermine bank
safety and stability. Put positively, the design features and operating protocols of a country's
safety net ought to evolve over time with changes in private and government regulators' capacity
for valuing banking institutions, for disciplining risk-taking and resolving insolvencies promptly,
and (above all) for being held accountable for how well they perform these tasks.
40



REFERENCES
Aspinwall, Richard C., 1992. "Privatization and Increased Accountability in Banking
Regulation," in Peter Dickson (ed.), Rebuilding Public Confidence Through Financial
Reform. Columbus: Ohio State University School of Business, pp. 80-82.
Belsley, David, Edwin Kuh, and Roy Welsch, 1980. Regression Diagnostics: Identifying
Influential Data and Sources of Collinearity. New York: John Wiley & Sons.
Benston, George, Robert Eisenbeis, Paul Horvitz, Edward Kane, and George Kaufinan, 1986.
Perspectives on Safe & Sound Banking: Past, Present, and Future. Cambridge: MIT
Press and the American Bankers Association.
Black, Fischer, Merton Miller, and Richard Posner, 1978. "An Approach to the Regulation of
Bank Holding Companies," Journal of Business, 5 (July), pp. 379-412.
Brock, Philip L., 1999. "Financial Safety Nets: Lessons from Chile." Seattle: University of
Washington, (March).
Calomiris, Charles, 1997. The Postlodem Safety Net: Lessons from Developed and Developing
Economies. Washington: American Enterprise Institute.
Campbell, Joseph, 1972. Myths to Live By, New York, Bantam Books.
Cukierman, Alex, Stephen Webb, and Bilin Neyapti, 1992. "Measuring the Independence of
Central Banks and its Effect on Policy Outcomes," World Bank Economic Review, 6
(No. 3), pp. 353-398.
Demirguc-Kunt, Aslh, and Enrica Detragiache, 1998. "The Determinants of Banking Crises:
Evidence from Developing and Developed Countries," IMF Staff Papers, 45 (no. 1), pp.
81-109.
Demirguc-Kunt, Aslt, and Tolga Tobact, 2000. "Deposit Insurance Around the World: A Data
Base," World Bank manuscript.
,1999a. "Financial Liberalization and Financial Fragility," Proceedings of the Annual
World Bank Conference on Development Economics, Washington: The World Bank, pp.
303-31.
1 1999b. "Does Deposit Insurance Increase Banking System Stability?: An Empirical
Investigation," (unpublished draft dated July).
Diamond, Douglas, 1984. "Financial Intermediation and Delegated Monitoring," Review of
Economic Studies, 51(July), pp. 343-414.
41



Diamond, Douglas, and Philip Dybvig, 1983. "Bank Runs, Deposit Insurance, and Liquidity,"
Journal of Political Economy, 91(June), pp. 401-19.
Ely, Bert, 1986. "Private Sector Depositor Protection Is Still a Viable Alternative to Federal
Deposit Insurance," Issues in Bank Regulation, (winter), pp. 40-47.
Garcia, Gillian G.H., 1999. "Deposit Insurance: A Survey of Actual and Best Practices,"
Washington: International Monetary Fund, Working Paper No. 99/54.
Goldstein, Morris, and Philip Turner, 1996. Banking Crises in Emerging Economies: Origins
and Policy Options. Basle: Bank for International Settlements.
Jensen, Michael C., and William Meckling, 1976. "Theory of the Firm: Managerial Behavior,
Agency Costs, and Ownership Structure," Journal of Financial Economics, 3(June), pp.
305-360.
Kane, Edward J., 1992. "Long-Run Benefits in Financial Regulation from Increased
Accountability and Privatization," in Peter Dickson (ed.), op. Cit., pp. 67-77.
, 1995. "Three Paradigms for the Role of Capitalization Requirements in Insured
Financial Institutions," Journal of Banking and Finance, 19 (June ), pp. 431-459.
La Porta, Rafael, Florencio Lopez-de-Silanes, and Andrei Shleifer, 1999. "Government
Ownership of Commercial Banks," Harvard University manuscript (November).
La Porta, Rafael; Florencio Lopez-de-Silanes, Andrei Shleifer, and Robert W. Vishny, 1998.
"Law and Finance," Journal of Political Economy, 106 (December), pp. 1113-1155.
Lindgren, Carl-Johan, Gillian Garcia, and Matthew I. Saal, 1996. Bank Soundness and
Macroeconomic Policy. Washington: International Monetary Fund.
Merton, Robert C., and Andre F. Perold, 1993. "Theory of Risk Capital in Financial Firms,"
Journal of Applied Corporate Finance, 6(Fall), pp. 16-32.
Miller, Geoffrey P., 1997. "Deposit Insurance for Economies in Transition," in Yearbook of
International Financial and Economic Law, Amsterdam: Kluwer Law International, pp.
103-138.
Schwartz, Anna J., 1999. "Is There a Need for An International Lender of Last Resort?," in
Shadow Open Market Committee, Rochester: Bradley Policy Research Center,
University of Rochester, Public Policy Study 99-01 (March 7-8), pp. 67-72.
Smith, Clifford, and Jerold Warner, 1979. "On Financial Contracting: An Analysis of Bond
Covenants," Joumal of Financial Economics, 7(June), pp. 117-161.
42



Talley, Samuel H., and Ignacio Mas, 1990. Deposit Insurance in Developing Countries, Policy,
Research, and External Affairs Working Paper Series No. 548, Washington: The World
Bank.
Wall, Larry, 1988. "A Plan for Reducing Future Deposit Insurance Losses: Puttable
Subordinated Debt," Economic Review, Federal Reserve Bank of Atlanta, 73
(July/August), pp. 2-17.
_,___ 1997. "Taking Note of the Deposit Insurance Fund: A Plan for the FDIC to Issue Capital
Notes," Economic Review, Federal Reserve Bank of Atlanta, 82(First Quarter), pp. 14-
31.
43



FIGURE 1
BEST PRACTICES FOR SAFETY-NET DESIGN ENVISIONED
BY IMF RESEARCHERS
All Countries Should Establish Explicit Deposit Insurance
The Insurance System Should Incorporate at Least the Following
Design Features:
- Prudential regulation
- Limitations on coverage
- Mandatory membership
- Political "independence" for regulatory officials.
44



FIGURE 2
CROSS-COUNTRY TREND IN THE ADOPTION OF EXPLICIT DEPOSIT INSURANCE
75
70 ---                                                                                                                6
65 -                                                                                                   _
60
60                                                                                                     _
55  -                  _    _   _    _   _        _   _    _    _   _    _    _ _    _   _    _5              5
t45   -                                                                                         41
s 40-
w 35 _              -                  _ _          _                                             _           _
20---= --X000=____    1920  lt                                                                    -l1"
10X6                                                                      30g|"   l    
25~~~~~~~~~~~~1
15                              12| " | | | |  1_
10       _  _   _1
5-
,99" Ncg NCO Ne  No             N-A N9E9  N4  Nc N9  N* N9G9  NcbO NCO "Ot9Bp 9t0A 90" NO   41 99qq 9      9N  99O  gg9
Yer
Num:berof 6i MSystem_s
Note: Countries that have adopted deposit insurance that are not in the sample we analyzed are: Bahrain, Bolivia, Bulgaria, Cameroon, Central African Rep.,
Chad, Congo, Croatia, Czech Rep., Dominican Republic, Ecuador, El Salvador, Equatorial Guinea, Estonia, Gabon, Gibraltar, Hungary, Latvia, Lebanon,
Lithuania, Macedonia, Marshall Islands, Micronesia, Poland, Romania, Slovak Republic, Tanzania, Uganda, Ukraine



TABLE 1
MEASURES OF CROSS-COUNTRY VARIATION IN THE
QUALITY OF ECONOMIC INFORMATION
Accounting  Corruption     Index of                       Accounting  Corruption    Index of
Standards     Index      Restrictions                     Standards     Index     Restrictions
on Press                                                 on Press
High Income                                                  (continued)
Australia              75          5.11          8.8         South Africa        70          5.35         30.6
Austria                54          5.14         14.6         Trinidad &                      1.80         27.6
Belgium                61          5.29          8.8         Tobago
Canada                 74         6.00          15.2         Uruguay             31          3.00         38.6
Cyprus                            2.60          21.2         Venezuela           40          2.82          35
Denmark                62         6.00           9.4
Finland                77          6.00         15.4         Middle Lower
France                 69         5.43          25.6         Bolivia                         1.35         18.4
Germany                62         5.36          14.4         Botswana                        2.30         27.4
Greece                 55         4.36          28.4         China                           2.55         83.8
Hong Kong              69          5.11         32.75        Colombia            50          3.00         52.2
Iceland                           3.60          12.4         Costa Rica                      3.00         17.4
Ireland                           5.11          17.8         Ecuador                         3.11         36.4
Israel                 64         5.00          29.2         Egypt               24          2.32         75
Italy                  62         3.68          27.8         Indonesia                       1.29         71.4
Japan                  65          5.11         20.2         Jamaica                         1.40         14.8
Korea                  62          3.18         26.4         Jordan                          3.29         50.6
Luxembourg                        3.60          10.4         Morocco                         1.80         52.4
Netherlands            64          6.00          14.8        Namibia                         2.60         27.2
New Zealand            70          6.00          6.8         Panama                          1.20         27.8
Norway                 74          6.00          6.6         Peru                38          2.82          58
Portugal               36         4.43           17          Philippines         65          1.75         44.6
Singapore              78         4.93          63.6         Sri Lanka                       3.00         46.8
Spain                  64         4.43           18          Thailand            64          3.11         39.8
Sweden                 83         6.00          10.2         Tunisia                         1.80         67.4
Switzerland            68         6.00           9.2         Turkey              51          3.11          68
Taiwan                 65         4.11          28.4
United Kingdom         78          5.46         22.2         Low Income
United States          71          5.18         12.8         Bangladesh                      0.85         52.8
Cote d'Ivoire                  2.30          69.2
Middle-Upper                                                 Ghana                           1.95         61.2
Income                                                       Honduras                        1.20         45.6
Argentina              45          3.61         31.2         India               57          2.75         42.4
Brazil                 54         3.79          29.8         Kenya                           2.89         59.2
Chile                  52         3.18           29          Nigeria             59          1.82         80.8
Malaysia               76         4.43           61          Pakistan                        1.79         57.8
Mexico                 60         2.86          54.4         Zimbabwe                        3.25         56.2
Accounting Standards: Index created by examining and rating companies' 1990 annual reports on their inclusion or omission
of 90 items. These items fall into 7 categories (general information, income statements, balance sheets, funds flow
statement, accounting standards, stock data and special items). A minimum of 3 companies in each country were studied.
The companies represent a cross-section of various industry groups of which 70%/o are industrial companies and 30%/O are
financial firms. Higher scores indicate better accounting standards. (Source: International Accounting andAuditing Trends,
Centerfor International Financial Analysis & Research, Inc.)
Corruption: ICR's assessment of corruption in government. Lower scores indicate "high government officials are likely to
demand special payments" and "illegal payments are generally expected throughout the lower levels of government" in the
form of "bribes connected with import and export licenses, exchange controls, tax, assessment, policy protection, or loans".
Scale runs from 0 to 6, with lower scores indicating higher levels of corruption. (Source: International Country Risk Guide)
Restrictions on Press: Assessment of repressive actions and laws, regulations, controls, and political pressures that influence
media content. Score reported is the average index assigned by Freedom House staff in Annual Press Freedom Reports,
1994-1998. Scale runs from 0 to 100, with lower scores indicating greater freedom.



TABLE 2
TIME SERIES OF CROSS-COUNTRY VARIATION IN BANKING-SYSTEM
CREDIT-RATING PROXIES*
1987   1988   1989   1990   1991   1992   1993   1994   1995   1996   1997
Argentina                                                           14.0   14.0   14.0   14.0   14.0
Australia                             3.0    4.1    4.8    4.8    4.8    4.4    4.6    4.2    4.2
Austria                                                             4.0    2.5    2.9    2.9    4.0
Bahrain                                                                                   12.0   12.0
Bahrain - Offshore                                                                               10.0
Belgium                                                     2.0    3.0    2.9    3.4    3.4    4.0
Bermuda                                                                                           6.4
Brazil                                                                     15.0   15.0   15.0   15.0
Canada                         3.1    3.1    3.3    3.8    3.7    3.7    3.7    3.7    3.7    3.7
Chile                                                                      9.0    8.5    8.4    8.0
China                                                               7.0    7.0    9.0    8.7    8.8
Colombia                                                                           11.0   11.0   11.0
Croatia                                                                                           11.0
Cyprus                                                                                    6.0    6.0
Czech Republic                                                                             9.1    9.0
Denmark                               2.0    3.0    3.4    4.2    5.0    4.6    4.6    4.6    4.7
Ecuador                                                                                           15.0
Egypt                                                                                             12.0
Estonia                                                                                           9.4
Finland                        3.0    3.0    3.0    3.0    6.0    6.0    6.0    6.0    6.0    6.0
France                         1.5    1.7    1.9    2.2    3.0    3.7    4.9    4.8    4.7    4.4
Germany                        1.0    1.0    1.0    1.2    1.2    1.6    1.8    1.8    2.1    2.3
Greece                                                                             10.0      10.0   8.1
Hong Kong                                                                   7.0    7.0    7.2    6.5
Hungary                                                                                   12.0   11.5
Iceland                                                                                           7.0
India                                                                              11.0   11.0
Indonesia                                                                                 11.3   13.0
Ireland                                       5.0    6.0    5.5    5.5    5.1    5.1    5.1    4.7
Israel                                                                                     7.0    7.0
Italy                                 2.0    1.9    2.6    4.2    5.5    5.7    6.0    5.9    5.6
Japan                          1.7    1.7    2.9    3,7    3.7    6.0    6.0    6.0    6.0    5.8
Jordan                                                                                            14.0
Kazakhstan                                                                -                       15.0
Korea                                                       7.0    7.0    6.8    7.5    7.2    13.7
Kuwait                                                                                     8.0    8.0
Lebanon                                                                                           15.0
47



TABLE 2 continued
Liechtenstein                                                                                4.0    4.0
Luxembourg                                                                           4.0    4.0    4.0
Malaysia                                                                             5.0    5.3    8.9
Malta                                                                                             . 6.0
Mauritius                                                                                            9.0
Mexico                                                                12.0   12.0   14.0   14.0   13.0
Netherlands                                            3.0    3.0    3.0    2.4    2.5    2.6    2.5
Norway                          4.0    4.0    5.0    6.4    7.0    7.0    7.0    6.0    6.1    6.1
Oman                                                                                         10.0   10.0
Pakistan                                                                                     16.0
Panama                                                                                               12.0
Peru                                                                                                 16.0
Philippines                                                                  13.0   12.6   12.7   12.2
Poland                                                                               12.1   11.8   11.7
Portugal                                                      6.0    6.0    6.7    6.7    6.7    6.9
Qatar                                                                                        9.0    9.0
Romania                                                                                              14.0
Russia                                                                                              13.0
Saudi Arabia                                                                                 10.0   10.0
Singapore                                                                             3.4    3.5    3.5
Slovakia                                                                                     11.0   11.0
Slovenia                                                                                             10.0
South Africa                                                                                 11.0 1- -0
Spain                                          3.5    3.5    4.0    4.2    4.4    4.3    4.3    4.3
Sweden                          1.0    1.0    1.0    2.5    4.0    5.5    5.5    5.5    5.5    4.9
Switzerland                                    1.0    1.0    2.0    2.7    2.7    2.6    3.1    3.2
Taiwan                                                                               4.0    5.2    5.9
Thailand                                                                      6.0    6.8    7.1    13.8
Tunisia                                                                                      11.0   11.0
Turkey                                                                10.0   14.0   14.0   14.0   15.0
United Arab Emirates                                                                         8.6    8.6
United Kingdom                  2.0    2.2    2.8    3.5    3.6    3.6    3.5    3.4    3.5    3.3
United States           6.0    4.7    4.0    4.9    5.2    5.2    4.8    4.6    4.5    4.3    3.9
o.     All ratings are assigned in the year shown. Except for the U.S., proxy values are weighted averages (using
asset weights) of Moody's Long-Term Deposit Ratings of Non-US banks operating in a country. Banks
with no LT-Deposit Ratings, and relatively smaller banks are not included in the sample; this is why the
averages are labelled as "proxies". Ratings were assigned the numerical values shown in the following
table:
i Aaa  Aal Aa2 I Aa3 I Al I A2 I A3 I Baal I Baa2 I Baa3  Bal  Ba2 I Ba3  E Bl  B2 I B3 I Caal I Caa2   Caa3 I Ca  C
I1    2  3   1  4   1  5   L6 j    7  8 S    1   9    1    10   I   11   1    12   1   B    1  14  1 15  1 16  17   1    18   1    19   r 20  21il
48



TABLE 3
PERCENTAGE OF BANK ASSETS RATED IN INDIVIDUAL COUNTRIES, 1987 to 1997
1987  1988  1989  1990  1991  1992  1993  1994  1995  1996  1997
Argentina                                                   12.4  15.0  59.9  63.5  44.4
Australia*                              86.3  79.4  78.9  74.7  76.6  82.1  83.6  87.1
Austria                                                     16.5  33.1  32.8  33.9  22.2
Bahrain                                                                        67.0  64.9
Belgium                                              15.3  48.9  50.0  56.6  60.3  44.0
Bermuda
Brazil                                                             5.4  82.7  80.9  91.5
Canada*                                 74.6  90.5  92.1  90.4  91.0  90.8  94.5  95.4
Chile*                                                            30.3  60.8  62.4  64.2
China                                                       25.4  80.4  78.2  71.4  65.4
Colombia                                                                 10.8  54.9  49.7
Croatia                                                                              26.6
Cyprus                                                                         83.1  85.6
Czech Republic                                                                 68.7  60.0
Denmark                                 44.2  72.2  67.4  69.9  68.7  75.5  79.9  96.7
Ecuador                                                                              38.6
Egypt, Arab Rep.                                                                     35.3
Estonia                                                                              48.5
Finland                                  9.5   9.3   10.2  12.7  12.3  62.3  61.4  67.7
France                                  56.9  59.0  64.8  67.2  65.2  67.2  70.8  70.6
Germany                                  18.3  23.4  24.2  30.6  31.0  33.5  36.5  40.3
Greece*                                                                 90.6  90.4  89.4
Hong Kong, China                                                  16.9  26.8  33.2  34.0
Hungary
Iceland                                                                              28.3
India                                                                   50.3  61.0
Indonesia                                                                      65.4  20.0
Ireland*                                95.5  94.3  86.1  86.9  74.2  62.1  63.7  73.3
Israel*                                                                        92.4  93.8
Italy                                   28.9  27.7  27.7  36.4  67.3  68.2  74.1  71.4
Japan                       13.0  14.0  79.7  80.5  77.9  84.2  85.6  86.3  87.0  84.7
Jordan                                                                               16.8
Kazakhstan                                                                           22.7
Korea, Rep.*                                         13.4  9.6  34.0  60.8  86.9  89.1
Kuwait                                                                         93.7  92.3
Lebanon                                                                              22.1
Liechtenstein
Luxembourg                                                               5.9  12.1  12.5
49



TABLE 3 continued
Malaysia                                                                      33.4  41.4  43.3
Malta                                                                                      50.9
Mauritius                                                                                  59.9
Mexico                                                                        9.6   13.2  89.2
Morocco                                                                              n.a.   n.a.
Netherlands                                       20.2  20.9  20.0  30.2  32.2  35.3  34.3
Norway                                     50.5  48.5  45.0  42.4  40.5  40.6  46.5  48.2
Oman                                                                                 66.5  46.2
Pakistan                                                                             76.0
Panama                                                                                      5.0
Peru                                                                                        69.0
Philippines                                                            15.5  73.8  64.9  57.3
Poland                                                                        16.9  26.0  22.1
Portugal                                                  14.4  25.5  34.6  49.1  67.7  51.0
Qatar                                                                                71.8  70.7
Romania                                                                                    40.1
Russian Federation*                                                                         23.7
Saudi Arabia                                                                         63.2  56.2
Singapore                                                                     90.5  97.2  93.7
Slovak Republic                                                                      66.9  47.2
Slovenia                                                                                    55.0
South Africa                                                                         93.1  86.5
Spain                                       17.2  19.3  25.4  59.2  58.9  63.3  67.3  74.4
Sweden                                     60.1  61.6  62.9  63.9  62.0  72.2  75.9  81.6
Switzerland                                44.1  45.6  45.7  58.0  55.7  65.5  78.2  90.6
Taiwan, China                                                                  9.5   30.0  20.1
Thailand                                                               25.1  59.6  71.1  65.4
Tunisia                                                                              62.5   6.5
Turkey                                                          25.0  25.9  43.4  86.1  38.2
United Arab Emirates                                                                 41.3  42.9
United Kingdom                             21.5  23.5  37.4  42.5  41.5  41.6  45.9  46.1
United States          0.1    0.8   3.8   32.8  33.0  38.1  43.7  51.7  53.5  57.8  55.9
Percentage of bank assets rated: The ratio of the total assets of rated banks to a proxy measure of the banking systems' assets.
Notes: The numerator is the sum of the total assets of the banks having long-term deposit credit ratings by Moody's credit rating
agency. The data source for the numerator is the Bankscope database. A small portion of the rated banks, for which no asset size
data are available, is excluded from the numerator.
The denominator is a proxy for the banking system's assets, and is calculated by adding the 206', 21', and 22nd, lines of the
Intemational Monetary Fund's International Financial Statistics (except for the countries with an '*"). Major aggregates of the
accounts on the assets side are Reserves (line 20), comprising domestic currency holdings with the monetary authorities, Foreign
Assets (line 21), and Claims on Other Resident Sectors (lines 22*). The denominator does not cover all the assets of the banks
and therefore the measure is downward biased. The denominators of the countries marked with an "*" are calculated with
BankScope data because using the IMF figures would produce a figure in excess of 100 percent.
50



TABLE 4
MEASURES OF CROSS-COUNTRY VARIATION IN CORPORATE GOVERNANCE
One            Shares not  Cufnulativ    Opprsse                   cap/itl to   AAnb- Restrictionsforr_NSecurede
share  Proxy   blocked       voting or    minorities   Preernptive   call an   director Mand;ttory    going into    automatic  creditors  Managernent Creditw    Legal
Enpglish    French   German   Scandinafvian   -owne  by mail   before    proportional  mechanism        rights   extraordina  rights   dividend   reorganization    stay on     first    does not stay   rights  reserve
vote            meeting  representation                               ry                                            secwr.ed
meeting                                          assets
High Income                                        __                                                                                   
AustrWiaI    0                                0                  0                   I                      0     ~ 0         0           0                               0.0  2                 0             i      _-3
Belgium                              -1             O0                       O    -                         0             0   ~ 0                                      0   0 O   0                      0.2 0                          2    0.1
Canada                         I         O          0            0           -O      _                                                           0.05        5         0         O                    0                                        -
Cyprus                         I         0          0o                                                                                                      _
Denmark            O                -          O           O           O                       1            0                         O           0.1        2    O                  1                                                 3 0._25r
Finland                        0         O0        -                         0    4                         0                                     0.1       -3         0             0      T                  I        -oI 
Fremce                        0                    -O                        0                 0                         0                       0.-51       3         0             ~0                       -O                       O     o.1
Germany                       O                     I     ~      0          -O0                             0             0                      0.05        1                                     l        10                        -3    0   .1
Grec-e                       0                     0            0               O-                         0            0      2                0.05      2        _6.35         ~O0                         0      ~                 1      03
Icelng Kong                                                     0 I                         0 .l Xl 00.                                                    5-         -0II                                                             4        0
Iceland                       0          0          0                        Q                                                                     .                    
Isand                          I         0          0            0           0       0                      0     .       1     3                 01         4           ~0                        -I-
Italy     0                    0                        0       0         0~~~~~~~~~~~~~~~~~~~~-  0  0                       0.2        -I        0              10                       12                              02
Japan                          0         0                                                     1 1            0                                  0.03        4         O                           0           1                       2      025-
Korea                 ~     ~    ~           ~          ~      ~~0  0    0  1  0              0        0                       0         0.05        2         0             0             I                                   3       0.5
Tetherlands               ii             -I  O-                                      O         O                                                  O    O ~   l. I  2   O                                  . O  1O                      2        O0
New ~Zealand                   I          0         0                  0    0        1                                                O          0.05      4           O-            I             I          -O0                      3        0
Norway                         0         0          0                        0       1                      0            0                        0.1      4- -~                     0             O                                 2        0 .2
i     ~     ~~~        ~    ~    ~ ~         ~     ~     ~     ~ ~~~~ 0  0  0      o  I        O             0                      0.05        3         OOO                                                 O                   0.
Portgapoel                               0                                                                                                                   3 0   0. 4              0                        1                                01
Sigpoen                       0                     0                        0       0         0                   ~                             0.05                  0                                                               2      
SweZ   O  O     O                        O   D                                       O                      ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~0.15    4  O           I                        2   0.2
Switmerlarsd                        0                      0     O           O       O    0          0                                l        0.1          -2         0                                          0.~ 101                     -F5
Taiwan                        0   .                              0        0- 0              oII                                                              3         0 .           0 3 T               o                 0         2          1
United S.,..                   I         0          0            0           0                              1                         0           0.1        5         O             0             0           1           0           I        0
Middle-upper Income                          
Argentina                     -O         -I        -             0          --O                0   .               0 . 0                           -o5  4-1o ~                                     0          -            0                02
Barao                          I_ '  O_
Brail                                                                                             --                                    -                    3        0.    O            Ol   O            O0   O O                   lOO 1-  0.2
Chile                         0           1         0            0           1       O         I            I             I           I          0.0 1       5        0.3            1             0          1            0           2     0.2
Malaysia                       I         0                          0 0                                     0                                     0.1                                           4- 0          l14                               0
Mauritius                      0          1         0                                                                                                       _ 
1 1       -         o          l         o             o        -o        o          o           o             o            l 0 0  0    0.33                 0 O           O                                                        0



TABLE 4 continued
% ofshar
One           Shares not  Cumulative   Oppressed              capil to   Anti-              Restrictions for    No     Secured
share  Proxy   blocked    voting or    minorities  Preemptive   call an   director Mandatory   goitng into    automatic  creditors  Mmtagement Creditor  Legal
English    French   German   Scandinavian   - one by mail  before   proportional  mecbanism    rights   extraordina  rights   dividend   reorganization    stay on    first   does not stay  rights  reserve
vote           meeting  reprsentation _y                                                                     secured
Saudi Arabia                                                                                                                           m
South Africa                           o        0            0          0       1                                 1 O 1   O   0       _     _               0             0            0         0           0          0        0
Trinidad & Tobao             I              0          0        0
Srugy                                  I 0           1                         0         0           0            1          0          0.2        2        0                          0          I                     2       0 2
VTnailan                     0                  0            0          0      0o       -I           1            0           0         0.2        2        0I                                                1         3       0.
Middkt Lower       ~        __ 
Bolivia                      0         I        0            0
Costa Rica         ~         ~0    _      . 
Ecuador                         I~ ~                ~          ~ ~ 0  -0  -0    . IO                       0                    -0.25      -2      0.5            ll                                            I4      0.5
Egypt                        0         1         0           O          0      -0        1                        O                  0    .1      -2   I     0     ~                                                     4      0 5
Indonesia                   0          1        0            0          0      0   - 1               0                        0         0.1   4 2                         I                                  0          2       0
Iran                         0         0        0            0                                                                                              _
hamaica                      I         0        0            0
Iordia                                 0        0                         00           1       °                  1           1         0° ;                0.I                        1                                         25 0.
Morocco
Namibia=
Panyma                       0         0        0            0
Nigerua                      I         0        0                           -0 0               I           0      t           0         0.2        3         0                         0          10                    4      02
Philippines                  0         1-1                                                           I   0 ) 0   0O    0                         _3           .       .                0   -0                0  °    °
Sraislan                     I         0        0            0           0     O                     O            I _         O         0.1         3        0O            1                     O            I         3        0
Thailand ~           ~    ~ ~~~   O         O        0          -O    -O          I            l                               0  2        2         O                                       ~                 3-    0.1
Tmunsia                      0         I                                                                            _                                                                            _
Turkey                       1         0        0            00T                                     0            0           0          05        3         0            1                      -I0                     2 4   0 2
B a n gl ad es h             I         0        0            0       0       I      0   .   14                   5
Ghanya                       I         0        0            01_°0.3 40
Ncpjal                       -         0    .   0         .                     0 _            _                                          .      -                             _~                                         
Nigeria                      I         o0                               0    o-          I                                    0      0    .1    -3           0     _      lI 
Pak~~~~~~~~i s a O                       -O     OI-o                              I.                                   I          -         5        -II                                                         4O
Zimbabwe l~~~~~ 0                        0            0          0 00I000      
52



Legal origin: Dummy variables that identify the legal origin of the Company Law of Commercial Code of each country. (English, French, German or Scandinavian).
Foreign Law Encyclopedia Commercial Laws of the World.
One share - one vote: Equals one if the Company Law of Commercial Code of the country requires that ordinary shares carry one vote per share, and zero otherwise. Equivalently, this
variable equals one when the law prohibits the existence of both multiple-voting and non-voting ordinary shares and does not allow firms to set a maximum number of votes per
shareholder irrespective of the number of shares she owns, and zero otherwise. Company Law of Commercial Code.
Proxy by mail: Equals one if the Company Law of Commercial Code allows shareholders to mail their proxy vote to the firm, and zero otherwise. Company Law of Commercial Code.
Share not blocked before meeting: Equals one if the Company Law of Commercial Code does not allow firms to require that shareholders deposit their shares prior to a General
Shareholders Meeting thus preventing them from selling those share for a number of days, and zero otherwise. Company Law of Commercial Code.
Cumulative voting or proportional representation: Equals one if the Company Law of Commercial Code allows shareholders to cast al of their votes for one candidate standing for
election to the board of directors (cumulative voting) or if the Company Law or Commercial Code allows a mechanism of proportional representation in the board by which minority
interests may name a proportional number of directors to the board, and zero otherwise. Company Law of Commercial Code.
Oppressed minorities mechanism: Equals one if the Company Law of Commercial Code grants minority shareholders either a judicial venue to challenge the decisions of management
or of the assembly or the right to step out of the company by requiring the company to purchase their shares when they object to certain fundamental changes, such as mergers, assets
dispositions and changes in the articles of incorporation. The variable equals zero otherwise. Minority shareholders are defined as those shareholders who own 10 percent of the
share capital or less. Company Law of Commercial Code.
Preemptive rights: Equals one if the Company Law of Commercial Code grants shareholders the first opportunity to buy new issues of stock and this right can only be waved by a
shareholders' vote, and zero otherwise. Company Law of Commercial Code.
Percentage of share capital to call an extraordinary shareholders' meeting: It is the minimum percentage ownership of share capital that entitled a shareholder to call for an
Extraordinary Shareholders' Meeting. It ranges from I to 33 percent. Company Law of Commercial Code.
Anti-director rights: An index aggregating the shareholder rights labeled as "anti-director" rights. The index is formed by adding I when: (I) the country allows shareholders to mail
their proxy vote to the firm; (2) shareholders are not required to deposit their shares prior to the General Shareholders' Meeting; (3) cumulative voting or proportional representation
of minorities in the board of directors is allowed; (4) an oppressed minorities mechanism is in place; (5) the minimum percentage of share capital that entitled a shareholder to call for
an Extraordinary Shareholders' Meeting is less than or equal to 10 percent (the sample median); or (6) shareholders have preemptive rights that can only be waived by a shareholders'
vote. The index ranges from 0 to 6. Company Law of Commercial Code.
Mandatory dividend: Equals the percentage of net income that the Company Law of Commercial Code requires firms to distribute as dividends among ordinary stockholders. It takes a
value of zero for countries without such restriction. Company Law of Commercial Code.
Restrictions for going into reorganization: Equals one if the reorganization procedure imposes restrictions, such as creditors' consent, to file for reorganization. It equals zero if there
are no such restrictions. Bankruptcy and Reorganization Laws.
No automatic stay on secured assets: Equals one if the reorganization procedure does not impose an automatic stay on the assets of the firm upon filing the reorganization petition. An
automatic stay prevents secured creditors from gaining possession of their security. It equals zero if such restriction does not exist in the law. Bankruptcy and Reorganization Laws.
Secured creditors first: Equals one if secured creditors are ranked first in the distribution of the proceeds that result from the disposition of the assets of a bankrupt firm. Equals zero if
non-secured creditors, such as the Government and workers, are given the absolute priority. Bankruptcy and Reorganization Laws.
Management does not stay: Equals one when an official appointed by the court, or by the creditors, is responsible for the operation of the business during reorganization. Equivalently,
this variable equals one if the debtor does not keep the administration of its property pending the resolution of the reorganization process, and zero otherwise. Bankruptcy and
Reorganization Laws.
Creditor rights: An index aggregating different creditor rights. The index is formed by adding I when: (1) the country imposes restrictions, such as creditors' consent or minimum
dividends to file for reorganization; (2) secured creditors are able to gain possession of the security once the reorganization petition has been approved (no automatic stay); (3)
secured creditors are ranked first in the distribution of the proceeds that result from the disposition of the assets of a bankrupt firm; and (4) the debtor does not retain the
administration of its property pending the resolution of the reorganization. The index ranges from 0 to 4. Company Law of Commercial Code.
Legal reserve: Equals the minimum percentage of total share capital mandated by the Corporate Law to avoid the dissolution of an existing firm. It takes a value of zero for countries
without such restriction. Company Law of Commercial Code.
Source: La Porta, Rafael; Lopez-de-Silanes, Florencio; Shleifer, Andrei; and Vishny, Robert W, 1998 "Law and Finance, " Journal of Political Economy, 106(December), pp. 1113-1155.
53



TABLE 5
MEASURES OF CROSS-COUNTRY VARIATION
IN COUNTERPARTY PROTECTIONS
Country                  Risk of         Rule of Law       Contract        Efficiency of    Bureaucratic
Expropriation                      Enforceability    Judicial System    Quality
High Income
Australia                  8.71             10.00             3.04             10 00            6.00
Austria                    9.60             10.00             3.30             9.50              5.64
Belgium                   9.48              10.00             3.29             9.50             6.00
Canada                     8.96             10.00             3.27             9.25              6.00
Cyprus                     7.50             5.98                                                4.32
Denmarkc                  9.31              10.00            3.24              10.00            6.00
Finland                   9.15              10.00            3.00              10.00            6.00
France                     9.19             8.99              2.47             8.00             6.00
Gennany                    9.77             9.23              3.40             9.00              5.96
Greece                     6.63             6.19              2.33             7.00              3.36
Hong Kong                  8.82             8.21                               10.00            4.14
Iceland                   9.25              10.00                                                6.00
Ireland                    8.96             7.80              3.17             8.75              5.46
Israel                     7.54             4.82              3.00             10.00             4.29
Italy                     9.17              8.33              2.10             6.75              4.43
Japan                      9.69             8.99              3.16             10.00             5.89
Korea                      8.59             5.36              2.19             6.00              4.18
Luxembourg                10.00             10.00                                                6.00
Netherlands                9.35             10.00             3.26             10.00             6.00
New Zealand                9.29             10.00                              10.00             6.00
Norway                     9.71             10.00             3.43             10.00             5.32
Portugal                   8.S7             8.69              1.92             5.50              3.70
Singapore                  8.86             8.57              3.22             10.00             5.11
Spain                      8.40             7.80              2.57             6.25              4.11
Sweden                    9.58              10.00             3.30             10.00             6.00
Switzerland                9.98             10.00             3.59             10.00             6.00
Taiwan                     9.16             8.52                               6.75
United Kingdom             9.63             8.57              3.43             10.00             6.00
United States              9.00             10.00             3.55             10.00             6.00
Middle-upper Income
Argentina                  4.91             5.36              2.01             6.00              3.00
Barbados
Brazil                    6.30              6.31              1.97             5.75              4.00
Chile                      6.80             7.02              2.44             7.25              3.36
Malaysia                   7.43             6.79              2.26             9.00              3.54
Mauritius                                   0.00
Mexico                    6.55              5.36              1.77             6.00              2.89
Oman
Saudi Arabia
South Africa              7.27              4.42              2.67             6.00              6.00
Trinidad & Tobago          6.63             6.67                                                 3.11
Uruguay                    7.29             5.00                               6.50              2.00
Venezuela                 6.30              6.37              1.64             6.50              2.89
54



TABLE 5 continued
Country                   Risk of         Rule of Law        Contract         Efficiency of    Bureaucratic
Expropriation                       Enforceability    Judicial System      Quality
Middle Lower
Bolivia                    4.57              2.20               1.76                               1.14
Botswana                   6.71              8.33                                                  3.71
China                      6.29              5.97              2.00                                3.04
Colombia                   7.02              2.08               1.90             7.25              4.00
Costa Rica                 5.79              6.67                                                  2.89
Ecuador                    5.18              6.67               1.86             6.25              3.00
Egypt                      6.05              4.17               2.09             6.50              2.64.
Indonesia                  6.09              3.99               1.76             2.50               1.50
Iran
Jamaica                    6.46              3.51                                                  3.04
Jordan                     4.86              4.35                                 8.66             3.00
Morocco                    5.43              4.46               1.95                               2.93
Namibia                    4.42              6.61                                                  4.42
Panama                     5.11              3.51                                                   1.11
Peru                       4.68              2.50               1.72             6.75              2.11
Philippines                4.80              2.74               1.75             4.75               1.46
Sri Lanka                  5.25               1.90                               7.00              3.00
Swaziland                                    0.00
Thailand                   7.57              6.25               2.23             3.25              4.39
Tunisia                    5.54              4.64                                                   3.00
Turkey                     5.95              5.18              2.00              4.00               3.29
Low Income
Bangladesh                 4.09              2.26                                                   1.21
Cote d'lvoire              6.40              5.64              2.58                                4.00
Ghana                      5.77              3.33                                                  2.71
Honduras                   5.20              3.45                                                   1.57
India                      6.11              4.17               2.00              8.00              3.82
Kenya                      5.66              5.42              2.16               5.75              3.61
Nepal                                         0.00
Nigeria                    4.36              2.74               1.68              7.25              2.29
Pakistan                   4.88              3.04               1.69              5.00              2.71
Zimbabwe                   5.04              3.69                                 7.50              3.43
Risk of Expropriation: International County Risk's (ICR) assessment of the risk of "outright confiscation" or "forced
nationalization". Average of the months of April and October of the monthly index between 1982 and 1995. Scale from 0 to 10, with
lower scores for higher risks. Source: International Country Risk Guide
Rule of Law: Assessment of the law and order tradition in the country produced by the country-risk rating agency Intemational
County Risk (ICR). Average of the months of April and October of the monthly index between 1982 and 1995. Scale from 0 to 10,
with lower scores for less tradition for law and order. Source: International Country Risk Guide
Contract Enforceability: Measures the "relative degree to which contractual agreements are honored and complications presented by
language and mentality differences". Scored 0-4, with higher scores for superior quality.
Source:  Business Environmental Risk Intelligence
Efficiency of Judicial System: Assessment of the "efficiency and integrity of the legal environment as it affects business, particularly
foreign firms" produced by the country risk-taking agency Business International Corporation. It "may be taken to represent
investors' assessments of conditions in the country in question". Average between 1980-1983. Scale from 0 to 10. with lower scores
for low efficiency levels.
Bureaucratic Quality: Average of "bureaucratic quality" assessment values assigned by ICRG between 1982-1995. Scored 0- 6,
with higher scores for superior quality.
55



TABLE 6
CROSS-COUNTRY VARIATION IN DEPOSIT-INSURANCE DESIGN FEATURES
DEPOSIT         Type        Date     Membership Administration Funding   Premium or Assessment          Annual Prtmium          Coverage Limit      Foreign    Interbanlk    Source of
INSURANCE                   Enected/                                                 Bast                                                           Currency    Deposits       Funding
SYSTEMS                    Revised
official=1                                             % of base               in ecu or U.S. dollars /6              0 = Banks Only
explicit=l            compulsory=l    joint--2    funded=l                                                                            yes=l        yes=l    1= Banks & Gov.
implicit=0             voluntary=0    private=3               unfunded=0                                                               no=         no=       2= Govemment
.______________________ ._________________________   __________________                              O nly
Africa                      _________
Benin                 0                                                                                                                             .
Burundi               0
Burkina Faso          0
Cameroon                        1999          0                         l2     deposits and nonperforming  0.15% of depositsT+O0.5% of  $5,336             0
loans                     npls
Central Afiican Rep.           01992                                    1     deposits and nonperforrning  0.1 5% of deposits + 0.5% of  S3,557            0
loans                     npis
Chad                  I         1999          0             2           1     deposits and nonperforming  0.15% of deposits +0.5% of    S3,557
loans                     npls
Congo                 1         1999          02                       -      deposits and nonperforming  0.15% of deposits +  0.5% of  $3,557             0           I             1
______ __  __ _____                              loans                     ispis                            .    _ _  _ _
Cote dIlvoire         0                                                                loans                      np_s
Equatorial Guinea     1         1999          0             2           I    deposits and nonperfonning  0. 15% of deposits + 0.5% of   $3,557             0           1             I
Gabon_________    ________                                                             loans                      npis
Gabon                 1         1999          0             2           I    deposits and nonperforming  0.15% of deposits+ 0.5%of      $5,336             0
___ __ __ __ __ __ __ __ __ __ _   __ ______                              loans                      npls            _ _ _ _ _ _ _ _
Ghana                 0loansnpls
Guinea                0
Kenya                           1985                                                  deposits                    0.15                  S1,50                                        I
Mali                  0                                     .                                                                             ,
Nigeria               I         1988                                                  deposits                   0.9375               S588 / $2435         0
Seychelles            0
South Africa          0
Tanzania                        1994                        3           1             deposits                    0.1                    S376___
Togo                                                                                                                                                                      .
Uganda                          1994          1             I           I             deposits                    0.2                   S2,310                                       I
Zaire                 0, 
Zambia                O                                                                                                                                                             __.__
Zimbabwe              0
Asia                         _ _  _  _     _  _  _      _  _  _ _      _  _     _  _  _  _  _  _  _      _  _  _  _  _  _  _  _   _.__      _ _  _ _  _  _ _  _    _ ' _  _     _ _  _  _



DEPOSIT          Type        Date      Membership Administration  Funding   Premium or Assessment             Annual Premium           Coverage Limit       Foreign    Interbank       Source of
INSURANCE                    Enacted/                                                     Base                                                               Currency    Deposits        Funding
SYSTEMS                     Revised
official=)                                                % of base          in ecu or U.S. dollars                          0 = Banks Only
/6
explicit=l              compulsory-I    joint-2      funded-I    .        __                                                                  yes=l        yes=I    1 Banks & Gov.
Implicit=o              voluntary=0     private=3    unfunded                                                                                 no=O         no=O      2= Government
-O_                                                                _ _    _   _       _   _                       Only
Bangladesh             1          1984           1                     _   I               deposits                    0.005                    $2,123             0            0
China                  0
Hong Kong-             0
India                  1          1961           1             1                  _        deposits                    0.005                    $2,355             I           O               I
Indonesia              O
Kiribati               0
Korea                  I          1996           1             1            1              deposits                     0.05               $14,600 but in full    0 1         01I
until the year 2000
Malaysia               0
Macshall Islands       I          1975           0             1                           deposits             risk-based 0.00 to 0.27 -      S100,000             I           I0
Micronesia                        1963           0             1            I              deposits             risk-based 0.00 to 0.28        $100 000             I           I              0
Nepal                  0                                                                                                                                                        _  _
New Zealand            0                                             _
Pakistan               0                                                   _
Papua New Guinea       0
Philippines            1          1963          T              I            I              deposits                      0.2                    $2,375              I           I
Singapore              0
Sri Lanka              I          1987           0             I            I              deposits                     0.15                    $1,470             0            0
Taiwan                 1          1985           0             1            I           insured deposits                0.015                  $38,500             0            0
Thailand               0
Vietnam                0
Transitional                                                                                                 __  
Soeialist
Economies
Bulgaria               I          1995           1             2            I          insured deposits            risk based to 0.5            $1,784             I            0
Croatia                I          1997           I            23/           I           insured deposits                 0.8                   $15,300            12/           0
Czech Rep.             I          1994                                      l I  I      insured deposits    commercial banks; 0.5, savings coinsurance to 11756    0            0
I                                                                         banks 0.1 
Estonia          l     1         1998           I             2            I         deposits until 2002             0.5 (max)           coinsurance 90°/ of      I            0
I__________                                                                                   1383, but ecu in 2010
Hungary                1          1993           I             2            i          insured deposits            risk based to 0.3       ecu 4165 orS4565        I            0



DEPOSIT         Type        Datc      Membership Administration  Funding   Prenium or Assessment         Annual Premium           Coverage Limit      Foreign    interbank      Source of
INSURANCE                   Enaeted/                                                  Base                                                             Currency    Deposits       Funding
SYSTEMS                    Revised
official=l                      _                         of base                                                       0= Banks Only
/6
explici      P I   compulsory=l    joint-2         funded=l                                                                             yesyl =Banks & Gov
implicitO               voluntary=O    private=3    unfunded                                                                             no=0        no=0O    2= Government
=0                                                                                                            Only
Poland                1          1995.         1             1           1      deposits, also risk-adjusted  not more than 0 4      ecu 1000, then 90%      1           0              I
assets                                      coinsurance for the
next ecu 4000
Romania               1          1996                        2                      insured deposits        risk-based 0.3 to 0.6         $3,600             1           0       _    _
Slovak Republic       1          1996          1             2           Iinsured deposits                   0.1 to 0.3 for banks         S7,900             I           0
Latvia                1          1998          1             I                      itinsured deposits              0.3             $830 until year 2000     1           0
Lithuania             I          99I                         I                      insured deposits                  5                  S6250 then         I             0 O
coinsurance
ICazakstan0                    _  _  _     _  _  _  __                            _  _  _  _  _  _  _     _  _  _  _ _  _  _  _  _  _  _    _  _  _  _
Kyrgyz Republic       0_
Macedonia             1          1996          0             2          -           insured deposits       1.5, risk-based 1% to 5%  coinsurance 75% to                   0             I
Russia                0
Ukraine               1          1998          1             1                _      total deposits        0.5 plus special charges        $250              1           0
Middle East                                             _                                                                                             _                        _
Bahrain    _                                           _    _ 1993  _ _ _2 _0 _        deposits                   ex post                 $5,640             1            0             0
Cyprus                0                  ._                 _ _  _   _  __  _   _  _  _ _  _  _  _  _                                                     _  _  _  _  _  _ _  _  _  _  _  _  _
Egypt                 0                                             _
Iran                  0
Lebanon               O          1967          1             2                       credit accounts                0.05                  $3,300             0            I      _
Libya-                0       ___
Otman                 1          1995                                                                              0 1   deposits    coinsurance 75% to      I            0
S52,630
Saudi Arabia          0
Turkecy               0          1983          I             I           I       insured savings deposits    risk-based 1.0to 1.2         int f                           0 
UnitedArab            0
Emirates                                                __  _  _  _   _  _  __                                                                              ___           __          _   _  _



DEPOSIT          Type   | Date         Membership Administration| Funding   Premium or Assessment             Annual Premium           Coverage Limit       Foreign    Interbank       Source of
INSURANCE                    Enaeted/                                                     Base                                                               Currency    Deposits        Funding
SYSTEMS                     Revised                                                                                                       ______                                     ______
officia=l                                                % of bisse          in ecu or U.S. dollars                         0 =  Banks Only
/6
explicit=I              compulsory= l   joirt=2      fiunded=l                              _                                                 yes=l        yes=t 1= Banks & Cov.
implicit-0               voluntary=0    private=-3    unfunded                                                                                no=0         no'=0     2= Government
.__ _                            =0 JO                              _______    _____                                                  __            Only
Central and Latin
America
Argentina                      1971, 1979,                                             insured deposits         risk-based, 0.36 to 0.72        30,000              I           0              0
1995
Belize                 0     _
Bolivia                0
BirTazi               T   _       1995          1   -          3            I          insured deposits                 0.3          1-  I7,000               -- I              0         --0
Chile                  1          1986                                 _ I  0           not applicable                  none              demand deposits in        I           0              2
full and 90%
coinsurance to UF 120
of $3,600 for savings
deposits
Colombia               I          1985           1                             _       insured deposits                  0.3               in full untill 2001,               1                0
then coinsurance to
_________  S__________                                                  $5,500
Costa Rica             0
Dominican Republic     I          1962           0             2            1              deposits                    0.1875            coinsuranceto $13000 O                 0       _      I
Ecuador                I          1999           I             I            I              deposits                     0.65               in full to year 2001     I                         n.a.
El Salvador            1          1999           1             1            I           insured deposits         risk-based, 0.1 to 0.3         $4,720                      _   0
Guatemala              0       _              _
Guyana            i    0
Honduras               0ii_
Jamaica                1          1998           1             1            1           insured deposits                 0.1                    $5,512              I           0
Mexico                 1          1986           1             1            I           all obligations       0.3 (max) plus 0.7 as needed   in full except        I            I
subordinatcd dcbt
_______________  i_______  i_________   __________   ____________   _______                                                 until  2005     ___________   _________   ________un i 20
Panama                 0      _                                        i                                                                                                  _                             i
Paraguay               0
Peru                      1       1992           I             2            1           insured deposits     risk-based from 0.65 to 1.45      $21,160                           I
Trinidad & Tobago      1    |    1986    |      I             I            I              deposits                      0.2                    $7,957             i1
Uruguay                0
Venezuela        §    I    |      j1985          I             I            I          insured deposits                  2                     $7,309              00                          I



DEPOSIT           Type         Date     Membership Administration  Funding   Premium or Assessment                Annual Premium            Coverage Limit        Foreign     Interbank       Source of
INSURANCE                     Enacted}                                                       Base                                                                  Currency      Deposits       Funding
SYSTEMS                      Revised
official=1                                                  % of bsse           in ecu or U.S. dollars                           0 = Banks Only
. ~ ~ ~ ~   ~     ~    ~ ~ ~ ~ ____                                                                                 /6             _____
explicitP1              compulsory-I    joint=2        fundcd=l                                                                                    yes=l         yes=l    1= Banks &Gov.
limip0                   voluntary=0     private3    unfunded                                                                                       no-0         no=O      2= Government
=0                                                                                                                    Only
Industrialized
Countries
Australia               0
Austria                            1979            1             3             0           insured deposits             pro rata, expost           $24,075 but          I * 
coinsurance for
businesses
Belgium                 l          1974            l             2             1          insured liabilities            0.02 + 0.04           ecu 15000, 20000 in      1 +            0
year 20000
Canada                  1           967            1              1            1           insured deposits               0.33 max         S40,770                        0            2
Denmark                 1          1988            I             2 -           I           insured deposits               0.2 (max)                 ecu 20000      .      1            0 _
Finland                 1    -     1969            1             3             1           insured deposits         risk based: 0.05 to 0.3          529,435              I            0
France                  1          1980            1             3             0                n.a.                on demand but limited             65387             1 *0                          0
Germany                 1          1966            1              3            1          insured deposits in      official is 0.03 but can be   private: 30% of                       0
commercial banks DIS, risk-          doubled                capital; official
assets in other DIS                                  coinsurance 90% to
ecu 20000
Gibraltar                          1998                      -   2             0           insured deposits       admin.expenses and expost       lesser of 900/     - 16                             0
coinsurance of ecu
20000
Greece                  1          1993            1             2             I              deposits           decreasing by size: 0.0025 to     20,000 ECU                          0              0
_____ ____ ____ ____ _____ ____ ____ ____ _____ ____ ____ __   _ _____ ____ ____ ____ ___  1.25
Iceland                 1          1985            1              1            l           insured deposits                 0.15                    ecu 20000                          0              0
Ireland                 1          1989            I              1            I       EU and EEA, i.e insured               0.2                coinsurance 90% to      I**                    0      0
deposits                                             ecu 15000
Italy                   I          19B7                         2 4/           0      protected funds adjusted for  risk adjusted expost 0.4 to 0.S  $ 12 5,000                        0
size and risk
Japan                   1          1971            I             2            T            insured deposits             0.0048 + 0036           $71000, but in fll   -   0             0
_________   __________  ___________  ____________                                                                       until M arch   2001                            _  _  __  _
Luxembourg                         1989                          3                         insured deposits                ex post              coinsurance 90% to0                                   0
ecu 15000 thru 1999,
.______________  ______ ___  _ ______ __   __ _____ ______ _____   ____ _____  then    to   ecu  20000



DEPOSIT         Type       Date    Membership Administration Funding   Premium or Assessment         Annual Premium          Coverage Limit     Foreign    interbank     Source of
INSURANCE                  Enacted/                                                Base                                                          Currency    Deposits      Funding
SYSTEMS                   Revised                                                                                          _____B_ks_O_-_y
official= I                                          % of base         in ecu or U.S. dollars                      0  Banks Only
,explicit-1           compulsory=l    joint--2    funded=l _yes=i                                                                             yes=l     = BanksG&o Gov.
impicii               voluntiiry=    private=    unfunded                                                                          no=0        no=0      2= Government
=0     __   _   _   _   _  _Only
Netherands           I         1979          1             I                      case by case                expost               ecu 20000           I           0
Norway                         1961                      3 5/         1    risk-weighted assets and total  0.005 assets 0.01 deposits  S260,S00        I           0
deposits
Portugal1                      1992          1            1           1          insured deposits   risk-based 0.08 to 0.12 + more  ecu 15000,         1           0
in emcrgencies       coinsurance to ecu
45000
Spain    iI                    1977          I            2           T          insured deposits           max. of 0.2         ecu 15000 through      1           0
1999, then ecu 20000
Sweden               I         199)6                      I           I          insured deposits    risk-based, 0.5 npw, 0.1 later   ecu 28663 $31412   I         0            I
Switzerland          1         1984          0                        0        balance sheet items          on demand               $19,700            0           0            0
United Kingdom                 192                        3           0      EEA deposits i.e. insured      on demand            larger of 90/.                    0            0
deposits                                     coinsurance to
1$33,333 or ecu 22,222
United States        I         1934          1            1           1         domestic deposits      risk-based, 0.00 to 0.27     S100,000         I * *
* Coverage is extended to deposits in the domestic currency, ecus or the currencies of other members of the EU.
** * Coverage is extended to deposits in sterling, ecus or the currencies of other members of the European Economic Area (which include EU).
1/ Korea has placed a temporary full guarantee on deposits.
2/ Croatia excludes foreign currency deposits placed prior to 1993 as they were covered by an issuance of government bonds.
3/ In Croatia, the system is administered by a private agency, but some decisions must be approved by the central bank.
4/ Although the scheme is privately run in Italy, all decisions must be approved by the central bank so the DIS has little authority.
5/ Norway has two schemes. Both are privately run, but each has two public members on its seven-member board - one from the central bank and the other from the Banking and
Securities Commission.
6/ Exchange rates at the end of June 1998.
7/ Coverage in Nigeria is higher at the official exchange rate than at the market rate.
8/ The premium in Belgium can be raised by a maximum of 0.004 percent when the funds' liquid assets fall below a critical level.
9/ The premium charged by the private deposit insurance schemes in Germany vary by scheme from 0.004 to 0.1 percent.
10/ Coverage in Iceland in principle is full. The minimum is ecu 20,000. Above that, payment is in proportion to the resources of the fund.
11/ Coverage in Latvia will rise gradually to ecu 20,000 by the year 2008.
12/ in the Netherlands, the ex post assessments are made case-by-case on the bases of several items of data recently reported to the central bank. A comparison is made between the
portfolios of the failed bank and the assessed bank. Costs are apportioned after consultation with the bankers' committee.
13/ Article 25 of the deposit insurance law in Poland sets premiums at no more than 0.4 percent of deposits. However, Article 13 states that premiums should not "exceed 0.4 percent of
the sum of assets rated according to risk".



14/ Turkey has provided unlimited coverage since May 1994.
15/ In Argentina, deposits that pay more than 200 basis points above the reference rate are not insured.
16/ Premiums in Colombia will become risk-based in the year 2000.
17/ The premium in Peru is computed to the maximum amount insured and applies only to deposits of individuals and non profit institutions. Banks pay 0.65 percent of total deposits plus
0.2 percent for each higher risk category.
Source: "Deposit Insurance: A Survey of Actual and Best Practices". Prepared by Gillian G.H. Garcia, Working Paper International Monetary Fund, Monetary and Exchange Affairs
Department, c 1999 (Washington, D.C.)



TABLE 7
PRINCIPAL COMPONENTS ANALSIS OF BINARY VARIABLES INDICATING
EXISTENCE OF EIGHT SAFETY-NET DESIGN FEATURES AS PART OF THE
EXPLICIT DEPOSIT-INSURANCE SYSTEMS OF 40 COUNTRIES
Component  Eigenvalue          Proportion of       Cumulative Proportion
Covariance Explained          Explained
1         2.35396             0.294                   0.294
2         1.54301             0.193                   0.487
3         1.03226             0.130                   0.616
4         0.92069             0.115                   0.731
5         0.74728             0.093                   0.825
6         0.67318             0.084                   0.909
7         0.41151             0.051                   0.960
8         0.31811             0.040                   1.000
Eigenvectors
Design         1        2        3         4        5        6         7        8
Feature
Involuntariness   0.03446   0.70951  0.13933   0.09557  0.05198   0.02037  -0.67204  -0.11001
CL             0.13924  -0.23528  0.75875   0.41237  -0.11961  0.39312  -0.00576  -0.10311
CI             0.38789   0.20163  -0.34568  0.25071  -0.63197  0.29205   0.09714   0.36390
PM             0.50853  -0.14058  0.07508   0.05889  0.51626  -0.15133  -0.16602  0.62892
Funded at all    -0.43254  0.15256  -0.16777  0.04716  0.39988   0.71419   0.11848   0.27433
Privately Funded  0.46589  -0.06180  -0.39694  0.29620  0.35720   0.20512   0.05075  -0.60133
Foreign-Den.   0.25809   0.57313   0.29976  -0.22588  0.13917  -0.05078  0.66413  -0.04658
Interbank      -0.31561  0.16250  -0.06841  0.78327  0.10107  -0.42677  0.23161   0.10139
Note: All variables are binary. Each variable takes on the value of unity when the
specified design feature is present. "Involuntariness" indicates that membership is
mandatory; CL indicates that nominal coverage limits are specified; CI indicates that
coinsurance exists for at least some depositors; "funded at all" indicates that deposit-
insurance obligations are funded in some way; "privately funded" indicates that funding
comes exclusively from private sources; "foreign-Den." indicates that foreign-
denominated deposits are explicitly covered; "interbank" indicates that interbank deposits
are formally guaranteed.



TABLE 8
CALCULATED VALUES OF THE FIRST THREE PRINCIPAL COMPONENTS
OF SAFETY-NET DESIGN FEATURES IN 40 COUNTRIES
Susceptibillty to idden
Countries             Degree of Privatization    Breadth of Coverage      Risk Shifting
Argentina                     0.974                 0.997                     0.708
Austria                       1.328                 1.107                     0.927
Belgium                     -0.001                  1.199                     1.030
Bangladesh                   -0.259                 0.627                     0.730
Brazil                        0.974                 0.997                     0.708
Canada                      -0.574                  0.789                     0.666
Switzerland                   1.114                -OA38                      0.437
Chile                         0.432                 1.046                     1.198
Colombia                      0.279                 0.928                    -0.077
Germany                       1.362                 1.198                     0.363
Denmark                      -0.001                 1.199                     1.030
Spain                       -0.001                  1.199                     1.030
Finland                       0.508                 1.058                     1.105
France                        1.406                 0.844                     0.876
United Kingdom                1.794                 1.046                     0.530
Greece                       0.465                  1.137                     0.633
India                        -0.001                 1.199                     1.030
Ireland                      0.853                  1.339                     0.287
Iceland                      0.714                  1.574                    -0.471
Italy                        0.820                  1.248                     0.852
Jamaica                      -0.001                 1.199                     1.030
Japan                        -0.398                 0.862                    -0.028
Kenya                        -0.316                 1.361                     0.965
Korea                        -0.259                 0.627                     0.730
Sri Lanka                    -0.293                -0.083                     0.591
Luxembourg                    1.794                 1.046                     0.530
Mexico                       -0.456                 1.597                     0.207
Nigeria                      -0.574                 0.789                     0.666
Netherlands                   0.432                 1.046                     1.198
Norway                        0.508                 1.058                     1.105
Oman                         0.387                  1.401                     0.684
Peru                         -0.001                 1.199                     1.030
Philippines                 -0.316                  1.361                     0.985
Portugal                     0.387                  1.401                     0.684
Sweden                       -0.001                 1.199                     1.030
Trinidad Tobago              -0.316                 1.361                     0.965
Turkey                       -0.140                 1.434                     0.271
Taiwan                       -0.293                -0.083                     0.591
United States                -0.316                 1.361                     0.965
Venezuela                    -0.259                 0.627                     0.730






Policy Research Working Paper Series
Contact
Title                            Author                  Date              for paper
WPS2429 Corruption, Composition of Capital   Shang-Jin Wei        August 2000        H. Sladovich
Flows, and Currency Crises                                                 37698
WPS2430 Financial Structure and Bank      Asli Demirguc-Kunt      August 2000        K. Labrie
Profitability                    Harry Huizinga                            31001
WPS2431 Inside the Crisis: An Empirical   Asli Demirguc-Kunt      August 2000        K. Labrie
Analysis of Banking Systems in   Enrica Detragiache                        31001
Distress                         Poonam Gupta
WPS2432 Funding Growth in Bank-Based and   Asli Demirguc-Kunt     August 2000        K. Labrie
Market-Based Financial Systems:    Vojislav Maksimovic                     31001
Evidence from Firm- Level Data
WPS2433 External Interventions and the    Ibrahim A. Elbadawi     September 2000     H. Sladovich
Duration of Civil Wars           Nicholas Sambanis                         37698
WPS2434 Socioeconomic Inequalities in Child   Adam Wagstaff       September 2000     A. Maranon
Malnutrition in the Developing World   Naoko Watanabe                      38009
WPS2435 The Evolution of Poverty during the   Asep Suryahadi      September 2000     P. Sader
Crisis in Indonesia, 1996-99     Sudarno Sumarto                           33902
Yusuf Suharso
Lant Pritchett
WPS2436 Safety Nets and Safety Ropes:     Sudarno Sumarto         September 2000     P. Sader
Who Benefited from Two Indonesian   Asep Suryahadi                         33902
Crisis Programs-the "Poor" or the    Lant Pritchett
"Shocked?"
WPS2437 Quantifying Vulnerability to Poverty:  Lant Pritchett     September 2000     P. Sader
A Proposed Measure, Applied to   Asep Suryahadi                            33902
Indonesia                       Sudarno Sumarto
WPS2438 Measurements of Poverty in        Menno Pradhan           September 2000     P. Sader
Indonesia: 1996,1999, and Beyond   Asep Suryahadi                          33902
Sudarno Sumarto
Lant Pritchett
WPS2439 State-Community Synergies in      Monica Das Gupta        September 2000     M. Das Gupta
Development: Laying the Basis for   Helene Grandvoinnet                    31983
Collective Action                Mattia Romani
WPS2440 Lessons from Uganda on Strategies  John Mackinnon         September 2000     H. Sladovich
to Fight Poverty                 Ritva Reinikka                            37698
WPS2441 Controlling the Fiscal Costs of   Patrick Honohan         September 2000     A. Yaptenco
Banking Crises                   Daniela Klingebiel                        38526



Policy Research Working Paper Series
Contact
Title                             Author                   Date               for paper
WPS2442 A Firms's-Eye View of Policy and    Bernard Gauthier         September 2000     L. Tabada
Fiscal Reforms in Cameroon        Isidro Soloaga                              36896
James Tybout
WPS2443 The Politics of Economic Policy     Richard H. Adams Jr.     September 2000     M. Coleridge-Taylor
Reform in Developing Countries                                                33704
WPS2444 Seize the State, Seize the Day:     Joel S. Hellman          September 2000     D. Billups
State Capture, Corruption, and    Geraint Jones                               35818
Influence in Transition           Daniel Kaufmann
WPS2445 Subsidies in Chilean Public Utilities   Pablo Serra          September 2000     G. Chenet-Smith
36370
WPS2446 Forecasting the Demand for          Lourdes Trujillo         September 2000     G. Chenet-Smith
Privatized Transport: What Economic Emile Quinet                              36370
Regulators Should Know, and Why    Antonio Estache
WPS2447 Attrition in Longitudinal Household    Harold Alderman       September 2000     P. Sader
Survey Data: Some Tests for Three   Jere R. Behrman                           33902
Developing-Country Samples        Hans-Peter Kohler
John A. Maluccio
Susan Cotts Watkins
WPS2448 On "Good" Politicians and "Bad"     Jo Ritzen                September 2000     A. Kibutu
Policies: Social Cohesion,        William Easterly                            34047
Institutions, and Growth          Michael Woolcock
WPS2449 Pricing Irrigation Water: A Literature   Robert C. Johansson  September 2000    M. Williams
Survey                                                                        87297
WPS2450 Which Firms Do Foreigners Buy?      Caroline Freund          September 2000     R. Vo
Evidence from the Republic of     Simeon Djankov                              33722
Korea
WPS2451 Can There Be Growth with Equity?    Klaus Deininger          September 2000     33766
An Initial Assessment of Land Reform  Julian May
in South Africa
WPS2452 Trends in Private Sector Development Shobhana Sosale        September 2000      S. Sosale
in World Bank Education Projects                                              36490
.,I,(,