July 8, 2004

                       AN INTERNATIONAL COMPARISON
                                Edward J. Kane
                                Boston College

         In every risky enterprise, safety is a paramount concern. Miners wear hard hats
and equip themselves with safety lamps. Cruise ships stock life vests and lifeboats and
routinely drill personnel and passengers in their location and proper use.
         Society’s concern for the safety of a particular enterprise increases with the
degree of adversity that attaches to bad outcomes and with the extent to which the bad
outcomes damage parties who cannot directly influence the tradeoffs between
profitability and safety that enterprise managers might feel free to make.
         Managers of financial institutions may be likened to a team of high-wire artists.
They deliberately throw themselves into risky positions and, when things turn out badly,
a messy multiparty disaster can ensue. What a government calls its financial safety net
consists of measures it takes to restrict the risky positions institutions assume in the first
place and to limit the damage customers, employees, and stockholders suffer when and if
disaster ensues.
         This paper begins by analyzing the safety-net metaphor and using its entailments
to clarify why the optimal design of financial safety nets varies with country
circumstances. Section II of the paper introduces the idea that individual countries have a
distinctive contracting environment and regulatory culture. The contracting environment
encompasses the degree to which political and legal institutions and operative concepts of
personal honor influence the accuracy of information flows and shape methods of private
contract enforcement. The regulatory culture simultaneously authorizes and restricts the
exercise of government power. Section III provides data on cross-country differences in
safety-net design. It also identifies the benefits of specific design features. Section IV
emphasizes the importance of incorporating accountability and disaster planning into
safety-net design. It warns officials against introducing explicit deposit insurance

 This paper updates and refocuses material first treated in Kane (2000). The author is grateful to the
Fundacion de Las Cajas de Ahorros Confederados in Madrid for financial support and to Richard
Aspinwall for helpful comments on an earlier draft.

                                              Page 1 of 49
                                                                                July 8, 2004

without first making sure that their country’s contracting environment and regulatory
culture can control the subsidy to risk-taking that it might generate.

I. Exploring the Safety-Net Metaphor
       A net is a cross-hatched mesh whose strands are composed of material strong
enough to serve the net’s particular purpose. Although a mosquito net may be made of
flimsy material, safety nets have to be strong enough to withstand the force of muscular
bodies falling with great force.

Circus Safety Nets.
       In a circus, a safety net serves two purposes. Its immediate function is to protect
acrobats from splattering themselves against the ground when they make a mistake and to
protect a paying audience from the trauma of witnessing such a disaster. Its deeper
purpose is to encourage acrobats to practice and perfect stunts that are challenging
enough to thrill an audience.
       Stunts must be perfected, because audiences will not pay to watch acrobats fall
repeatedly into a net. Carrying out a relatively simple trick is more entertaining than
botching a difficult one. From an economic point of view, safety nets are erected not to
inspire wild risk-taking, but to embolden performers enough to free their minds to
analyze and undertake risks in prudent and skillful ways.
       Safety-net design has many dimensions. Moreover, the costs and benefits of
particular features unfold unevenly and over many periods. At the margin, increases in
the present discounted value of the entertainment engendered by enhanced risk-taking
must be traded off against the incremental costs of generating additional safety. Costs
may be divided into: immediate costs of acquisition; future monitoring and maintenance
expense; risks of technological obsolescence; and the residual risk of incurring
catastrophes the net cannot cover. Benefits consist of cash flows that the circus earns
from the performance of entertaining stunts.
       Wide-reaching and sturdy meshes are safer, but their acquisition cost is higher
and the greater visibility of the protection they supply lessens the sense of risk sharing on
which audience enjoyment depends. Sturdy filament materials exist whose thinness can

                                        Page 2 of 49
                                                                                 July 8, 2004

make the net less visible, but these materials are more costly to acquire and maintain.
Finally, because no net can stop every possible disaster, conscientious operators must
make plans to access emergency medical treatment promptly and to withstand whatever
criticism subsequently emerges in the press.

Comparison with Financial Safety Nets
       The economic principles that govern the design of a circus net apply equally well
to the design of a country’s safety net. To design and operate a safety net optimally,
authorities must consider costs and benefits over many periods and many of the tradeoffs
they must make are subtle ones. Metaphorically, these tradeoffs share four entailments
with a circus net.
       First, like acrobats, financial institutions that enjoy safety-net protection may be
tempted to engage in activities so risky that critical mistakes can still cripple them or
even end their existence. This phenomenon is called moral hazard. To explain moral-
hazard incentives, it helps to draw a correspondence between stakeholders in a bank and
stakeholders in a circus. In a banking enterprise, borrowers and depositors correspond to
individuals who manufacture the trapezes and tightwire platforms on which stunts are
performed. To assess how fully their claims will be serviced, customers need to be able
to monitor how well the bank is doing. Unless reliable public information about bank
and borrower performance is in easy supply, customers and regulators have to devise
procedures with which to extract reliable data. Other things equal, the weaker a country’s
information environment, the more an unguaranteed customer’s stake is at risk. Of
course, government efforts to protect bank customers by monitoring, controlling, and
bailing out banks generate costs that must be financed, usually by assessments levied
both explicitly and implicitly on banks and taxpayers. This financing pattern means
taxpayers are not simply the equivalent of audience members. The government may call
upon taxpayers to pay for any deficits the regulated “circus” might suffer. Financial
safety nets protect borrowers, depositors, and taxpayers only from being directly harmed
by financial-institution missteps. Indirectly, they harm them by encouraging individual
institutions to shift some of the risks generated by their lending and funding activities
onto the safety net.

                                        Page 3 of 49
                                                                                 July 8, 2004

       Second, thrust into the role of circus managers, regulators have to balance
conflicting concerns. They must guard against excessive risk-taking, while also
controlling over time the other costs and benefits the safety net produces. Safety nets for
banks comprise a series of policy arrangements that help stakeholders to avoid or weather
loss-causing financial shocks (Diamond and Dybvig, 1983: Talley and Mas, 1990; Kane,
1995; Goldstein and Turner, 1996;Brock, 1999). These multidimensional policy schemes
seek to align the costs and benefits generated by:
           1. limiting aggressive risk-taking by banks;
           2. protecting bank customers from being blindsided by bank insolvencies;
           3. preventing and controlling damage from bank runs;
           4. detecting and resolving insolvent banks; and
           5. allocating across society the losses officials uncover when an insolvent
               bank is closed.
       Third, decisions about the design of a country’s financial safety net are not
necessarily made selflessly. Decisions are influenced by incentives that result from
political and bureaucratic arrangements that convey to regulatory officials and bank
stakeholders a collection of mutually reinforcing rights and duties. To fashion a net of
the right size and strength, decisions bearing on the cost and effectiveness of net
components should be observable enough to be disciplined by healthy market and
budgetary pressures. Unless officials feel fully accountable for policy-induced flaws in
risk-taking incentives, the net’s overall effectiveness is bound to be shortchanged. For
strategies of risk control to be maximally successful, employment contracts must assign
top regulators the duty of measuring and managing the social costs (and especially the
risk-taking incentives) generated by decisions about net design and incent all regulatory
personnel to perform these duties well. Just as differences in the size and shape of a
circus tent call for adjustments in the dimensions of the mesh and in the strength and
location of its supporting piers, every regulator’s rights and duties must be anchored in
the characteristics of the legal and private-information environment of the sponsoring
financial system and national government.
       Finally, the effectiveness of a safety net is never perfect, nor is it constant over
time. Authorities must understand that, on occasion, even the best-designed safety net

                                        Page 4 of 49
                                                                                 July 8, 2004

may 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
evolving crisis-management protocols. Paradoxically, unless a strong safety net is
backed up by solid disaster planning, its very strength may foster a series of infrequent,
but devastating crises. This is because, 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 crisis pressures.
       Considering how and where the circus metaphor breaks down underscores an
additional lesson. The major difference is that—unlike the splattering of an unlucky,
incompetent, or overly well-connected acrobat—breakdowns in financial safety nets are
not immediately visible to the naked eye. This lack of transparency intensifies the
conflict between the tasks of maximizing a net’s effectiveness and minimizing its costs.
To protect the reputation of their agency or top officers during their particular terms in
office, self-interested regulators may be tempted to conceal and sugarcoat information
about emerging or accelerating difficulties. As short-timers, they may callously fail to
challenge misleading bank condition reports or even encourage disinformation to be
entered on balance sheets and income statements. Precisely because top regulators do not
want their professional reputation besmirched by blame for financial-institution failures,
their authority over reporting protocols can be abused. Having the ability to reduce
transparency supports an incentive to delay insolvency resolution in tough times or tough
cases, so as to allow banks whose insolvency cannot yet be publicly recognized an
opportunity to gamble for resurrection at taxpayer and competitor expense.
       Knowing that regulators can block the flow of adverse information and that they
dislike public criticism tempts managers and owners of insolvent institutions to exploit
regulators’ incentive conflicts. Unscrupulous bankers routinely praise officials for
neglecting their duty of truth-telling. To justify their willingness to cover up bank
weakness, regulators claim they have the right to conceal or mischaracterize evidence of
widespread losses whenever they believe that forthright statements might generate a
financial panic. But public fears could be calmed without postponing cost-effective
action designed to preserve banking solvency. An explanation for ignoring banking losses
is that in some cases top officials may be trying to achieve a reputationally clean

                                        Page 5 of 49
                                                                                  July 8, 2004

getaway. Depending on the quality of a nation’s information and bureaucratic
environments, high officials may hope to escape blame for incurring insolvency-
resolution costs on their watch by doctoring and suppressing evidence long enough to
pass the bill for safety-net losses onto the next generation of regulatory officials.
Recognizing the availability of this escape hatch undermines the urgency of promptly
understanding newly emerging forms of risk-taking and engaging in extensive crisis
planning. In turns, underinvesting in these activities disposes regulators to treat future
financial breakdowns as if they were unique events that they can plan to handle in an ad
hoc manner.

II. Regulatory Culture and the Contracting Environment
       Modern finance theory emphasizes that bank depositors must worry about
controlling incentives for opportunistic behavior by their bank’s managers, owners, and
borrowers (Jensen and Meckling, 1976; Diamond, 1984; La Porta, Lopez-de-Silanes,
Shleifer, and Vishny, 1998). Managerial opportunism has three 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. difficulties an individual depositor faces in adequately analyzing and
           responding to whatever information its monitoring activity turns up (policing
       3. difficulties depositors face in coordinating collective action (coordination
In most information and contracting environments, by stepping into depositors’ shoes,
government regulators can reduce these costs.
       For this reason, 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

                                        Page 6 of 49
                                                                                 July 8, 2004

banks to misrepresent their economic condition to depositors and to pursue profit-making
opportunities that might exploit depositors’ informational disadvantage.
Dimensions of Regulatory Culture
       Financial regulation and supervision are cooperative endeavors. Anyone can
propose rules, but to command compliance requires legitimacy: a mutual understanding
that the rules are conceived and enforced to increase social welfare.
       To win and sustain a legitimate right to wield coercive force, regulators must
accept and respect appropriate checks on their authority. An “appropriate” check is one
that is consistent with the country’s political culture and its citizens’ understanding of the
country’s past regulatory experience.
       A culture is defined as customs, ideas, and attitudes that members share and
transmit from generation by systems of subtle and unsubtle rewards and punishments.
Carnell (1993) and Kane (2003) assign regulatory culture the role of defining,
authenticating, and promoting the financial common good. One country’s regulatory
culture may differ from another in any of six dimensions:
   1. in the character of the statutory grant of authority and the reporting
       responsibilities a regulatory enterprise receives;
   2. in the specific rules the enterprise formulates and how it develops and
       promulgates them;
   3. in the methods the enterprise uses to monitor for violations;
   4. in the penalties it can and does impose on clients when it finds material violations;
   5. in the nature and extent of due-process restrictions (including specific burdens of
       proof) that protect regulated institutions from unfair administrative procedures;
   6. in the extent of insured institutions’ rights to appeal regulatory decisions to a
       higher authority.
To be viable, the taboos and traditions incorporated in a regulatory culture must embody
community standards of fair play and proper use of government power.
Transparency and Deterrency
       Regulators’ tools of damage control are rulemaking and enforcement. To
understand the economic role these tools play, it is helpful to imagine a world in which
depositors’ monitoring and policing costs would be uniformly zero. In this world, each

                                        Page 7 of 49
                                                                                  July 8, 2004

deposit contract would be self-enforcing. Coordination costs would be irrelevant.
Establishing a team of centralized monitors and enforcers to thwart risk-shifting and
misconduct by bank insiders would offer no incremental 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
(MT) 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 maximal deterrency (MD) to describe a situation
in which depositors would immediately and perfectly understand the implications of
information flows 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
and maximal deterrency, the less ex ante 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 is implicitly a five-party contract. The net imposes mutual rights and duties on:
bankers, borrowers, depositors, safety-net managers, and suppliers of safety-net capital
(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

                                        Page 8 of 49
                                                                                July 8, 2004

counterparties and the efficiency with which they manage the diverse social costs of
coping with divergences from MT and MD conditions.
       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 financial triage function of shielding solvent, but illiquid
institutions from having to sell assets into momentarily disorderly markets.
       In practice, a safety-net manager must assemble a staff that can 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 delimit the net managers’ authority to regulate the bank. Along
with banks’ rights to challenge and appeal adverse actions, the last two categories provide
credible ways for regulators to bond themselves to exercise their supervisory authority in
the joint interests of banks and their various creditors. To complete the web of social
contract enforcement, the suppliers of regulatory risk capital—banks and 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 reasonable
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 and the exercise of appeal rights by regulated institutions would
otherwise tend to make regulatory personnel overly responsive to bank and depositor

                                        Page 9 of 49
                                                                                    July 8, 2004

        Theoretically, a nation’s safety net is a multiparty web of contractual duties and
obligations. Its most palpable features are deposit guarantees and lender-of-last-resort
credit facilities. The ideal safety net is one that efficiently mitigates the particular
monitoring, policing, and coordination 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 regulation and supervision to establish incentives for banks and regulators
that are compatible with the interests of all other parties, net design must be environment-
specific. One issue is how transparent information systems and supervisory technology
for monitoring bank capital and risk exposures are, either to outside experts or to the
financial press. A second issue is the extent to which regulator incentives lead to patterns
of discipline that reinforce or displace 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 the 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

                                         Page 10 of 49
                                                                                 July 8, 2004

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:” professional accountants, credit bureaus, credit-rating agencies, an
independent financial press, investment advisors, and even academic researchers.
However, even in high-income countries, the information-verification mission of these
watchdogs often conflicts with their other economic interests. Elsewhere,
interinstitutional competition is usually weak, reporting standards relatively
uninformative, and validity checks on bank and borrower disclosures allow many
informational impurities to survive the data-testing process.
       Recapping the discussion, the quality of regulation varies across countries with
informational transparency (T). In turn, T varies with accounting integrity (AI), 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 the governance
and information environments for depositors and taxpayers in different countries. Table
1 reproduces 1990-1998 measures of the relative informativeness of a country’s
accounting standards, the degree of corruption observed in government or business

                                       Page 11 of 49
                                                                                   July 8, 2004

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 was available for 73 countries) shows an r=.67 with per capita
       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 the level of economic and financial development as an
imperfect control for evolving social and cultural attitudes that strengthen the
enforceability of financial contracts.
       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.
Evidence of Variation in Depositors’ Deterrent Capacity

                                         Page 12 of 49
                                                                                 July 8, 2004

       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 the 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 to default. Deterrency (D) depends on a country’s systems
for policing corporate and public governance (G):
                                       D = D(G, T).                                      (2)
       Weaknesses in D disadvantage three groups with stakes in the safety net: banks as
lenders; depositors as bank creditors; and all suppliers of safety-net capital. In low-
deterrency environments, a rational saver will be reluctant to trust its funds to unrelated
parties, unless the borrower bonds its repayment capacity in a credible way. Table 2
ranks 195 countries according to the 2002 level of a World Bank index of their ability to
control corruption. Data for 1996, 1998, and 2000 are included to show that most
countries’ average rating remains in the same quartile over time.
       As we observed with the indexes of informational integrity arrayed in Table 1,
five other measures of deterrent protections that are posted on the World Bank website
are highly correlated with one another and with this index. As with transparency, the
level of deterrency increases on average as per capita income rises. Both types of
collinearity reinforce our contention that unmeasured socio-cultural norms and freedoms
explain differences in transparency and deterrency. This encourages us to investigate
how differences in the quality of a country’s corruption control and level of development
affect safety-net design.
        A climate of corruption makes it simultaneously harder for depositors to monitor
insolvent banks and harder for taxpayers and solvent banks to monitor regulators. Used
as a proxy for informational transparency, the index of corruption control can help us to
explore how increases in informational reliability affect the dimensions of the safety net.

                                       Page 13 of 49
                                                                               July 8, 2004

However, viewed as a proxy for deterrency and public accountability, enhancements in
corruption control have contradictory effects. This because, in lessening banking
fragility, effective corruption control simultaneously reduces the costs and the benefits of
safety-net protection.
Evidence of Cross-Country Variation in Accountability
       No existing dataset 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 engage in circumventive behavior that both limits
and underscores the potential damage that inefficient or unfair policies might otherwise
generate. Finally, for central-bank officials, legal independence reduces political
subservience. Cukierman et al. (1992, pp. 380-381) measure 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 taxpayers 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 correlate significantly with per capita GDP, with correlation coefficients
running as high as .70. Increases in central-bank independence and decreases in
economic repression make it easier for suppliers of safety-net capital to hold safety-net
managers accountable for the costs of the policies they follow.
Role of Foreign-Bank Entry
       For 72 countries for which banking-market indexes existed in 1999, the three-firm
banking concentration ratio correlates positively with economic repression (r ≈ .30 for
five of the Heritage Foundation’s indexes). This correlation between banking-market

                                       Page 14 of 49
                                                                                 July 8, 2004

concentration and economic repression suggests that a country’s banking markets are
more contestable when customers may transact more freely.
       In weak contracting environments, transparency, deterrency, and accountability
are enhanced by the entry of foreign banks. In countries where actual competitors are
few and entry protections have previously proved effective, 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. Given an
opportunity to extract monopoly rents from their customers, the net interest margin
between loan and deposit interest rates would be high. In today’s world, however, high
margins engender entry from abroad. Outside entry not only squeezes the benefits that
entry restrictions can deliver to local banks, it drives the value of these benefits below the
costs of cultivating the political influence necessary to support them.
       To persuade government officials to resist entry requires 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, 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 still produce. New products and communications channels expand
opportunities for customers and outside institutions to circumvent transaction controls.
Monopoly rents create incentives for customers and foreign banks to innovate around
government restrictions. Technological change has continually undermined the
effectiveness of entry barriers countries have previously used. In a world of rapid
technological change, officials can at best temporarily slow the rate of outside entry.
Thanks to the communications and information revolutions, globalization is wiping out
the capitalized net value of regulation-induced bank rents in high-barrier countries.

                                       Page 15 of 49
                                                                                 July 8, 2004

       Over time, entry tends to benefit host-country customers by providing access to
better-regulated banks and by pushing loan and deposit interest rates in their favor. This
path dependence makes it impossible to sign a priori the effect that current banking
concentration has on safety-net design. The more energetically authorities have tried to
limit banking competition to sustain bank profits in the past, the more rapidly they must
expect foreign and domestic nonbank competitors to take high-margin business away
from local banks in the future. Hence, high concentration simultaneously signals local
banks’ prior investment in political clout and their future vulnerability to global

III. Cross-Country Differences in Safety-Net Design Features
       In the absence of MT and MD, depositors must watch for harm from two
       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.
Sources of Implicit Coverage
       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 depositors may collect 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

                                        Page 16 of 49
                                                                                July 8, 2004

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.
       Politically enforced implicit coverage is 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
Eastern, 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, all of the
information measures, economic freedom, and various corporate-governance variables.
These correlations support the hypothesis that government banking presence tends to be
larger in environments where informational integrity, deterrent rights, and accountability
are weak.
       Implicit insurance also exists in explicit schemes, because formal limits on
government support can easily be ignored. Officials have the option to extend coverages
beyond statutory limits whenever that serves their bureaucratic or political interests.
Unfortunately, extracontractual coverages are often provided when market-mimicking
regulatory discipline would better promote taxpayer interests.
Spread of Explicit Schemes

                                       Page 17 of 49
                                                                                July 8, 2004

       Figure 1 shows that during the last four decades, explicit deposit insurance has
spread rapidly. Especially in Europe, socio-cultural expectations and cross-country
agreements have added deposit insurance to the mix of baseline governmental
responsibilities. Deposit-insurance guarantees sometimes protect even against the risk of
currency devaluation by covering accounts denominated in foreign currencies.
       Table 3 documents that, during the last 25 years, countries in different regions
adopted explicit deposit insurance in different degrees and in different ways. Figure 2
indicates that explicit insurance was part of the “best practices” policy standards that
were promulgated in IMF policy reports during the last half of the 1990s (Lindgren,
Garcia, and Saal, 1996 and Garcia, 1999). It is now widely understood that these
standards are not in fact “best practices” in environments where transparency, deterrency,
and accountability are weak (Demirgüç-Kunt and Kane, 2002).
       In poor informational and contracting environments, it is especially dangerous to
undermine depositors’ ethical responsibility to look out for themselves. The lack of
controls and the ambiguous and unfunded nature of purely implicit deposit insurance
leads depositors to demand a risk premium that is broadly commensurate with the risk-
taking capacity of their bank. By providing bureaucratic and political mechanisms for
patching weak banks, explicit guarantees make depositors less eager to gather
information about an institution’s financial condition and less likely to react promptly to
bad news about this condition. The more completely and reliably government or private
insurers attempt to “bulletproof” depositors against loss, the less incentive individual
depositors have to police the risks their banks can or do take.
       More than half the nations of the world still limit themselves to implicit
guarantees. Hovakimian, Kane, and Laeven (2003) show econometrically that, when
explicit guarantees are adopted, unfavorable country characteristics adversely influence
deposit-insurance design. For explicit deposit insurance to improve social welfare, the
regulatory culture must be strong enough to impel authorities to generate as much
supervisory discipline as the private discipline that deposit insurance displaces.
Anesthetization of depositor concerns can permit minor bank insolvencies to fester and
grow into hopeless insolvencies unless the regulatory culture focuses on addressing
supervisory problems promptly.

                                       Page 18 of 49
                                                                                  July 8, 2004

          Together, Tables 2 and 3 show that success in controlling corruption helps to
predict the introduction of explicit deposit insurance. However, corruption control is
neither a necessary nor a sufficient condition for including explicit guarantees in the
safety net. One-third of the countries in each of the lowest three quartiles of Table 2 offer
explicit guarantees, while one-third of the 40 countries with the least-corrupt
environments do not. However, the recent surge in adoption by lower-ranking countries
has been distorted by pressure from multinational organizations. For example, Eastern
European countries have been eager to meet criteria for membership in the European
Union and these criteria include explicit deposit insurance. Also, throughout the 1990s,
IMF personnel encouraged countries in Latin America and Asia that experienced banking
crises to craft explicit deposit insurance systems as a way to back away from unlimited
guarantees installed in the heat of crisis.
Special Dangers of Introducing Explicit Deposit Insurance in Difficult Circumstances
          In downturns, deposit insurance can block the exit of insolvent banks. Instead of
assuring the prompt exit of crippled or unprofitable firms, political pressure may urge
regulators to narrow the industry’s “exit drainpipe.” This is most likely to occur when
politicians are eager to preserve these institutions’ contribution to politically inspired
credit-allocation programs. Allowing unprofitable deposit institutions to issue guaranteed
deposits can artificially prolong their life at levels of net worth that would otherwise be
too low to sustain their existence. These deeply insolvent firms may be likened to
“zombies” in that they are kept alive by the black magic of government guarantees.
Zombie competitors are dangerous to society because they can bid down industry profit
margins to levels so low that even healthy and well-managed banks can no longer turn a
          In the United States, exit pressure that built up in the mid-1960s was not released
until well into the 1980s, when an explosion of belated deposit-institution exits occurred.
By then, most of the departing firms were zombie institutions whose insolvency could
and should have been resolved long ago. Some of the others were marginal institutions
that might have survived had safety-net managers not allowed industry profit margins and
risk-taking incentives to become so badly distorted.

                                         Page 19 of 49
                                                                                  July 8, 2004

        When deposit insurance retards the exit of poorly performing deposit institutions,
not only is new entry discouraged, but healthy competitors and taxpayers (as suppliers of
safety-net capital) are forced to back up high-risk gambles by crippled firms. Bailing out
zombie institutions distorts the stock of real capital by encouraging loans to high-risk
enterprises and assigns taxpayers an unbalanced option on industry profits. Taxpayers
are required to pay off future losses, but receive little opportunity to participate in gains.
The most that taxpayers can receive from a recovered institution is relief from the need to
continue to fund bailout arrangements.
        Explicit deposit insurance is especially dangerous when it is introduced in crisis
circumstances. Typically, adopting explicit insurance as an emergency measure serves 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 waste a nation’s scarce savings and reduce the present
discounted value of its aggregate stock of real capital.
        The problem is governmental myopia. The immediate benefits of a banking-
system rescue come at a time when the long-run control of supervisory and bank risk-
taking incentives has little urgency. Once a country’s banking system experiences large
losses, bank stakeholders lobby for a massive infusion of state resources. Introducing
explicit deposit insurance provides a convenient way for these stakeholders to mask the
size of the subsidies they extract. Government guarantees can rescue a deeply troubled
banking system without immediately requiring economic policymakers either to
recognize the size of bank losses or to impose new taxes. Because explicit insurance
reduces depositor concern for transparency and deterrency, opportunities for engaging in
unsound and corrupt banking practices tend to expand unless and until subsequent
reforms strengthen government banking supervision appropriately.

                                        Page 20 of 49
                                                                                  July 8, 2004

       Confirming these concerns, Demirgüç-Kunt and Detragiache (1998 and 2002)
find that, when they control statistically for the impact of exogenous crisis-generating
forces, the 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.
Desirable and Undesirable Safety-Net Features
       In principle, several deposit-insurance design features can constrain banks’ ability
to exploit weaknesses in transparency, deterrency, and accountability. Market discipline
can be generated by increasing the number of private parties that are responsible for
absorbing a “first-loss share” of the losses bank insolvencies entail. In practice, private
loss bearers are either very large depositors, bonding companies, or subordinated
debtholders. The value to society of incorporating this design feature turns on the
credibility of two expectations: (1) that government officials will force private parties to
live up to their contractual responsibilities and (2) that loss-sharing private parties will
not stand by if governmental forbearance exposes them to increasing risks.
       Several ways exist to expand first-loss exposure. One is to make private parties
underwrite and manage some of the loss exposures inherent in the deposit-insurance
system. For example, clearing members of stock and commodities exchanges operate a
private safety net in which they promise to absorb losses traders suffer due to failures in
settlement. In Table 3, when the column labeled “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. Tables 3 and 4 show that, of the 11

                                        Page 21 of 49
                                                                                   July 8, 2004

countries whose stock exchanges Siow and Aitken (2003) rate highest in efficiency and
integrity, seven incorporate privatization features in their banking safety net.
       A second way to enlist private parties in constraining bank risk shifting is to use
exclusions and coverage limits to render insurance coverage incomplete. Most countries
formally specify an upper limit to the size of deposit balance they protect. Relatively few
countries cover interbank deposits, but it is fairly common to insure accounts
denominated in foreign currency. A contractual device for controlling moral hazard
similar to exclusions and coverage limits that private insurance companies use is
coinsurance (Calomiris, 1998; Kane, 1995). Coinsurance requires that the insured party
bear a specified share of the value that is destroyed when a loss-causing event occurs.
       Although risk-control benefits of coverage limits and coinsurance require assured
enforcement, they can be realized without turning small depositors into loss-bearers. A
deposit insurer can exempt small accounts and structure larger depositors’ coinsurance
responsibility as a combination of percentage and fixed deductions from each
reimbursement claim. What matters is to impart to designated private stakeholders the
incentives and information they need to control bank risk-taking. Table 3 shows that only
about 15 countries currently ask depositors to coinsure safety-net losses.

IV. Matching 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.
       However, officials’ conscientiousness varies with cross-country differences in
transparency, deterrency, and accountability. Hovakinian, Kane, and Laeven show that,
in weak institutional environments, authorities are unlikely to adopt a mix of loss-sharing
rules, risk-sensitive premiums, and coverage limits strong enough to control moral
hazard. This implies that European Union, IMF, and World Bank personnel should

                                       Page 22 of 49
                                                                                  July 8, 2004

recommend changes in the structure of a country’s safety net only after carefully
analyzing how each change would affect transparency, deterrency, and accountability.
Accountability for Implicit Coverages as a Design Feature
       Taxpayers’ investment in the safety net consists of the contingent tax liability they
accept in explicitly and implicitly backing up the net’s guarantees. In the absence of
taxpayer back-up, private and government deposit-insurance managers would have to
expend additional resources to convince contractual counterparties that the enterprise can
be relied upon to fulfill its contractual commitments (Merton and Perold, 1993). The
capitalized value of the annual saving in enterprise expense may be defined as “risk
capital” that taxpayers contribute to the deposit-insurance system. Unless taxpayer-
contributed risk capital generates a fair return in the form of safety-net benefits to society,
deposit-insurance schemes unwisely subsidize bank risk-taking.
       In managing safety-net capital, regulators are pulled in contradictory directions.
On the one hand, they are expected to minimize the risk of a banking disaster. For this
reason, regulators who show mercy to a troubled bank garner public praise for themselves
and their agency. On the other hand, regulators are also expected to minimize the cost of
bailing out troubled banks and to subject all banks to market-mimicking discipline.
However, regulators who strongly and transparently discipline a weak bank risk being
blamed for aggravating the bank’s problems.
       Suppressing evidence of banking weakness and relaxing capital requirements is a
rational way to resolve this tension, especially if safety-net officials expect to hold office
briefly and can derive bureaucratic and reputational benefits from currying industry
support. This incentive conflict is not easily resolved. Unless regulatory decisions take
place in a MTMD environment for taxpayers, safety-net managers cannot be made fully
accountable in a timely manner for managing taxpayers’ economic stake in the safety net.
Though helpful, efforts to privatize the loss exposure can never be complete, because
even a privately managed and funded deposit-insurance scheme enjoys implicit
catastrophic taxpayer back-up. The taxpayer remains a silent partner whose investment
in the net is both unfunded and unlikely to be formally acknowledged by net managers.
       To the extent that the informational environment allows, it is important to make
specific officials responsible for tracking the aggregate losses to which the safety net

                                        Page 23 of 49
                                                                                  July 8, 2004

exposes taxpayers and for pricing and managing this exposure appropriately. However,
the difficulty of implementing this principle in environments where information is
unreliable and corruption is rampant implies that introducing explicit guarantees may end
up substituting corrupt government supervision for value-preserving private discipline.
       To guard against this unhappy result, political independence for safety-net
officials is not enough. In low-accountability environments, safety-net designers must
incorporate features that can generate strong private discipline on safety-net managers
and bankers alike. A good starting point is for authorities to take steps to make risk more
visible to outsiders at the individual-bank level. In particular, a positive obligation might
be placed on every insured bank’s top officers to personally certify the material accuracy
of marked-to-market estimates of bank net worth and to report promptly all substantial
changes in risk exposures and net worth to regulators. Where mark-to-market estimates
are hard to verify, insured banks should be asked to hold proportionately higher levels of
accounting capital and their officers should be subject to particularly severe penalties for
failing to disclose material adverse information.
       As better information on bank capital and loss exposure becomes available,
government regulators can be assigned two further specific tasks: (1) to calculate in a
reproducible manner an accounting estimate of the opportunity-cost value of deposit-
insurance risk exposures in insured banks, and (2) to tailor insurance premiums and
regulatory intervention to control the risk each bank passes on to the insurer.
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 they would forfeit if statistical analysis of
available data and subsequent events could prove that the agency’s risk calculations were
       It is also possible to lessen incentive conflicts by directly curtailing the benefits
that stockholders of insolvent institutions can extract by go-for-broke risk-taking. In
practice, the profitability of endgame plays can be blunted by extending stockholder
liability for liquidation losses beyond the level of the capital actually paid-in at the
corporate level. Several now-industrialized countries (including the U.K., the U.S. and

                                        Page 24 of 49
                                                                                 July 8, 2004

Canada) imposed extended liability on bank shares when their contacting environments
were more rudimentary.
        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 traded banks into a clearer signal of institutional strength or
weakness. Extended liability means that the insurer’s right to liquidate an insolvent bank
carries with it a right to collect additional funds from the personal assets of every
stockholder. As compared to limited-liability shareholding, deterrency is enhanced by
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. Stock markets
would imbed the value of this contingency into the price of each bank’s shares. The
value of the contingency would be negligible for banks that were performing well and
adequately supporting their risk with paid-in capital. However, the insurer’s right would
become increasingly valuable whenever a bank began to take poorly supported risks or to
slide into serious trouble. 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 in advance of their falling into
complete economic insolvency.
        This kind of stock sell-off would increase regulatory deterrency 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.
Moreover, in contracting environments where the chance that stockholders might over-
react to bad news is strong, the deterrent effect of extended liability would be particularly
        Extended liability would increase regulatory accountability because adverse stock
price movements would generate public pressure on regulators to investigate and take
remedial action. Extended liability would also raise the news value to outside
stockholders, taxpayers, and the financial press of staying abreast of 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 each

                                        Page 25 of 49
                                                                                          July 8, 2004

stockholder to bond its extended-liability obligation by depositing earning assets in a
collateral account at the central bank.1 Stockholders could be free to move individual
assets into and out of the collateral account over time as long as the aggregate market
value of the collateral account remains sufficient and the value of all posted assets 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 a stock sale.
        For any economy that ranks low in informational and ethical integrity, extended
liability provides an elegant economic solution to the problem of deposit-insurance risk-
shifting. It would improve the private and public contracting environments by
insinuating an observable and market-driven wedge between the economic interests of
ethical and unethical bankers and regulators. Of course, this is precisely why it is
politically difficult to build an effective constituency for it. One must expect bank
stockholders to resist efforts to narrow their risk-shifting capacity. One must also expect
corrupt regulators to defend limited liability and the kickbacks that risk subsidies
generate. Still, for conscientious officials, extended liability represents a workable
alternative to more intrusive government supervision of bank activity and is a strategy
that would work little harm on banks that expect to stay healthy.
Desirability of Rehearsing Crisis Management Procedures
        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 drilled in containing and
cleaning up 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 breaks down or threatens to break down,
similar problems of priority and coordination arise. Although the first line of crisis
management is an unstinting program of inspection, testing, and prompt repair, it must be
recognized that troubled banks have strong incentives to circumvent the prevention

 Where appropriate, this account might be required to include some kind of currency hedge and could be
held in safekeeping at the insured institution.

                                            Page 26 of 49
                                                                                   July 8, 2004

system. Whenever prevention fails to contain circumvention, a number of banks go splat.
The second line of crisis management is resolving the insolvency of damaged banks.
Insolvency resolution determines who loses what amounts when the net worth of many
banks is wiped out at the same time. For authorities to allocate losses efficiently, they
must establish defensible 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 exits, but specific staffmembers must take up the task of helping
depositors who want to extract their funds to do so in an informed and orderly manner.
The third priority is to clean up the mess so that the show may resume without an undue
         Rescue and Triage. In the midst of an emergency, triage decisions cannot wait for
formal ratification by less-informed higher-ups. Although staff judgments 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 bureaucratic decision-making to a decentralized
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.
         In any medical disaster, making decisions about the urgency of treating different
patients is called triage. Triage begins with evaluating each casualty’s particular needs.
This assessment seeks to determine which parties are and are not beyond help. Its second
objective is to set priorities for treating those that can benefit from specific assistance.
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 take on the roles of comforting moribund patients and helping
noncritical patients wait for treatment.
         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 they have formulated an

                                        Page 27 of 49
                                                                                  July 8, 2004

integrated disaster plan and drilled personnel in its execution. Regulatory staff must be
prepared to react 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 needed 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 different stakeholders
would require to make them whole. Without this information, higher officials cannot
evaluate the reasonableness of forcing taxpayers to supply that help.
       It is easy to see that it makes no sense in the midst of an emergency to divert
limited surgical resources to sewing up 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.
       Until the size of individual insolvencies can be sorted out, depositor access to
funds must be suspended. 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. 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.
Procedures for setting the transactable fraction of different deposit accounts should be
founded in conservative valuation techniques whose application is 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 is unfrozen in stages when and as the depth and intangible
elements of each bank’s insolvency can be sized more precisely.
       The most straightforward way to preserve the liquidity of bank depositors is to
assure insured and uninsured customers of each insolvent bank that arrangements are
being made to grant them direct or indirect access to central-bank funds based on the
recoverable value of their net claims on the bank and its insurer. Preparing examiners to

                                        Page 28 of 49
                                                                                 July 8, 2004

calculate promptly the value at which bank assets could be liquidated in an orderly
manner is the key step in restoring the liquidity of deposit accounts.
       Once examiners have made these quick-and-dirty calculations, depositors can be
granted at least 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 reliable information on asset
values can be developed.
       Crowd Control. Banking panics are triggered by adverse information that
destroys customer confidence in the repayment capacity of a group of banks. The trigger
may be information either about broad economic events, about specific banks, or about
assets banks are known to hold.
       A panic is defined by two conditions. First, in a panic, depositor runs are so
widespread that affected banks cannot raise funds quickly by selling liquid 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, an obvious concern is to halt the spread of illiquidity and financial
dislocation. Using the words “panic” and “depositor runs” to describe the onset of this
problem builds a subliminal case for authorities to throw themselves into rescue mode
without stopping to consider longer-run priorities. The urgency of stopping the spread of
confusion must not over-ride the simultaneous need to identify zombie institutions and
begin the process of winding up their affairs. Issuing blanket government loans and
guarantees to all troubled banks shifts the burden of absorbing the losses imbedded in
zombie portfolios to taxpayers and relieves managers, stockholders, and creditors from
bearing due responsibility for loss-making decisions they had effectively ratified.
       Injections of liquidity should be offered only in exchange for good assets and
control over zombie institutions must be put into new hands. This principle allows
authorities to maintain aggregate liquidity by open-market operations, but requires them

                                       Page 29 of 49
                                                                                July 8, 2004

to take the measure of a bank’s wounds before expanding its individual access to
emergency loans. It should be understood that insolvency-assessment timeouts will occur
and that, during these timeouts, would-be transactors can be expected to use standard and
innovative forms of credit to prevent transactions from grinding to a halt. Checks and
credit-card slips will still circulate. However, instead of being cleared promptly, they
will stand as evidence of personal indebtedness. When the timeout ends, if the issuing
bank cannot cover them, these debts may be discharged from other sources.
       The spread of confusion can and should be countered by an intelligent mix of
structural and macro policies. Macro control may be effected by avoiding monetary
contraction and by offering government loans and guarantees to certifiably healthy banks.
Structural policy measures should enlist nongovernmental parties in the tasks of
identifying viable institutions and of focusing liquidity assistance on them. Policymakers
should make sure that, in judging which banks and assets deserve to be shielded from the
spreading harm, authorities do not blunt the analytic capacities of nongovernmental
entities, customers, local clearinghouses, and foreign banks.
       Cleanup. Banking crises are the deep downside of a process of lending to parties
whose ability to repay cannot be accurately assessed in advance. A panic occurs when
profound weaknesses in customer repayment capacities surface so suddenly that large
losses need to be assigned not just to bank stockholders, but also to bank creditors and
their guarantors. To clean up the red ink that has been spilled, declines in bank asset
values must pass through the liability side of bank balance sheets. In designing an
interim treatment plan, the long-run cost of interfering with the transfer of a bank’s losses
to its creditors must be weighed against the benefits of bringing the panic to a quicker
       It is unreasonable to suppose that keeping stockholders in deeply insolvent banks
from paying for managerial mistakes is a desirable public policy. Inefficiently managing
a bank’s resources destroys its ownership capital. This loss of capital should unleash
market forces that transfer the valuable parts of any insolvent bank’s franchise into
better-capitalized and potentially more-skillful hands. Closing an insolvent bank or
assigning its business to a new owner is not a barbarously cruel thing to do. Unless
creditors and investors expect inefficient managers and undercapitalized firms to be

                                       Page 30 of 49
                                                                                   July 8, 2004

promptly and appropriately disciplined, the incentives that govern the evaluation and
selection of risky investment projects will break down. On average, the more fairly an
insolvent banking system’s losses can be allocated, the more likely it is that socially
desirable patterns of bank lending will be restored. The more efficiently and more fairly
the process of loss resolution is conducted, the smaller the long-run cost in economic
resources and social demoralization a country must pay to make its banking system
whole again.

V. Summary
        The risks banks take originate in the real economy. To allow banks to
intermediate a nation’s flow of savings and investment, depositors must be confident in
banks’ ability to monitor the quality of borrowers.
        Accumulating stockholder equity is a way for individual banks to bond the quality
of their monitoring activity, but it is not the only way. In particular, because
governments have an interest both in assuring the efficiency with which savings are
invested and in limiting the damage that bad banking may engender, governments erect a
financial safety net.
        Safety-net design must address differences in transparency, deterrency, and
accountability that exist across countries. The weaker is a country’s informational,
ethical, and corporate-governance environment, the greater the danger than a wholly
governmental system of explicit deposit guarantees will undermine bank safety and
        Barth, Caprio, and Nolle (2004) provide a snapshot of how banking and
regulatory environments differ across countries. The analysis presented here stresses that
the design features incorporated into a country’s net ought to evolve over time with
changes in private and government regulators’ capacity for valuing banking institutions,
for monitoring and disciplining risk-taking, for resolving insolvencies promptly, and for
being held accountable for how well they perform these tasks.
        Safety-net managers should neither subsidize nor tax bank risk-taking. No
reputable body of economic thought argues that taxing or subsidizing risk-taking is an
optimal way to run a developing nation’s economy. Even if a case for such a policy

                                       Page 31 of 49
                                                                             July 8, 2004

could be made, it would be inefficient to channel taxes and subsidies through a
government-run safety net whose managers are not required either to plan for disaster or
to account publicly for taxpayers’ stake in its operations.


Barth, James R., Gerard Caprio, and Daniel E. Nolle, 2004. “Comparative International
       Characteristics of Banking,” Washington: Comptroller of the Currency, Economic
       and Policy Analysis Working Paper 2004-1 (January).
Black, Fischer, Merton Miller, and Richard Posner, 1978. “An Approach to the
       Regulation of Bank Holding Companies,” Journal of Business, 5 (July), pp. 379-
Brock, Philip L., 1999. “Financial Safety Nets: Lessons from Chile.” Seattle: University
       of Washington, (March).
Calomiris, Charles W., 1999. “Building an Incentive-Compatible Safety Net,” Journal of
       Banking and Finance, 23(October), pp. 1499-1519.
_____, 1997. The Postmodern Safety Net: Lessons from Developed and Developing
       Economies. Washington: American Enterprise Institute.
Carnell, Richard, 1993. “The Culture of Ad Hoc Discretion,” in George Kaufman and
       Robert Litan (ed.), Assessing Bank Reform: FDICIA One Year Later,
       Washington: Brookings Institution, pp. 113-121.
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.
Demirgüç-Kunt, Asl, and Enrica Detragiache, 1998. “The Determinants of Banking
       Crises: Evidence from Developing and Developed Countries,” IMF Staff Papers,
       45 (no. 1), pp. 81-109.
_____, 1999a. “Financial Liberalization and Financial Fragility,” Proceedings of the
       Annual World Bank Conference on Development Economics, Washington: The
       World Bank, pp. 303-31.

                                       Page 32 of 49
                                                                            July 8, 2004

_____, 2002. “Does Deposit Insurance Increase Banking System Stability?: An
       Empirical Investigation,” Journal of Monetary Economics, 49, pp. 1373-1406.
Demirgüç-Kunt, Asle, and Edward J. Kane, 2002. Journal of Economic Perspectives, 16
       (No. 2), pp. 175-195.
Demirgüç-Kunt, Asl, and Tolga Tobac, 2000. “Deposit Insurance Around the World: A
       Data Base,” World Bank manuscript.
Diamond, Douglas, 1984. “Financial Intermediation and Delegated Monitoring,” Review
       of Economic Studies, 51(July), pp. 343-414.
Diamond, Douglas, and Philip Dybvig, 1983. “Bank Runs, Deposit Insurance, and
       Liquidity,” Journal of Political Economy, 91(June), pp. 401-19.
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.
Hovakimian, Armen, Edward Kane, and Luc Laeven, 2003. “How Country and Safety-
       Net Characteristics Affect Bank’s Risk-Shifting,” Journal of Financial Services
       Research, 23 (No. 2) pp. 177-204.
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., 2000. “Designing Financial Safety Nets to Fit Country
       Circumstances.” (unpublished).
_____, 1995. “Three Paradigms for the Role of Capitalization Requirements in Insured
       Financial Institutions,” Journal of Banking and Finance, 19 (June ), pp. 431-459.
__________, 2003. “What Kind of Multinational Deposit-Insurance Arrangements
       Might best Enhance World Welfare?,” Pacific-Basin Finance Journal, 11 (No. 4),
       pp. 413-428.
La Porta, Rafael, Florencio Lopez-de-Silanes, and Andrei Shleifer, 1999. “Government
       Ownership of Commercial Banks,” Harvard University manuscript (November).

                                     Page 33 of 49
                                                                           July 8, 2004

La Porta, Rafael; Florencio Lopez-de-Silanes, Andrei Shleifer, and Robert W. Vishny,
       1998. “Law and Finance,” Journal of Political Economy, 106 (December), pp.
Lindgren, Carl-Johan, Gillian Garcia, and Matthew I. Saal, 1996. Bank Soundness and
       Macroeconomic Policy. Washington: International Monetary Fund.
Merton, Robert C., and André 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.
Siow, Audris S. and Aitken, Michael J., "Ranking World Equity Markets on the Basis of
       Market Efficiency and Integrity," The HP Handbook of World Stock, Derivative
       & Commodity Exchanges 2003, Herbie Skeete, ed., pp. xlix-lv, Mondo Visione
       Ltd., 2003
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.

                                     Page 34 of 49
                                                                                                                                                                      July 8, 2004

                                                     Figure 1: Cross-Country Trend in the Adoption of Explicit Deposit Insurance

                                85                                                                                                                                              83
                                80                                                                                                                                   77 78
                                65                                                                                                                         63

                                                                                                                                                   56 57
Number of Explicit DI Systems

                                45                                                                                                  43
                                35                                                                              32
                                30                                                                  27
                                25                                                            22
                                                                                      19 20
                                20                                              17 18
                                                                        14 15
                                15                                12 13
                                                         9   10
                                10                   7
                                         3   4

                                                                                        Number of DI Systems

                                                                                      Page 35 of 49
                                                                   July 8, 2004

                       FIGURE 2

 All Countries Should Establish Explicit Deposit Insurance

 The Insurance System Should Incorporate at Least the Following Design
   Prudential regulation
   Limitations on coverage
   Mandatory membership
   Political “independence” for regulatory officials.

                                  Page 36 of 49
                July 8, 2004

Page 37 of 49
                                                                                                             July 8, 2004

  Table 2: Control of Corruption in 195 Countries Between 1996 and 2002, Listed in Order of Percentile Rank in 2002

Country                                                  2002               2000              1998           1996

FINLAND                                                  100                100               98.9            99.3
SINGAPORE                                                99.5               99.5              97.3            97.3
NEW ZEALAND                                               99                97.3              97.8            98.7
DENMARK                                                  98.5               97.8              99.5            100
SWEDEN                                                   97.9               98.9              98.4             98
ICELAND                                                  97.4               98.4              95.1            91.3
SWITZERLAND                                              96.9               95.7              100             95.3
NETHERLANDS                                              96.4               96.7              96.2             96
CANADA                                                   95.9               96.2              96.7            96.7
LUXEMBOURG                                               95.4                94               92.9             92
NORWAY                                                   94.8               94.6              95.6            94.7
UNITED KINGDOM                                           94.3               95.1              94.5             94
AUSTRALIA                                                93.8               93.5              93.4            93.3
AUSTRIA                                                  93.3               92.9              91.8            89.3
GERMANY                                                  92.8               91.8               94             90.7
UNITED STATES                                            92.3               92.4              91.3             90
IRELAND                                                  91.8               90.8              92.3            92.7
BELGIUM                                                  91.2                87               86.3            83.3
CHILE                                                    90.7               90.2              85.8             86
HONG KONG                                                90.2               89.1              90.2             88
SPAIN                                                    89.7               91.3              89.6            81.3
FRANCE                                                   89.2               89.7              90.7            86.7
BAHAMAS                                                  88.7               81.5              78.7             66
PORTUGAL                                                 88.1               88.6              89.1            85.3
ANDORRA                                                  85.6               N/A               N/A             N/A
BARBADOS                                                 85.6               N/A               N/A             N/A
BERMUDA                                                  85.6               N/A               N/A             N/A
CAYMAN ISLANDS                                           85.6               N/A               N/A             N/A
LIECHTENSTEIN                                            85.6               N/A               N/A             N/A
JAPAN                                                    85.1               87.5              86.9            84.7
PUERTO RICO                                              84.5                88               88.5             84
UNITED ARAB EMIRATES                                      84                76.1              81.4             64
ISRAEL                                                   83.5               85.9               88             87.3
KUWAIT                                                    83                82.1              85.2            78.7
OMAN                                                     82.5               78.3              83.6            63.3
BAHRAIN                                                   82                70.7              73.8             62
QATAR                                                    81.4               77.2               82             57.3
BHUTAN                                                   80.9               86.4              75.4            N/A
SLOVENIA                                                 80.4               84.2              82.5            82.7
CYPRUS                                                   79.9               84.8              87.4            88.7
COSTA RICA                                               79.4               83.2              80.3            80.7
ANTIGUA AND BARBUDA                                      77.8               N/A               N/A             N/A
FRENCH GUIANA                                            77.8               N/A               N/A             N/A
MARTINIQUE                                               77.8               N/A               N/A             N/A
TAIWAN                                                   77.3               77.7              84.2             80

                                                    Page 38 of 49
                                                               July 8, 2004

MALTA                                76.8        68.5   78.7    66
ITALY                                76.3        82.6   84.7   74.7
URUGUAY                              75.8        78.8   74.9    74
BOTSWANA                             75.3        83.7   78.1   71.3
GRENADA                              74.7         63    63.4   N/A
ESTONIA                              74.2        79.3   76.5   60.7
HUNGARY                              73.7        79.9   79.8    78
GREECE                               73.2         81    83.1   70.7
SAUDI ARABIA                         72.7        60.9   72.7    42
MAURITIUS                            72.2         75     71    75.3
DOMINICA                             71.1        48.9   48.6   N/A
SEYCHELLES                           71.1         63    48.6   N/A
ST. KITTS AND NEVIS                  69.6         63    63.4   N/A
ST. LUCIA                            69.6        75.5   63.4   N/A
ST. VINCENT AND THE GRENADINES       69.6         63    63.4   N/A
POLAND                               69.1        73.4    77    72.7
CZECH REPUBLIC                       68.6        72.3   73.2   77.3
MALAYSIA                              68         67.9   80.9    76
SOUTH AFRICA                         67.5        73.9   74.3   79.3
TUNISIA                               67         76.6   68.3    58
KOREA, SOUTH                         66.5        72.8   69.9   76.7
CAPE VERDE                            66          63    48.6   N/A
BRUNEI                               65.5        55.4   66.7    66
SLOVAK REPUBLIC                      64.9        69.6   62.8   73.3
LITHUANIA                            64.4        70.1   67.8   53.3
CROATIA                              63.9        60.3   46.4   34.7
MAURITANIA                           63.4        26.6   48.6   N/A
NAMIBIA                              62.9        85.3   72.1    82
SURINAME                             62.4        68.5   66.7    44
MADAGASCAR                           61.9        21.2   15.3    66
FIJI                                 61.3        74.5   70.5   N/A
LATVIA                               60.8        59.2   61.7   31.3
ERITREA                              60.3        58.2   75.4   N/A
MALDIVES                             59.8        33.2   34.4   N/A
JORDAN                               59.3         62    71.6   56.7
MARSHALL ISLANDS                     58.8         63    34.4   N/A
MOROCCO                              58.2        71.2   62.3   64.7
BURKINA FASO                         57.7         25    38.8    44
TRINIDAD AND TOBAGO                  57.2        71.7   69.4   65.3
BRAZIL                               56.7        59.8   68.9   55.3
SAMOA                                56.2         63    48.6   N/A
MACAO                                55.7        N/A    N/A    N/A
CUBA                                 55.2        45.7   53.6   59.3
SRI LANKA                            54.6        58.7   57.4    50
MONGOLIA                             54.1        41.8   54.6    66
THAILAND                             53.6        46.2   61.2   42.7
SENEGAL                              53.1        42.9   41.5   38.7
BULGARIA                             52.6        54.9   39.9   29.3
MEXICO                               52.1         44     41    39.3
PERU                                 51.5        56.5   58.5    56

                                 Page 39 of 49
                                                      July 8, 2004

PANAMA                       51         45.1   54.1   33.3
BELIZE                       50         62.5   48.6   N/A
SAO TOME AND PRINCIPE        50          63    15.8   N/A
INDIA                       49.5        52.7   60.1   43.3
SWAZILAND                    49         48.9    59    N/A
LESOTHO                     48.5         63    66.1   N/A
EGYPT                       47.9        54.3   56.3   62.7
SYRIA                       47.4        23.9   31.7   28.7
NEPAL                       46.9        40.2   30.1    48
MALI                        46.4        35.9   31.1    44
LEBANON                     45.9        36.4    47     52
ROMANIA                     45.4         38    44.3   51.3
ETHIOPIA                    44.8        57.1   56.8   12.7
IRAN                        44.3        32.6   27.9   23.3
TURKEY                      43.8        47.8   65.6   61.3
DOMINICAN REPUBLIC          43.3        47.3   37.7   40.7
GHANA                       42.8        40.8   43.2    36
CHINA                       42.3        46.7   57.9   58.7
KIRIBATI                    40.2        33.2   34.4   N/A
MICRONESIA                  40.2        48.9   48.6   N/A
TONGA                       40.2        48.9   48.6   N/A
VANUATU                     40.2        48.9   48.6   N/A
NICARAGUA                   39.7        17.4   25.7   52.7
JAMAICA                     39.2        53.3   55.7   40.7
COLOMBIA                    38.7        39.7   30.6   37.3
GUYANA                      38.1        43.5   55.2    44
PHILIPPINES                 37.6        37.5   45.9    38
TIMOR, EAST                 37.1        N/A    N/A    N/A
EL SALVADOR                 36.6        53.8   47.5   25.3
GABON                       36.1        25.5   11.5    4.7
GUINEA                      35.6        41.3   14.8    66
RWANDA                      35.1        61.4   34.4   N/A
BOSNIA-HERZEGOVINA          34.5         37    45.4   N/A
BENIN                        34         48.9   19.7   N/A
GUINEA-BISSAU               33.5        42.4   32.2   12.7
VIETNAM                      33         23.4   28.4   30.7
TOGO                        32.5        28.8   42.6   12.7
YEMEN                        32         26.1   32.8   49.3
ALGERIA                     31.4        29.9   25.1    40
GUATEMALA                   30.9        29.3   23.5    18
ARMENIA                     30.4        22.8    23     30
PAKISTAN                    29.9        27.2   20.2   15.3
MACEDONIA                   29.4        38.6   48.1   10.7
COMOROS                     28.4        33.2   15.8   N/A
DJIBOUTI                    28.4        12.5   15.8   N/A
ARGENTINA                   27.8        44.6   59.6    54
HONDURAS                    27.3        28.3   21.9   17.3
BELARUS                     26.8        57.6   29.5    20
YUGOSLAVIA                  26.3         9.8    8.2   20.7
SIERRA LEONE                25.8        22.3   22.4    0.7

                        Page 40 of 49
                                                                                                                July 8, 2004

BOLIVIA                                                         25.3                 27.7               42.1     22
LIBYA                                                           24.7                 15.2               10.9    21.3
GAMBIA                                                          24.2                  56                40.4     66
KYRGYZ REPUBLIC                                                 23.7                 20.1               26.2    24.7
ALBANIA                                                         23.2                 32.1                9.8     60
COTE D'IVOIRE                                                   22.7                 31.5               44.8     72
SOLOMON ISLANDS                                                 22.2                 48.9               34.4    N/A
MOLDOVA                                                         21.6                 19.6               38.3    50.7
RUSSIA                                                          21.1                 10.3               26.8    27.3
PAPUA NEW GUINEA                                                20.6                  8.2                24     48.7
CAMBODIA                                                        20.1                 35.3                2.7    18.7
MALAWI                                                          19.6                 48.4               39.3     12
UGANDA                                                          19.1                 18.5                29      32
VENEZUELA                                                       18.6                  31                21.3     28
CONGO                                                            18                  16.3                7.1     24
UKRAINE                                                         17.5                 14.1                12     26.7
ZAMBIA                                                           17                  20.7               33.3     16
LIBERIA                                                         16.5                  3.8                0.5     0.7
WEST BANK/GAZA                                                   16                  80.4               77.6    N/A
TANZANIA                                                        15.5                  12                 8.7     9.3
MOZAMBIQUE                                                      14.9                 39.1               18.6    32.7
ECUADOR                                                         14.4                 13.6               19.1     26
BURUNDI                                                         12.9                  2.7               15.8    N/A
CENTRAL AFRICAN REPUBLIC                                        12.9                 12.5               34.4    N/A
CHAD                                                            12.9                 33.2               14.2    N/A
GEORGIA                                                         12.4                 24.5               27.3      8
UZBEKISTAN                                                      11.9                 21.7                7.7    11.3
KENYA                                                           11.3                  9.2               10.4     8.7
KAZAKHSTAN                                                      10.8                  19                13.1    22.7
TAJIKISTAN                                                      10.3                  4.9                3.8     2.7
AZERBAIJAN                                                       9.8                  6.5                5.5    16.7
SUDAN                                                            9.3                   6                20.8     7.3
CAMEROON                                                         8.8                  7.6                4.4     6.7
NIGER                                                            8.2                 16.8               12.6     44
BANGLADESH                                                       7.7                 30.4               43.7    36.7
ANGOLA                                                           7.2                  2.2                4.9     10
INDONESIA                                                        6.7                  8.7                6.6    35.3
ZIMBABWE                                                         6.2                 15.8               60.7    54.7
KOREA, NORTH                                                     5.7                 14.7               33.9     44
SOMALIA                                                          5.2                  0.5                0.5     0.7
TURKMENISTAN                                                     4.6                  5.4                3.3     3.3
PARAGUAY                                                         4.1                 10.9                9.3     34
LAOS                                                             3.6                 17.9               24.6    18.7
NIGERIA                                                          3.1                  7.1                 6      5.3
AFGHANISTAN                                                      2.6                  1.1               N/A     N/A
MYANMAR                                                          2.1                  3.3                2.2      6
CONGO, DEM. REP.                                                 1.5                  1.6                 0       0
IRAQ                                                              1                   4.3                1.6      4
HAITI                                                            0.5                 11.4               13.7    12.7
EQUATORIAL GUINEA                                                 0                    0                15.8    N/A
Source: World Bank’s Governance Databank, accessible at

                                                           Page 41 of 49
                                                                                                                                          July 8, 2004

                                            Table 3: Cross-Country Variation in Deposit-Insurance Design Features
    DEPOSIT         Date
                                                                            Premium or        Annual                        Foreign   Interbank     Source of
  INSURANCE       Enacted/   Membership      Administration    Funding                                   Coverage Limit
                                                                          Assessment Base    Premium                       Currency   Deposits      Funding
   SYSTEMS        Revised
                                                                                                         in ecu (euros)                           0 = Banks
                                               official=1                                   % of base    or U.S. dollars                             Only
                                                                                                                                                  1= Banks &
                             compulsory=1       joint=2       funded=1                                                      yes=1      yes=1         Gov.
                                                              unfunded=                                                                           Government
                             voluntary=0       private=3          0                                                         no=0       no=0          Only
                                                                                             0.15% of
                                                                            deposits and    deposits +
                                                                           nonperforming     0.5% of
Cameroon           1999           0                2             1             loans           npls         $5,336            0          1               1
                                                                                             0.15% of
                                                                            deposits and    deposits +
Central African                                                            nonperforming     0.5% of
Rep.               1999           0                2             1             loans           npls         $3,557            0          1               1
                                                                                             0.15% of
                                                                            deposits and    deposits +
                                                                           nonperforming     0.5% of
Chad               1999           0                2             1             loans           npls         $3,557            0          1               1
                                                                                             0.15% of
                                                                            deposits and    deposits +
                                                                           nonperforming     0.5% of
Congo              1999           0                2             1             loans           npls         $3,557            0          1               1
                                                                                             0.15% of
                                                                            deposits and    deposits +
Equatorial                                                                 nonperforming     0.5% of
Guinea             1999           0                2             1             loans           npls         $3,557            0          1               1
                                                                                             0.15% of
                                                                            deposits and    deposits +
                                                                           nonperforming     0.5% of
Gabon              1999           0                2             1             loans           npls         $5,336            0          1               1
Kenya              1985           1                1             1            deposits         0.15         $1,750            1          1               1
Nigeria            1988           1                1             1           deposits        0.9375      $588 / $2435         0          1               1

Tanzania           1994           1                3             1           deposits          0.1           $376             0          0               1

                                                                          Page 42 of 49
                                                                                                                                     July 8, 2004

Uganda          1994          1              1            1          deposits          0.2            $2,310             0          0               1

                                                                    Premium or        Annual                           Foreign   Interbank     Source of
               Enacted/   Membership   Administration   Funding                                   Coverage Limit
    DEPOSIT                                                       Assessment Base    Premium                          Currency   Deposits      Funding

Bangladesh      1984          1              1            1          deposits         0.005           $2,123             0          0               1
India           1961          1              1            1          deposits         0.005           $2,355             1          0               1
                                                                                                   $14,600 but
                                                                                                  in full until the
South Korea     1996          1              1            1          deposits           0.05        year 2000            0          0               1
Marshall                                                                              0.00 to
Islands         1975          0              1            1          deposits           0.27        $100,000             1          1               0
                                                                                      0.00 to
Micronesia      1963          0              1            1          deposits           0.28        $100,000             1          1               0

Philippines     1963          1              1            1          deposits           0.2           $2,375             1          1               1
Sri Lanka       1987          0              1            1          deposits          0.15           $1,470             0          0               1
Taiwan          1985          0              1            1          deposits         0.015          $38,500             0          0               1
Vietnam         2000

Albania         2002
Belarus                      n.a.          n.a.           1            n.a.             n.a.           2000             n.a.       n.a.             0
                                                                                     0.5% until
                                                                                    July 2001,
Bosnia-                                                                             changed to
Herzegovina     1998          1              1            1          deposits          0.3%            2764             n.a.       n.a.             1

                                                                  Page 43 of 49
                                                                                                                                 July 8, 2004

                                                                   Premium or        Annual                        Foreign   Interbank     Source of
              Enacted/   Membership   Administration   Funding                                   Coverage Limit
                                                                 Assessment Base    Premium                       Currency   Deposits      Funding
    DEPOSIT   Revised
                                                                    insured        risk based
Bulgaria       1995          1              2            1          deposits          to 0.5        $1,784           1          0               1
Croatia        1997          1              2            1          deposits           0.8         $15,300           1          0               1
                                                                                   banks: 0.5,
                                                                    insured          savings     coinsurance
Czech Rep.     1994          1              1            1          deposits        banks 0.1      to 11756          0          0               1
                                                                                                 90% of 1383,
                                                                  deposits until                  but ecu in
Estonia        1998          1              2            1            2002          0.5 (max)        2010            1          0               1
                                                                    insured        risk based    ecu 44165 or
Hungary        1993          1              2            1          deposits          to 0.3        $4565            1          0               1
Kazakstan      2003          1              1            1            n.a.             1%            2599            1         n.a.             1
                                                                    insured                       $830 intil
Latvia         1998          1              1            1          deposits           0.3        year 2000          1          0               1
                                                                    insured                      $6250 then
Lithuania      1996          1              1            1          deposits           1.5       coinsurance         1          0               1
                                                                                    1.5, risk-
                                                                    insured        based 1%      coinsurance
Macedonia      1996          0              2            1          deposits         to 5%       75% to $183         1          0               1
                                                                                                   ecu 1000,
                                                                                                   then 90%
                                                                  deposits, also                 coinsurance
                                                                  risk-adjusted     not more      for the next
Poland         1995          1              1            1            assets         than 0.4      ecu 4000          1          0               1
                                                                     insured       risk-based
Romania        1996          1              2            1           deposits       0.3 to 0.6      $3,600           1          0               1
Russia         2003
Serbia and
Montenegro     2001
Slovak                                                              insured         0.1 to 0.3
Republic       1996          1              2            1          deposits        for banks       $7,900           1          0               1
Slovenia       2003

                                                                 Page 44 of 49
                                                                                                                                   July 8, 2004

                                                                     Premium or        Annual                        Foreign   Interbank     Source of
                Enacted/   Membership   Administration   Funding                                   Coverage Limit
    DEPOSIT                                                        Assessment Base    Premium                       Currency   Deposits      Funding
                                                                                      0.5 plus
Ukraine          1998          1              1            1        total deposits    charges          $250            1          0                1

Middle East
Bahrain          1993          1              2            0          deposits        ex post         $5,640           1          0                0
Cyprus           2000
Jordan           2000
Kuwait                        n.a.          n.a.          n.a.           n.a.           n.a.           n.a.           n.a.       n.a.             n.a.
Lebanon          1967          1             2             1       credit accounts      0.05          $3,300           0          1                1
                                                                                                     75% to
Oman             1995          1              1            1           deposits          0.02        $52,630           1          0                1
                                                                   insured savings   risk-based
Turkey           1983          1              1            1           deposits       1.0 to 1.2       in full         1          0                1

Latin America
and the
                 1971,                                                               risk-based,
                 1979,                                                insured           0.36 to
Argentina        1995          1              3            1          deposits           0.72         30,000           1          0                0
Bahamas          1,999         1              1            1          deposits       120 of 1%        40,816           0          0                0
Bolivia          1999
Brazil           1995          1              3            1          deposits          0.3           $17,000          1          0                0
                                                                                                    deposits in
                                                                                                   full and 90%
                                                                                                   to UF 120 of
                                                                                                    $3,600 for
Chile            1986          1              1            0        not applicable      none         deposits          1          0                2

                                                                   Page 45 of 49
                                                                                                                                      July 8, 2004

                                                                      Premium or         Annual                         Foreign   Interbank     Source of
                 Enacted/   Membership   Administration   Funding                                    Coverage Limit
    DEPOSIT                                                         Assessment Base     Premium                        Currency   Deposits      Funding
                                                                                                       in full until
                                                                                                       2001, then
                                                                       insured                        coinsurance
Colombia          1985          1              1            1          deposits           0.3           to $5,500         0          1                0
Dominican                                                                                             coinsurance
Republic          1962          0              2            1          deposits          0.1875        to $13000          1          0                1
                                                                                                     in full to year
Ecuador           1999          1              1            1          deposits            0.65            2001           1          1               n.a.
                                                                       insured         risk-based,
El Salvador       1999          1              1            1          deposits         0.1 to 0.3     C 30,000           1          0                1
Jamaica           1998          1              1            1          deposits           0.1            $5,512           1          0                1
                                                                                                     in full except
                                                                                        0.3 (max)    subordinated
                                                                                       plus 0.7 as     debt until
Mexico            1986          1              1            1        all obligations     needed           2005            1          1                1
Paraguay          2003
                                                                       insured          from 0.65
Peru              1992          1              2            1          deposits           to 1.45      $21,160            1          0                1
Trinidad &
Tobago            1986          1              1            1          deposits           0.2           $7,957            1          1                1
Venezuela         1985          1              1            1          deposits            2            $7,309            0          0                1

                                                                                                      $24,075 but
                                                                       insured          pro rata,          for
Austria           1979          1              3            0          deposits          expost       businesses          1          0                1
                                                                                                       ecu 1500,
                                                                        insured                      20000 in year
Belgium           1974          1              2            1          liabilities     0.02 + 0.04        2000            1          0                1
Canada            1967          1              1            1          deposits         0.33 max       $40,770            0          1                1

                                                                    Page 46 of 49
                                                                                                                                      July 8, 2004

     DEPOSIT                                                        Premium or         Annual                           Foreign   Interbank     Source of
               Enacted/   Membership   Administration   Funding                                       Coverage Limit
   INSURANCE                                                      Assessment Base     Premium                          Currency   Deposits      Funding
Denmark         1988          1              2            1          deposits        0.2 (max)         ecu 20000          1          0               1
                                                                     insured        risk based:
Finland         1969          1              3            1          deposits       0.05 to 0.3         $29,435           1          0               1
                                                                                    on demand
France          1980          1              3            0              n.a.       but limited            65387          1          0               0
                                                                       insured                         private: 30%
                                                                     deposits in                         of capital;
                                                                    commercial        official is          official
                                                                     banks DIS,       0.03 but         coinsurance
                                                                   risk-assets in      can be           90% to ecu
Germany         1966          1              3            1           other DIS       doubled              20000          1          0               0
                                                                                                      lesser of 90%
                                                                     insured        .expenses and      coinsurance
Gibraltar       1998          1              2            0          deposits           expost         of ecu 20000       1                          0
                                                                                      by size:
                                                                                     0.0025 to
Greece          1993          1              2            1           deposits         1.25           20,000 ECU          1          0               0
Iceland         1985          1              1            1           deposits          0.15           ecu 20000          1          0               0
                                                                   EU and EEA,                        coinsurance
                                                                    i.e insured                       90% to ecu
Ireland         1989          1              1            1           deposits           0.2             15000            1          0               0
                                                                                     the greater of
                                                                                      25,000 £ &
                                                                                      0.0125% of       the lower of
                                                                                     deposit base       $24,941 or
                                                                                    subject to max
                                                                                                         75% of
                                                                                      contribution       amount
Isle of Man     1991          1              1            0          deposits       of 250,000 £        deposited         1          0               0
                                                                  protected funds    adjusted
                                                                    adjusted for    expost 0.4
Italy           1987          1              2            0        size and risk      to 0.8            $125,000          1          0               1

                                                                  Page 47 of 49
                                                                                                                                                                                              July 8, 2004

                                                                                                        Premium or                 Annual                                   Foreign       Interbank     Source of
                        Enacted/         Membership            Administration          Funding                                                     Coverage Limit
                                                                                                      Assessment Base             Premium                                  Currency       Deposits      Funding
    DEPOSIT             Revised
                                                                                                                                                    $71000, but
                                                                                                           insured               0.0048 +             in full in
Japan                    1971                   1                      2                   1               deposits                0.036            recent crisis              0             0               1
Liechtenstein             003                   1                      1                   1                  n.a.                   n.a.              n.a.                   n.a.          n.a.             0
                                                                                                                                                    90% to ecu
                                                                                                                                                    15000 thru
                                                                                                           insured                                 1999, then to
Luxembourg               1989                   1                      3                   0               deposits               ex post           ecu 20000                  1             0               0
Malta                    2003
Netherlands              1979                   1                      1                   0           case by case               expost             ecu 20000                 1             0               1
                                                                                                       risk-weighted              0.005
                                                                                                      assets and total          assets 0.01
Norway                   1961                   1                      3                   1              deposits               deposits            $260,800                  1             0               1
                                                                                                                                                    ecu 15000,
                                                                                                                                0.08 to 0.12 +
                                                                                                           insured                  more in         coinsurance
Portugal                 1992                   1                      1                   1               deposits             emergencies        to ecu 45000                1             0               1
                                                                                                                                                     ecu 15000
                                                                                                           insured                                   1999, then
Spain                    1977                   1                      2                   1               deposits             max. of 0.2          ecu 20000                 1             0               1
                                                                                                          insured                 0.5 first,         ecu 28663
Sweden                   1996                   1                      1                   1              deposits                0.1 later           $31412                   1             0               1
                                                                                                       balance sheet
Switzerland              1984                   0                      3                   0               items                on demand             $19,700                  0             0               0
                                                                                                                                                   larger of 90%
United                                                                                                     insured                                 to $33,333 or
Kingdom                  1982                   1                      3                   0               deposits             on demand            ecu 22,222                1             0               0
                                                                                                           domestic                0.00 to
United States            1934                   1                      1                   1               deposits                 0.27              $100,000                 1             1               1
           Source: Demirgüç-Kunt and Sobaci (2001), as updated by Luc Laeven of the World Bank through February 6, 2004.
           Note: Blank spaces and n.a. indicates a detail of the deposit-insurance scheme that either has not yet been determined or has not yet been verified by World Bank personnel.
           See Demirgüç-Kunt and Sobaci (2001) for qualifications and additional explanation.

                                                                                                     Page 48 of 49
                                                                                     July 8, 2004

                        Table 4: 25 Highest-Ranking Stock Exchanges on
                            Market Integrity and Market Efficiency

  Market                                          Integrity        Efficiency
  Deustche Boerse-Xetra (electronic)                           4                 3
  New York Stock Exchange                                     10                 1
  Stockholmborson                                              5                 7
  Toronto Stock Exchange                                       7                 5
  Helsinki Stock Exchange                                      3                11
  Borsa Italia                                                11                 4
  Deustche Boerse-Frankfurt                                    1                14
  Copenhagen Stock Exchange                                    8                12
  Euronext Paris                                              19                 2
  London Stock Exchange                                       14                 8
  New Zealand Stock Exchange                                   6                16
  Tokyo Stock Exchange                                        17                 6
  Cyprus Stock Exchange                                        2                23
  Hong Kong Stock Exchange                                    16                10
  Australian Stock Exchange                                   15                13
  Singapore Stock Exchange                                    13                17
  Cairo & Alexandria Exchanges                                 9                22
  Taiwan Stock Exchange                                       22                 9
  NASDAQ Stock Market                                         20                15
  Philippines Stock Exchange                                  12                25
  Oslo Bors                                                   18                20
  Istanbul Stock Exchange                                     21                19
  Kuala Lumpur Stock Exchange                                 23                18
  American Stock Exchange                                     25                21
  Jakarta Stock Exchange                                      24                24

  Correlation                                                 0.226

Source: Siow and Aitken (2003), p.20.

                                        Page 49 of 49

To top