## _wp14118

## Source details

**Canonical URL:** [_wp14118](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14118.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14118.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14118.pdf.json)

---

### I. Introduction — key findings and context
- Database covers IMF membership of 188 countries plus Liechtenstein through end of 2013.
- Preliminary assessment:
  - DIS largely fulfilled primary objective of preventing open runs on bank deposits during the global crisis; widespread bank runs by insured depositors were not observed, with notable exceptions (e.g., Northern Rock in the UK) and protracted withdrawals by uninsured depositors.
  - Crisis response pattern: countries expanded coverage in existing DIS or adopted deposit insurance where absent.
- Policy concerns highlighted:
  - Expansion of national safety nets raises questions about government finances’ adequacy to support DIS promises in future stress.
  - Tension between preventing bank runs and DIS-induced moral hazard and incentives for aggressive risk-taking.
- Paper organization: Section 2 describes the database and variables; Section 3 surveys DIS worldwide; Section 4 reviews crisis-period policies; Section 5 concludes.

### II. The database — scope, sources, and construction
- Builds on earlier dataset (Demirgüç-Kunt, Karacaovali, and Laeven (2005)) that covered DIS through 2003; updated and extended to 2013.
- Primary sources used where possible: World Bank 2010 comprehensive survey on financial sector regulations, International Association of Deposit Insurers surveys (2008, 2010, 2011), European Commission (2011) for European countries, national sources (laws and regulations), and IMF staff reports.
- Crisis-era government actions sourced from Laeven and Valencia (2012), FSB (2010, 2012), Schich (2008, 2009), Schich and Kim (2011), and IMF staff reports.
- Focus is on deposit insurance for commercial banks; for countries with multiple DIS, data relate only to the national statutory scheme (may understate actual effective coverage).
- Full database, including arrangements beyond the national statutory scheme, is available in spreadsheet format as an online Appendix (per source).

### III. Variable definitions — key coding rules and enumerated variables
- Type of deposit insurance
  - Explicit defined by formal legislation/regulation; implicit assumed otherwise.
  - Explicit takes value 1 if country has explicit deposit insurance, 0 if implicit.
- Coverage
  - Coverage = statutory coverage limit in local currency (or “unlimited” if a full guarantee).
  - Coverage / GDP per Capita = ratio of coverage limit to per capita GDP, expressed as a percentage, based on statutory coverage limit.
  - Table 2 reports coverage limits in reported currency and US dollars (end-of-year exchange rates). GDP per capita from April 2014 IMF WEO database unless noted.
  - Coinsurance rules described where applicable.
- Organization and administration
  - DIS organization: legally separate entity, or contained within central bank, banking supervisor, or government ministry. Type coded 1 if legally separate, 2 if contained within an authority.
  - Administration: coded 1 if privately administered, 2 if publicly administered, 3 if jointly administered.
- Role
  - DIS roles: paybox only (coded 1) or paybox plus / loss or risk minimizer (coded 2).
  - “Paybox plus” can include resolution powers (create bridge banks, replace management) and mandate to minimize taxpayer losses.
- Multiple systems
  - Multiple = 1 if multiple statutory deposit insurance schemes exist within a country, 0 otherwise.
- Participation and types of deposits
  - Domestic banks, foreign subsidiaries, foreign branches, foreign currency deposits, interbank deposits all separately coded (1/0).
- Funding and backstops
  - Funding method: ex ante (coded 1) or ex post (coded 2).
  - Funding source: government (coded 1), privately by covered institutions (coded 2), or jointly (coded 3).
  - Backstop = 1 if legislation/regulation explicitly provides government support in case of DIS fund shortfall, 0 otherwise.
- Risk-adjusted premiums, assessment base, and payouts
  - Risk-adjusted premiums = 1 if premiums are adjusted for risk, 0 otherwise.
  - Assessment base categories: covered deposits, eligible deposits, total deposits, total liabilities.
  - Payout coding: 1 if coverage is per depositor account, 2 if per depositor per institution, 3 if per depositor.
  - Deposit losses: coded 1 for countries with substantial losses imposed on insured deposits despite explicit DIS — Argentina (1989 and 2001) and Iceland (2008); coded 0 otherwise.
- Banking crises
  - Indicator whether country experienced a banking crisis between 2007 and 2012; dates from Laeven and Valencia (2012), with Cyprus added as of 2012.

### IV. Introduction of deposit guarantee scheme — crisis-era policy actions and coding
- Introduction of explicit deposit insurance schemes
  - Several countries introduced explicit DIS during the financial crisis period (examples: Australia and Singapore).
  - Some countries transitioned from unlimited government guarantees already in place before the onset of the crisis into an explicit DIS with capped coverage limits (example: Thailand).
  - Coding: Introduction = 1 if introduced an explicit DIS during 2008-2013, 0 otherwise.
- Increase in statutory deposit coverage
  - Many countries raised coverage limits during the crisis.
  - Reasons: indexing to inflation/currency pegs/income measures; discouraging deposit outflows; EU convergence/harmonization.
  - Some countries expanded the range of accounts covered to include foreign currency or interbank deposits.
  - Coding: Increase in coverage = 1 if coverage limits increased during 2008-2013, 0 otherwise.
- Abolishment of coinsurance
  - Pre-crisis: coinsurance used to preserve stability benefits while keeping monitoring incentives.
  - Crisis outcome: coinsurance proved politically difficult to maintain.
  - Numerical change: Sixteen countries had coinsurance in 2003 – by 2010, only three remained.
  - Coding: Coinsurance = 1 if coinsurance was abolished during 2008-2013, 0 otherwise.
- Government guarantees on deposits, non-deposit liabilities, and bank assets
  - Several countries instituted temporary unlimited guarantees on deposits alongside increases in statutory coverage.
  - Classification distinguishes Limited vs Full coverage, year introduced (In place) and year expired (Expired); details in Table 5 footnotes.
  - Many countries extended guarantees on non-deposit liabilities during systemic crises; scope varied from a few major institutions to specific debt classes or unlimited guarantees.
  - Some governments guaranteed particular asset classes on banks’ balance sheets (examples: Netherlands—ING; Switzerland—UBS).
  - Coding: Guarantee on non-deposit liabilities = 1 if guarantees applied during 2008-2013, 0 otherwise. Guarantee on bank assets = 1 if applied during 2008-2013, 0 otherwise.
- Significant nationalizations of banks
  - Identified 17 countries where a significant portion of the financial system was nationalized since 2008 (including Belgium, Iceland, Ireland, the Netherlands, the United Kingdom, and the United States).
  - Nationalization implies implicit government backstop of deposits within nationalized institutions, reduction in counterparty risk elsewhere, and contingent future liability for governments.
  - Coding: Significant nationalization = 1 if significant nationalization occurred during 2008-2013.

### V. Main features of deposit insurance schemes around the world (end-2013) — aggregated facts and exact measures
- Coverage and adoption
  - Out of 189 countries covered, 112 countries (or 59 percent) had explicit deposit insurance by year-end 2013.
  - Increased from 84 countries (or 44 percent) in 2003.
  - 5 countries adopted deposit insurance in the year 2008 alone.
  - About 84 percent of high income countries had explicit deposit insurance by year-end 2013; Israel and San Marino exceptions.
  - Explicit deposit insurance in low income countries: about 32 percent.
  - Regional variation: Europe 96 percent; Africa 24 percent.
- Funding models and governance
  - 88 percent of countries with explicit deposit insurance have an ex ante (pre-funded) scheme.
  - Ex post schemes exist in about one-fourth of high income countries and are absent in low income and lower middle income countries.
  - About 43 percent of deposit insurance funds in ex ante schemes are a paybox (reimburse insured depositors only); 57 percent have extended powers (licensing, supervisory authority, information collection, minimizing losses/risks).
  - 14 percent of schemes are not legally separate from central bank, supervisory agency, or ministry of finance; this rises to 27 percent in low income countries.
  - Administration: about 66 percent of schemes are administered publicly; low income countries: 82 percent publicly administered; high income countries: 44 percent publicly administered, 21 percent privately administered, 35 percent jointly administered.
  - Funding source: 77 percent of all schemes funded privately, 2 percent funded exclusively by government, 21 percent funded jointly.
  - High income countries: 91 percent of schemes funded by the private sector.
- Coverage limits and relative generosity
  - Statutory coverage limit examples preserved from source:
    - Moldova: US$460
    - United States: US$250,000
    - Norway: US$327,172
    - Thailand: US$1,523,322 (blanket guarantee being phased out)
    - Turkmenistan: full guarantee
    - Uzbekistan: full guarantee
  - By end-2013, average coverage limits relative to per capita income:
    - High income countries: 5.3 times per capita income.
    - Upper middle income countries: 6.3 times per capita income.
    - Lower middle income countries: 11.3 times per capita income.
    - Low income countries: 5.0 times per capita income.
- Co-insurance and risk-adjusted premiums
  - Co-insurance usage fell from 16 schemes in 2003 to 3 by end-2013 (remaining: Bahrain, Chile, Libya).
  - By end-2013, 31 percent of schemes adjusted premium contributions for risk.
- Government backstops and contingent liabilities
  - About 38 percent of schemes have government backstops (credit lines or guarantees on debt issuances).
  - Presence of backstops is slightly higher in high income countries.

### VI. Depositor protection during the global financial crisis (2008 onward) — crisis-era statistics and dynamics
- Scope of safety-net expansions since 2008:
  - 14 countries introduced explicit deposit insurance since 2008.
  - For countries with deposit insurance that experienced a banking crisis:
    - 96 percent increased the statutory coverage limit.
    - 32 percent introduced government guarantees on deposits.
      - 38 percent of these deposit guarantees were blanket guarantees.
    - 72 percent extended guarantees on bank liabilities.
    - 36 percent used government guarantees on bank assets.
    - 64 percent undertook bank nationalizations.
- Effects and dynamics observed during the crisis:
  - Deposit insurance, combined with central bank liquidity support and fiscal backstops, contributed to preventing open bank runs (example: Federal Reserve liquidity + US Treasury backstop to FDIC).
  - Money market funds experienced massive outflows (notably after September 16, 2008) and required temporary federal insurance.
  - Notable exceptions to stability in Europe:
    - Northern Rock (UK) and DSB Bank (Netherlands) had isolated runs quickly contained.
    - Greece experienced a slow-moving deposit decline of about 20 percent between 2010 and 2012.
    - Cyprus experienced a generalized run when authorities suggested a tax on insured deposits (eventually not imposed).
  - Runs on uninsured deposits and non-deposit liabilities were widespread (e.g., wholesale deposit runs and repo runs on broker-dealers in the US), highlighting interconnectivity between banks and markets.
- Design shortcomings highlighted by crisis experience:
  - Many DIS did not foster market discipline nor adequately compensate for risks transferred to them.
  - Co-insurance largely phased out pre-crisis.
  - DIS premiums often failed to reflect tail risk; example: about 97 percent of banks in the US were assigned the lowest risk category and charged a zero percent explicit premium during the run-up to the crisis.
  - Ex ante funds were often small relative to insured deposits and below target sizes.
  - Many DIS appear underfunded, especially in countries with large financial systems.
  - The size of DIS funds seldom exceeds the percentage of deposits covered by DIS.
  - Large bank deposits relative to GDP combined with high government indebtedness raise doubts about governments’ ability to backstop the safety net.
- Distributional and cross-border complications:
  - Differential treatment of foreign and domestic depositors emerged (example: Iceland did not honor obligations to some foreign depositors).
  - Concerns about “bail-in” of large corporate depositors in Europe.

### VII. Safety Net Index (SNI) and political economy
- Construction and components
  - SNI constructed via principal component analysis of standardized design features increasing in moral hazard, summing the first six principal components with eigenvalues > 1.
  - Design features included coverage limit / GDP per capita and dummy variables for unlimited guarantees, foreign currency coverage, interbank coverage, no co-insurance, payout rules, no risk-adjusted premiums, ex ante fund, government funding, government backstops, no losses on uninsured deposits, government guarantees on deposits/non-deposit liabilities/bank assets since 2008.
- Selected SNI values observed in the dataset:
  - Argentina: -11.9 (imposed losses on insured depositors).
  - Iceland: -10.5 (imposed losses on insured depositors).
  - Ireland: 4.6 (issued temporary guarantees on deposits and non-deposit liabilities).
  - United States: 4.6 (issued temporary guarantees on deposits and non-deposit liabilities).
  - Turkmenistan: 4.5 (blanket guarantees).
  - Uzbekistan: 7.8 (blanket guarantees).
- Political economy constraints and time inconsistency
  - Governments rewrote statutes during crises to increase coverage, abolish coinsurance, and extend guarantees; these contingent liabilities often did not appear in official budgets.
  - It is politically difficult to unwind guarantees or decrease DIS coverage post-crisis.
  - Raising post-crisis premiums to recoup bailout subsidies is politically challenging; surviving banks often have increased political clout to resist actuarially fair premiums.
  - Bail-in policies suggested to reduce contingent fiscal liabilities face time-inconsistency problems—governments may renege during systemic crises due to contagion concerns.

### VIII. Empirical evidence on effects of deposit insurance
- Prior academic prescriptions: limited deposit insurance to reduce moral hazard (low coverage-to-income limits, co-insurance, exclude wholesale deposits).
- Anginer, Demirgüç-Kunt, and Min (2014):
  - Generous safety nets increased bank risk and systemic fragility in pre-crisis years (moral hazard effect).
  - During the crisis, bank risk was lower and systemic stability greater in countries with deposit insurance (stabilization effect).
  - Overall effect across full sample remained negative: destabilizing moral hazard effect dominated stabilizing effect.

### IX. Policy implications and recommendations
- Reforms needed to contain and mitigate contingent liability risks:
  - Monitor, assess, and report fiscal risks related to deposit insurance (DSI).
  - Consider reforms to reduce moral hazard while preserving short-run stabilization benefits.
  - Strengthen funding and risk-adjusted premium structures so DIS premiums better reflect tail risks.
  - Ensure ex ante funds target adequate sizes relative to insured deposits, and consider credible government backstops that are transparent and budgeted.
  - Evaluate gradual introduction of bail-in policies for uninsured creditors and debtholders, while acknowledging contagion and political-economy constraints and time-inconsistency problems.
  - Avoid widening the perimeter of insured instruments without careful assessment of systemic spillovers from protecting deposit-like instruments.

### X. Concluding remarks
- Deposit insurance helped prevent large-scale depositor runs during the global financial crisis but has generally underpriced risk, encouraging potentially ruinous bank risk-taking.
- Crisis-driven expansions of safety nets raised questions about governments’ ability to honor obligations and increased moral hazard.
- The growing reliance on short-term wholesale funding and market linkages raises doubts about the optimal perimeter of protection.
- A gradual move to bail-in policies could reduce fiscal contingent liabilities but faces political and systemic constraints.
- The paper provides a comprehensive database of DIS features and government guarantees to facilitate further research.

*Source: _wp14118 - References (content unit provided).*

### References .............................................................................................................

### _wp14118 - References

### Figures and Tables (Inventory)
- Figures listed: Figure 1 through Figure 16, covering topics such as Explicit Deposit Insurance by Income Group, Type of DI Scheme, Coverage Changes, Coinsurance Trends, Risk Adjustment of DI Premiums, Government Support of DI Schemes, Potential Deposit Liabilities and Ability to Pay, Size of DIS Fund Relative to Covered Deposits and Government Indebtedness, and Safety Net Index. (Figures numbered 1–16; specific figure titles present in source.)
- Tables listed: Table 1 through Table 5, covering Explicit Deposit Insurance Schemes Around the World, Coverage of Explicit Deposit Insurance Schemes, Design of Explicit Deposit Insurance Schemes, Recent Changes to Depositor Protection, and Fund Size and Coverage of Existing DIS. (Tables numbered 1–5; specific table titles present in source.)

### I. Introduction — key findings and context
- Database covers IMF membership of 188 countries plus Liechtenstein through end of 2013.
- Preliminary assessment: DIS largely fulfilled primary objective of preventing open runs on bank deposits during the global crisis; widespread bank runs by insured depositors were not observed, with notable exceptions (e.g., Northern Rock in the UK) and protracted withdrawals by uninsured depositors.
- Crisis response pattern: countries expanded coverage in existing DIS or adopted deposit insurance where absent.
- Policy concerns highlighted:
  - Expansion of national safety nets raises questions about government finances’ adequacy to support DIS promises in future stress.
  - Tension between preventing bank runs and DIS-induced moral hazard and incentives for aggressive risk-taking.
- Paper organization: Section 2 describes the database and variables; Section 3 surveys DIS worldwide; Section 4 reviews crisis-period policies; Section 5 concludes.

### II. The Database — scope, sources, and construction
- Builds on earlier dataset (Demirgüç-Kunt, Karacaovali, and Laeven (2005)) that covered DIS through 2003; updated and extended to 2013.
- Primary sources used where possible: World Bank 2010 comprehensive survey on financial sector regulations, International Association of Deposit Insurers surveys (2008, 2010, 2011), European Commission (2011) for European countries, national sources (laws and regulations), and IMF staff reports.
- Crisis-era government actions sourced from Laeven and Valencia (2012), FSB (2010, 2012), Schich (2008, 2009), Schich and Kim (2011), and IMF staff reports.
- Focus is on deposit insurance for commercial banks; for countries with multiple DIS, data relate only to the national statutory scheme (may understate actual effective coverage).
- Full database, including arrangements beyond the national statutory scheme, is available in spreadsheet format as an online Appendix (per source).

### A. Variable Definitions — enumerated variables and coding
- Type of deposit insurance
  - Explicit defined by formal legislation/regulation; implicit assumed otherwise.
  - Explicit takes value 1 if country has explicit deposit insurance, 0 if implicit.
- Coverage
  - Coverage = statutory coverage limit in local currency (or “unlimited” if a full guarantee).
  - Coverage / GDP per Capita = ratio of coverage limit to per capita GDP, expressed as a percentage, based on statutory coverage limit.
  - Table 2 reports coverage limits in reported currency and US dollars (end-of-year exchange rates). GDP per capita from April 2014 IMF WEO database unless noted.
  - Coinsurance rules described where applicable.
- Organization and administration
  - DIS organization: legally separate entity, or contained within central bank, banking supervisor, or government ministry. Type coded 1 if legally separate, 2 if contained within an authority.
  - Administration: coded 1 if privately administered, 2 if publicly administered, 3 if jointly administered.
  - Examples: some schemes legally separate but hosted within central bank; coded as legally separate.
- Role
  - DIS roles: paybox only (coded 1) or paybox plus / loss or risk minimizer (coded 2).
  - “Paybox plus” can include resolution powers (create bridge banks, replace management) and mandate to minimize taxpayer losses.
- Multiple systems
  - Multiple = 1 if multiple statutory deposit insurance schemes exist within a country, 0 otherwise.
  - Footnotes to Table 3 provide details on names/coverage of multiple DIS; focus of paper is on main statutory scheme applying to private commercial banks.
- Participation
  - Domestic banks = 1 if domestic banks are covered, 0 otherwise.
  - Foreign subsidiaries = 1 if local subsidiaries of foreign banks are covered, 0 otherwise.
  - Foreign branches = 1 if local branches of foreign banks are covered, 0 otherwise.
  - Example: EEA rules—coverage distinctions between foreign branches and subsidiaries; EEA deposits in non-EEA branches generally not covered.
- Types of deposits
  - Foreign currency deposits = 1 if DIS covers deposits in currencies other than the official domestic currency, 0 otherwise.
  - Interbank deposits = 1 if interbank deposits are covered, 0 otherwise.
  - Note: some DIS cover only a limited set of foreign currencies; payments on foreign currency deposits often made in local currency.
- Funding
  - Funding method: ex ante (coded 1) or ex post (coded 2).
  - Funding source: government (coded 1), privately by covered institutions (coded 2), or jointly (coded 3).
  - Example classification: Portugal classified as funded jointly where participating banks primarily fund DIS but government funding was provided.
- Government support (backstop)
  - Backstop = 1 if legislation/regulation explicitly provides government support in case of DIS fund shortfall, 0 otherwise.
  - Government support refers only to central government support, not central bank.
- Risk-adjusted premiums
  - Risk-adjusted premiums = 1 if premiums are adjusted for risk, 0 otherwise.
  - Examples: Italy groups banks into six risk categories mapped to six premium levels; Greece uses a risk coefficient between 0.9 and 1.1 from January 2009 based on solvency, liquidity, and internal control measures.
- Assessment base
  - Assessment base categories (mutually exclusive): covered deposits, eligible deposits, total deposits, total liabilities.
  - Definitions: eligible deposits = deposits repayable by DIS before coverage cap applied; covered deposits = eligible deposits after applying coverage limit.
  - Example: US FDIC changed assessment base (per Dodd-Frank) from total domestic deposits to average total assets minus tangible equity (Tier 1 capital).
- Payouts
  - Payout coding: 1 if coverage is per depositor account, 2 if per depositor per institution, 3 if per depositor.
  - Coverage per depositor account > per depositor per institution > per depositor in generosity order.
  - Special account rules (e.g., joint accounts in the US can double effective coverage) noted in footnotes to Table 3.
  - Deposit losses: coded 1 for countries with substantial losses imposed on insured deposits despite explicit DIS — Argentina (1989 and 2001) and Iceland (2008); coded 0 otherwise.
- Banking crises
  - Collected indicator whether country experienced a banking crisis between 2007 and 2012.
  - Banking crisis date denotes year of crisis. Dates for 2007–2011 from Laeven and Valencia (2012); Cyprus added as of 2012.

*Source: _wp14118 - References (content unit provided).*

### Introduction of deposit guarantee scheme

### Introduction of deposit guarantee scheme

### Introduction of explicit deposit insurance schemes
- Several countries introduced explicit deposit insurance schemes during the financial crisis period (examples given: Australia and Singapore).
- Some countries transitioned from unlimited government guarantees already in place before the onset of the crisis into an explicit DIS with capped coverage limits (example given: Thailand).
- Coding rule: Introduction is coded one if the country introduced an explicit deposit insurance scheme during the period 2008-2013, and zero otherwise.

### Increase in statutory deposit coverage
- Many countries raised coverage limits during the crisis.
- Reasons for increases:
  - Ex ante decisions to index coverage limits to inflation-adjusted units, currency pegs, or measures of income such as a multiple of the minimum wage.
  - To discourage deposit outflows from the banking system.
  - EU policies emphasizing convergence and harmonization in deposit insurance coverage led to large coverage increases in the EU.
- Some countries expanded the range of accounts covered to include foreign currency or interbank deposits.
- Coding rule: Increase in coverage is coded one if there was increase in coverage limits during the period 2008-2013, and zero otherwise.
- Reference: Table 4 specifies all countries that introduced temporary increases in coverage during the recent financial crisis.

### Abolishment of co-insurance
- Pre-crisis, co-insurance was used to preserve financial stability benefits of an explicit DIS while keeping monitoring incentives.
- Definition/example: With co-insurance, depositors are insured for only a pre-specified portion of their funds (i.e., less than 100 percent of their insured deposits). For example, in a country with 20 percent co-insurance and a maximum coverage limit of $100, depositors with less than $125 would receive 80 percent of the money within their account. Depositors with any amount greater than $125 (where 80 percent is the maximum of $100) would receive only $100.
- During the crisis, co-insurance proved politically difficult to maintain.
- Numerical change: Sixteen countries had co-insurance in 2003 – by 2010, only three remained.
- Coding rule: Coinsurance is coded one if co-insurance was abolished during the period 2008-2013, and zero otherwise.

### Government guarantee on deposits
- Several countries instituted a temporary unlimited guarantee on deposits alongside increases in statutory coverage.
- Coding rule: Government guarantee on deposits is coded one if a (partial or full) government guarantee on deposits was put in place during the period 2008-2013, and zero otherwise.
- Classification:
  - Distinguish whether the deposit guarantee covered only some deposits (Limited) or all deposits (Full).
  - Indicate the year when the guarantee was introduced (In place) and when the guarantee expired (Expired).
- Further details on specific deposit guarantees are provided in the footnotes to Table 5.

### Government guarantee on non-deposit liabilities
- Many countries extended guarantees on the non-deposit liabilities of financial institutions during systemic banking crises in the global financial crisis.
- Scope varied:
  - Guarantees limited to a small number of major institutions (example: Ireland).
  - Guarantees on specific debt classes or only new debt issuances (examples: Republic of Korea, United States).
  - Unlimited guarantees (example: Australia).
- Coding rule: Guarantee on non-deposit liabilities is coded one if there were government guarantees applied to non-deposit liabilities during the period 2008-2013, and zero otherwise.
- Further details on types of non-deposit guarantees are provided in the footnotes to Table 5.

### Government guarantee on bank assets
- Some governments guaranteed particular asset classes on banks’ balance sheets in addition to guaranteeing deposits and other liabilities.
- Examples: The governments of the Netherlands and Switzerland guaranteed the asset values of some hard-to-value assets on the balance sheets of ING and UBS, respectively.
- Coding rule: Guarantee on bank assets is coded one if there were government guarantees on banking assets during the period 2008-2013, and zero otherwise.

### Significant nationalizations of banks
- Government intervention also included taking control over financial institutions through nationalizations.
- Coding rule: Significant nationalization of banks is coded one if there was significant nationalization of banks and other financial institutions since the period covered (2008-2013).

*Source: Introduction of deposit guarantee scheme (excerpt).*

### 2008. Nationalizations are defined broadly to include explicit nationalizations as well as cases

### _wp14118 - 2008. Nationalizations are defined broadly to include explicit nationalizations as well as cases

### Nationalizations and implicit backstops
- Identified 17 countries where a significant portion of the financial system was nationalized since 2008 (including Belgium, Iceland, Ireland, the Netherlands, the United Kingdom, and the United States).
- Nationalization implies:
  - An implicit government backstop of all deposits within nationalized institutions.
  - A reduction in counterparty risk for the rest of the financial system.
  - A contingent future liability for national governments.

### Main features of deposit insurance schemes around the world
- Coverage and adoption
  - Out of 189 countries covered, 112 countries (or 59 percent) had explicit deposit insurance by year-end 2013.
  - Increased from 84 countries (or 44 percent) in 2003.
  - 5 countries adopted deposit insurance in the year 2008 alone.
  - Example: Australia adopted explicit deposit insurance in 2008.
  - EU harmonization spurred adoption in Central and Eastern Europe.
  - About 84 percent of high income countries had explicit deposit insurance by year-end 2013; Israel and San Marino were exceptions.
  - Explicit deposit insurance in low income countries: about 32 percent.
  - Regional variation: Europe 96 percent; Africa 24 percent.

- Funding models and governance
  - 88 percent of countries with explicit deposit insurance have an ex ante (pre-funded) scheme.
  - Ex post schemes exist in about one-fourth of high income countries and are absent in low income and lower middle income countries.
  - About 43 percent of deposit insurance funds in ex ante schemes are a paybox (reimburse insured depositors only); 57 percent have extended powers (licensing, supervisory authority, information collection, minimizing losses/risks).
  - 14 percent of schemes are not legally separate from central bank, supervisory agency, or ministry of finance; this rises to 27 percent in low income countries.
  - Administration: about 66 percent of schemes are administered publicly; low income countries: 82 percent publicly administered; high income countries: 44 percent publicly administered, 21 percent privately administered, 35 percent jointly administered.
  - Funding source: 77 percent of all schemes funded privately, 2 percent funded exclusively by government, 21 percent funded jointly.
  - High income countries: 91 percent of schemes funded by the private sector.

- Coverage limits and relative generosity
  - Statutory coverage limits examples: US$460 in Moldova; US$250,000 in the United States; US$327,172 in Norway; US$1,523,322 in Thailand (blanket guarantee being phased out); full guarantees in Turkmenistan and Uzbekistan.
  - By end-2013, average coverage limits relative to per capita income:
    - High income countries: 5.3 times per capita income.
    - Upper middle income countries: 6.3 times per capita income.
    - Lower middle income countries: 11.3 times per capita income.
    - Low income countries: 5.0 times per capita income.

- Co-insurance and risk-adjusted premiums
  - Co-insurance usage fell from 16 schemes in 2003 to 3 by end-2013 (remaining: Bahrain, Chile, Libya).
  - By end-2013, 31 percent of schemes adjusted premium contributions for risk.
  - Risk assessment methods varied widely.

- Government backstops and contingent liabilities
  - About 38 percent of schemes have government backstops (credit lines or guarantees on debt issuances).
  - Presence of backstops is slightly higher in high income countries.

### Depositor protection during the global financial crisis (2008 onward)
- Scope of safety-net expansions since 2008 (actions tracked in Figure 12):
  - 14 countries introduced explicit deposit insurance since 2008.
  - For countries with deposit insurance that experienced a banking crisis:
    - 96 percent increased the statutory coverage limit.
    - 32 percent introduced government guarantees on deposits.
      - 38 percent of these deposit guarantees were blanket guarantees.
    - 72 percent extended guarantees on bank liabilities.
    - 36 percent used government guarantees on bank assets.
    - 64 percent undertook bank nationalizations.

- Effects and dynamics observed during the crisis
  - Deposit insurance, combined with central bank liquidity support and fiscal backstops, contributed to preventing open bank runs (example: Federal Reserve liquidity + US Treasury backstop to FDIC).
  - Money market funds experienced massive outflows (notably after September 16, 2008) and required temporary federal insurance.
  - Notable exceptions to stability in Europe:
    - Northern Rock (UK) and DSB Bank (Netherlands) had isolated runs quickly contained.
    - Greece experienced a slow-moving deposit decline of about 20 percent between 2010 and 2012.
    - Cyprus experienced a generalized run when authorities suggested a tax on insured deposits (eventually not imposed).

  - Runs on uninsured deposits and non-deposit liabilities were widespread (e.g., wholesale deposit runs and repo runs on broker-dealers in the US), highlighting interconnectivity between banks and markets.

- Design shortcomings highlighted by crisis experience
  - Many DIS did not foster market discipline nor adequately compensate for risks transferred to them.
  - Co-insurance (market discipline tool) largely phased out pre-crisis.
  - DIS premiums often failed to reflect tail risk; example: about 97 percent of banks in the US were assigned the lowest risk category and charged a zero percent explicit premium during the run-up to the crisis.
  - Ex ante funds were often small relative to insured deposits and below target sizes.
  - Many DIS appear underfunded, especially in countries with large financial systems.
  - The size of DIS funds seldom exceeds the percentage of deposits covered by DIS.
  - Large bank deposits relative to GDP combined with high government indebtedness raise doubts about governments’ ability to backstop the safety net.

- Distributional and cross-border complications
  - Differential treatment of foreign and domestic depositors emerged (example: Iceland did not honor obligations to some foreign depositors).
  - Concerns about “bail-in” of large corporate depositors in Europe.

- Safety Net Index (SNI) and moral hazard
  - SNI constructed via principal component analysis of standardized design features increasing in moral hazard, summing the first six principal components with eigenvalues > 1.
  - Design features included coverage limit / GDP per capita and dummy variables for unlimited guarantees, foreign currency coverage, interbank coverage, no co-insurance, payout rules, no risk-adjusted premiums, ex ante fund, government funding, government backstops, no losses on uninsured deposits, government guarantees on deposits/non-deposit liabilities/bank assets since 2008.
  - SNI values observed:
    - Argentina: -11.9 (imposed losses on insured depositors).
    - Iceland: -10.5 (imposed losses on insured depositors).
    - Ireland: 4.6 (issued temporary guarantees on deposits and non-deposit liabilities).
    - United States: 4.6 (issued temporary guarantees on deposits and non-deposit liabilities).
    - Turkmenistan: 4.5 (blanket guarantees).
    - Uzbekistan: 7.8 (blanket guarantees).

- Political economy constraints and time inconsistency
  - Governments rewrote statutes during crises to increase coverage, abolish coinsurance, and extend guarantees; these contingent liabilities often did not appear in official budgets.
  - It is politically difficult to unwind guarantees or decrease DIS coverage post-crisis.
  - Raising post-crisis premiums to recoup bailout subsidies is politically challenging; surviving banks often have increased political clout to resist actuarially fair premiums.
  - Bail-in policies are suggested as a way to reduce contingent fiscal liabilities, but they suffer from time inconsistency — governments may renege during systemic crises due to contagion concerns.

- Empirical assessments of deposit insurance effects
  - Prior academic work advocated limited deposit insurance to reduce moral hazard (low coverage-to-income limits, co-insurance, exclude wholesale deposits).
  - Anginer, Demirgüç-Kunt, and Min (2014) findings:
    - Generous safety nets increased bank risk and systemic fragility in pre-crisis years (moral hazard effect).
    - During the crisis, bank risk was lower and systemic stability greater in countries with deposit insurance (stabilization effect).
    - Overall effect across full sample remained negative: destabilizing moral hazard effect dominated stabilizing effect.

### Policy implications and recommendations
- Reforms needed to contain and mitigate contingent liability risks:
  - Monitor, assess, and report fiscal risks related to deposit insurance (DSI).
  - Consider reforms to reduce moral hazard while preserving short-run stabilization benefits.
  - Strengthen funding and risk-adjusted premium structures so DIS premiums better reflect tail risks.
  - Ensure ex ante funds target adequate sizes relative to insured deposits, and consider credible government backstops that are transparent and budgeted.
  - Evaluate gradual introduction of bail-in policies for uninsured creditors and debtholders, while acknowledging contagion and political-economy constraints and time-inconsistency problems.
  - Avoid widening the perimeter of insured instruments without careful assessment of systemic spillovers from protecting deposit-like instruments.

### Concluding remarks
- Deposit insurance helped prevent large-scale depositor runs during the global financial crisis but has generally underpriced risk, encouraging potentially ruinous bank risk-taking.
- Crisis-driven expansions of safety nets raised questions about governments’ ability to honor obligations and increased moral hazard.
- The growing reliance on short-term wholesale funding and market linkages raises doubts about the optimal perimeter of protection.
- A gradual move to bail-in policies could reduce fiscal contingent liabilities but faces political and systemic constraints.
- The paper provides a comprehensive database of DIS features and government guarantees to facilitate further research.

*Source: _wp14118 - 2008. Nationalizations are defined broadly to include explicit nationalizations as well as cases*

### REFERENCES

### _wp14118 - REFERENCES

### Key bibliographic sources cited
- Anginer, D., A. Demirgüç-Kunt, and M. Zhu, 2014, “How Does Deposit Insurance Affect Bank Risk? Evidence from the Recent Crisis,” Journal of Banking and Finance; also World Bank Policy Research Paper, WPS6289.
- Basel Committee on Banking Supervision (BCBS) and International Association of Deposit Insurers (IADI), 2009, “Core Principles for Effective Deposit Insurance Systems,” https://www.bis.org/publ/bcbs156.pdf
- Blinder, A.S., and R.F. Wescott, 2001, “Reform of Deposit Insurance: A Report to the FDIC,” FDIC and Princeton University. Mimeo. http://www.fdic.gov/deposit/insurance/initiative/reform.html
- Demirgüç-Kunt, A., and E. Detragiache, 2002, “Does Deposit Insurance Increase Banking System Stability? An Empirical Investigation,” Journal of Monetary Economics, Vol. 49, No. 7, pp. 1373-1406.
- Demirgüç-Kunt, A., and H. Huizinga, 2004, “Market Discipline and Deposit Insurance,” Journal of Monetary Economics, Vol. 51, No. 2, pp.375-399.
- Demirgüç-Kunt, A., B. Karacaovali, and L. Laeven, 2005, “Deposit Insurance around the World: A Comprehensive Database,” Policy Research Working Paper No. 3628 (Washington, DC: World Bank).
- Demirgüç-Kunt, A., E.J. Kane, and L. Laeven, 2008a, “Determinants of Deposit-Insurance Adoption and Design,” Journal of Financial Intermediation, Vol. 17, No. 3, pp. 407-438.
- Demirgüç-Kunt, A., E.J. Kane, and L. Laeven (Eds.), 2008b, Deposit Insurance around the World: Issues of Design and Implementation (Cambridge, MA: MIT Press).
- European Commission, 2004, Report on minimum guarantee level of Deposit Guarantee Schemes Directive 94/19/EC, http://ec.europa.eu/internal_market/bank/docs/guarantee/report_en.pdf. Annex available from: http://ec.europa.eu/internal_market/bank/docs/guarantee/annexes_en.pdf
- European Commission, 2010, Impact Assessment for Proposal on Deposit Guarantee Schemes, http://ec.europa.eu/internal_market/bank/docs/guarantee/20100712_ia_en.pdf. Annex II available from: http://ec.europa.eu/internal_market/bank/docs/guarantee/jrc-annex2_en.pdf
- European Commission, 2011, JRC Report under Article 12 of Directive 94/19/EEC, http://ec.europa.eu/internal_market/bank/docs/guarantee/jrc-rep_en.pdf. Annex I and II available from: http://ec.europa.eu/internal_market/bank/docs/guarantee/jrc-annex1_en.pdf and http://ec.europa.eu/internal_market/bank/docs/guarantee/jrc-annex2_en.pdf
- European Federation of Deposit Insurance, 2006, Deposit Guarantee Systems: EFDI’s First Report, http://efdi.eu/fileadmin/user/publications/EFDI%20publications/Deposit%20Guarantee%20Systems%20EFDIs%20First%20Report%20%28Manuela%20De%20Cesare%202006%29.pdf
- Financial Stability Board, 2012, Thematic Review on Deposit Insurance Systems, http://www.financialstabilityboard.org/publications/r_120208.pdf
- Financial Stability Board, 2010, Update on Unwinding Temporary Deposit Insurance Arrangements, http://www.financialstabilityboard.org/publications/r_1006.pdf
- Garcia, G., 2000, Deposit Insurance: Actual and Good Practices, IMF Occasional Paper No. 197 (Washington, DC: International Monetary Fund).
- International Association of Deposit Insurance, 2008, 2010, 2011 Annual Survey Results of Deposit Insurance, http://www.iadi.org/Research.aspx?id=58
- Kane, E.J., 1989, The S&L Insurance Mess: How Did it Happen? (Washington, DC: The Urban Institute Press).
- Laeven, L., 2013, European Union: Publication of Financial Sector Assessment Program Documentation—Technical Note on Deposit Insurance, IMF Country Report No. 13/66. http://www.imf.org/external/pubs/ft/scr/2013/cr1366.pdf
- Laeven, L., and F. Valencia, 2012, “Systemic Banking Crises Database: An Update,” IMF Working Paper no. 12/163. http://www.imf.org/external/pubs/ft/wp/2012/wp12163.pdf
- Schich, S., 2008, “Financial Crisis: Deposit Insurance and Related Safety Net Aspects,” Financial Market Trends, OECD. http://www.oecd.org/finance/financial-markets/41894959.pdf
- Schich, S., 2009, “Expanded Guarantees for Banks: Benefits, Costs, and Exit Issues,” Financial Market Trends, OECD. http://www.oecd.org/finance/financial-markets/42779438.pdf
- Schich, S., and B.-H. Kim, 2011, Guarantee Arrangements for Financial Promises: How Widely Should the Safety Net Be Cast? Financial Market Trends, OECD. http://www.oecd.org/finance/financial-markets/48297609.pdf
- World Bank, 2003 and 2011, Survey of Banking Supervision and Regulation (data files referenced).

### Figures listed (titles and focus)
- Figure 1. Explicit Deposit Insurance by Income Group, 2013
- Figure 2. Explicit Deposit Insurance by Region, 2013
- Figure 3. Type of DI Scheme, 2013
- Figure 4. Objective of the DI Scheme, 2013
- Figure 5. Organization of the DI Scheme, 2013
- Figure 6. Administration of the DI Scheme, 2013
- Figure 7. Funding of the DI Scheme, 2013
- Figure 8. Coverage Increased During Crisis and Remains Above Pre-Crisis Levels
  - Data series labels shown: Coverage limit / GDP per capita, 2003; Coverage limit / GDP per capita, 2010; Coverage limit / GDP per capita, 2013
- Figure 9. Decline of Coinsurance, 2003−2013
- Figure 10. Risk Adjustment of DI Premiums, 2013
- Figure 11. Government Support of DI Schemes, 2013
- Figure 12. Increase in Depositor Protection, 2007−2013 (% of countries with explicit DIS)
- Figure 13. Potential Deposit Liabilities and Ability to Pay by the DIS Fund, end-2010 (note: Middle income includes lower and upper middle income countries. Insufficient data to report figures on low income countries.)
- Figure 14. Total Deposits and Ability to Pay by the Government, end-2010 (note: Middle income includes lower and upper middle income countries. Insufficient data to report figures on low income countries.)
- Figure 15. Size of DIS Fund Relative to Covered Deposits and Government Indebtedness
- Figure 16. Safety Net Index, 2013
  - Notes: "The safety net index is a principal components index of DI design and other safety net features that is increasing in the generosity of the safety net." "Countries with safety net index (SNI) values between -1 and +1 are excluded from the chart."

### Tables and selected exact figures or labels preserved from the source
- Table 1. Explicit Deposit Insurance Schemes Around the World, end-2013
  - Includes country lists and annotations such as: "As of 2013", and country-specific notes numbered 1/ through 10/.
- Table 2. Coverage of Explicit Deposit Insurance Schemes Around the World, end-2013
  - Column headings preserved exactly: Country | 2003 | 2010 | 2013 | 2003 | 2010 | 2013 | 2003 | 2010 | 2013 | 2003 | 2010 | 2013
  - Example entries preserved exactly where present in the source (sample rows):
    - Afghanistan: n.a., AF100,000, AF100,000; n.a., 222, 217, 67; n.a., 222, 217, 67; n.a., 412, 260
    - Albania: 100% of first LEK350,000; 85% of next LEK411,765 (up to maximum of LEK700,000); LEK2,500,000; LEK2,500,000; 579; 624; 0322449857962403224498319586531
    - Australia: n.a.; AUD1,000,000; AUD250,000; n.a.; 917431221625; n.a.; UNLIMITED 1/221625; n.a.; 1628342
    - United States: USD100,000; USD250,000; USD250,000; 100000250000250000; 100000250000 13/250000; 262518471
  - Notes section includes numbered clarifications such as:
    - 1/ On October 12, 2008, Australia announced an unlimited guarantee scheme...
    - 2/ Blanket guarantee introduced in 2008 expired at the end of 2010.
    - 3/ Indonesia introduced explicit deposit insurance in 2004. The 2003 coverage limit refers to blanket guarantee in place.
    - ...and additional numbered notes up to at least 21/.
- Table 3. Design of Explicit Deposit Insurance Schemes Around the World, end-2013
  - Detailed scheme features and design elements captured as column labels and markers in the table, including:
    - Type of Deposit Insurance Scheme: explicit, legally separate, central bank/supervisor/ministry
    - Administration: administered publicly, administered privately, administered jointly
    - Scheme function: paybox only, paybox plus, loss or risk minimizer, multiple schemes
    - Participation and Coverage: compulsory for domestic banks, local subsidiaries of foreign banks, local branches of foreign banks, foreign currency deposits, interbank deposits, coinsurance
    - Funding: ex-ante fund, ex-post scheme, funded by government, funded privately, funded jointly, backstop
    - Contributions and Assessment Base: risk-adjusted premiums, assessment base 34/, covered deposits, eligible deposits, total deposits, total liabilities
    - Payouts to Depositors: per deposit account 32/, per deposit or per institution, deposit losses imposed 21/xx
  - Extensive numbered notes (1/ through at least 35/) clarifying national practices and special cases.
- Table 4. Recent Changes to Depositor Protection, 2007−2013
  - Tracks for each country: Experienced banking crisis between 2007 - 2013 1/; Increase in deposit protection since 2008 (introduction of DIS, increase in statutory DIS coverage, government guarantee on deposits, abolished coinsurance, limited/unlimited guarantees); Government guarantees on bank deposits since 2008; Significant nationalizations of banks since 2008; Government guarantees on non-deposit liabilities since 2008; Government guarantees on bank assets since 2008.
  - Notes include clarifications such as: "1/ Banking crisis dates for the period 2007-2011 according to Laeven and Valencia (2012). Cyprus is added to this list as of 2012."
- Table 5. Fund Size and Coverage of Existing DIS, 2010
  - Column headings preserved: Country | Total deposits (US$bn) | Eligible deposits (US$bn) | Covered deposits (US$bn) | Size of the DIS fund (US$bn) | Public debt (as % of GDP) 1/ | GDP (US$bn) | Total deposits / GDP | Size of DIS fund / Covered deposits
  - Selected exact numeric entries from the table preserved verbatim:
    - Argentina: Total deposits 95; Eligible deposits 27.6; Covered deposits 21.4; Size of the DIS fund 9.2367.6; Public debt 25.8; GDP 4.4
    - Australia: Total deposits 1336; Eligible deposits 1269.2815020.5; Covered deposits 1247.2; Size of DIS fund 107.1; GDP 0
    - United States: Total deposits 788878886231.5-7.4; Public debt 894.8; GDP 14958.352.7-0.12
    - Japan: Total deposits 111019990.97881.73.15; Public debt 216.05495.4202.00.04
  - Notes: "1/ General government gross debt (as % of GDP). Source: IMF World Economic Outlook April 2014." "2/ Size of DIS fund / Eligible deposits."
  - Sources line for table: European Commission, International Association for Deposit Insurers, Financial Stability Board (2010, 2012), FDIC, Laeven and Valencia (2012), IFS, IMF staff reports, and national deposit insurance agencies.

*Italic: Sources listed in the reference section of _wp14118 (as provided).*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2014/_wp14118.pdf_