## _wp05188

## Source details

**Canonical URL:** [_wp05188](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05188.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05188.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05188.pdf.json)

---

### I. Introduction
- Dollarized economies have experienced severe banking and currency crises driven largely by runs on bank deposits, mainly dollar deposits.
- Examples and dynamics:
  - The Argentine crisis (2002) and spillovers to Uruguay and Paraguay illustrate that dollarization can complicate crisis management and create major financial vulnerabilities.
  - Systemic runs in dollarized economies have an important self-fulfilling component (Diamond and Dybvig (1983); Calvo (1988); Cole and Kehoe (1998)).
  - Dollar deposits can only be paid to the extent banks have sufficient dollar liquidity or access (through lender of last resort, or LOLR) to dollar liquidity on-lent by the central bank.
  - In highly dollarized fractional reserve systems, only a limited proportion of deposits can be backed by central bank dollar liquidity.
  - Runs are exacerbated by firesales of bank assets and by the threat of last-resort measures such as deposit freezes.
  - In mixed-currency economies, additional incentives to run include threats of forced conversions into pesos and large exchange rate depreciations (as occurred during the Argentine crisis of 2002).
  - Floating the exchange rate to stem peso runs can further harm dollar assets and worsen banks’ positions.
- Paper objectives:
  - Review recent experiences with liquidity insurance.
  - Argue in favor of liquid asset requirements (LARs).
  - Present the concept of circuit breakers (CBR) — temporary, efficient, pre-programmed suspension of convertibility — combined with limited earmarked dollar liquidity and automatic reprogramming of term dollar deposits.
  - Conclude that CBR, if well designed and accompanied by prudential policies and a bank resolution framework, could limit the need and cost of dollar liquidity buffers and reduce destabilizing runs.

### II. Can Liquidity Be Borrowed?
- Two broad approaches to insure a dollarized financial system against dollar liquidity shortages:
  - Self-insurance: holding a substantial stock of foreign-currency-denominated liquid assets, either by the central bank or by individual banks (LARs).
  - External insurance: contracts with private providers of dollar liquidity (consortium of financial institutions) or with international financial institutions (IFIs).
- Cost and practical considerations of self-insurance:
  - Self-insurance entails a non-trivial cost equal to the excess financing cost over return on liquid foreign assets, combining a maturity and sovereign risk premia.
  - Back-of-the-envelope estimate approach: cost ≈ difference between average yield of external debt (e.g., J.P. Morgan’s EMBI index) and return on central bank reserves — a non-negligible number for most emerging economies.
  - If runs were purely self-fulfilling, external insurance would be less expensive than holding liquidity, and a perfectly insured country would be immune to non-fundamental runs.
  - Practical complications: required insurance packages are typically large; risks difficult to diversify; insurance structured as a loan with solvency preconditions; defining insurable event and verifiable losses is difficult (moral hazard).
  - Insurance premia are likely prohibitively expensive unless the facility is collateralized; collateral availability in emerging economies is generally limited and of dubious value.

### II.B Private Insurance — Recent Experiences and Lessons
- Argentine contingent credit line (late 1990s) — operational facts and constraints:
  - Contract allowed central bank and participating local banks to engage in repurchase agreements against Argentine sovereign securities for up to US$6.7 billion.
  - Execution was delayed until August 2001, when the run was well underway.
  - Ultimately provided US$1.77 billion (out of US$4.75 billion available at the beginning of 2001).
  - Bond price declines used as collateral implied reductions in the size of the line; line dropped to US$1.35 billion at the first three-month renewal.
  - Contract features:
    - Maturity: three years with an ever-greening clause extending life every three months.
    - Argentine dollar-denominated bonds taken at 80 percent of market value (if bond price fell by more than 5 percent, more bonds had to be delivered as margin).
    - Annual commitment fee: 32 basis points.
    - Interest on withdrawn funds: roughly LIBOR plus 205 basis points.
    - In 1998, average government bond yield was around 940 bps.
    - Illustrative cost comparison: increasing reserves by the committed amount would entail a “fee” cost of about 570 bps (alternatively, a flow cost of roughly US$380 million per year), well above the 32 bps commitment fee under the contract.
  - Execution constraints: decline in collateral value and legal/budgetary restrictions prevented timely issuance of additional collateral; issuing bonds in private markets was not an option during the run.
- Mexican contingent credit line (Box 2 summary):
  - The line was withdrawn in its entirety by the Mexican government on September 30, 1998 due to deterioration in international capital market access and a decline in oil prices that reduced fiscal resources.
  - Private insurers disputed the government’s decision but ultimately extended the loan; subsequent refusal to extend financing illustrates insurer actions to limit downside exposure under rising risk.
- General lessons:
  - Insurers facing poor collateral and rising risk will use contractual rights to limit downside exposure, reducing the effectiveness of the facility when needed most.
  - Insurance premia and contractual terms (margins, valuation rules, frequent renewals) can undermine the intended stabilizing role of private liquidity lines.
  - Market-driven collateral deterioration and contract design can lead to a paradoxical reduction in available liquidity precisely during runs.

### III. Centralized versus Decentralized Liquidity Buffers
- Trade-off overview:
  - Centralizing reserves at the central bank introduces agency costs leading to suboptimal dollar liquidity holdings and implicit subsidies to dollar intermediation.
  - Decentralized holdings via LARs internalize currency-risk externalities but forgo diversification benefits of a common liquidity fund.
  - For largely systemic shocks, benefits of decentralized LARs outweigh costs; a positive LAR (and one greater than the one corresponding to peso deposits) is optimal from a prudential perspective.
  - Once LARs are introduced, a limited dollar liquidity recycling facility can be justified to address idiosyncratic liquidity risk and enhance banking system resilience.
- Model insights on LOLR vs LAR:
  - Presence of a LOLR reduces banks’ demand for precautionary liquidity because borrowing cost from LOLR replaces the marginal cost of falling short.
  - Fundamental asymmetry: peso LOLR can be provided at virtually no cost via issuance of peso liabilities; dollar LOLR can only be provided to the extent of dollar reserves and is costly.
  - Distortions from dollar LOLR:
    - Acts as a subsidy to commercial banks (at taxpayer expense), lowering cost of dollar deposits and promoting excessive dollar intermediation.
    - Reliable dollar LOLR may subsidize risk-taking: risky banks rely more on LOLR, hold fewer reserves, and compete unfairly with conservative banks.
  - Policy implication: require banks to hold minimum liquid dollar reserves in proportion to their liquid dollar liabilities (introduce LAR).
- Design features and post-LAR role of central bank:
  - LARs should induce banks to hold socially optimal liquidity levels; eligible assets should be foreign assets with deep, liquid markets.
  - Central bank dollar LOLR can be credibly limited to exceptional support for isolated banks facing very large liquidity shocks.
  - Official international reserves are needed to the extent central bank obligations to liquidity-supplying banks may need partial backing by foreign reserves, particularly for very short maturities.

### IV. Circuit Breakers (CBR) — Concept and Design
- Concept:
  - CBR: a temporary, efficient, pre-programmed suspension of convertibility designed to economize on reserves and stop runs with less trauma than severe last-resort measures.
- Core elements of the proposed scheme:
  - Limited dollar liquidity earmarked to ensure full convertibility of transactional dollar deposits.
  - A mechanism that automatically reprograms dollar-term deposits once triggered by a sufficient decline in banks’ liquidity.
- Potential benefits:
  - Reduce the need for a large dollar liquidity buffer and its cost.
  - Narrow scope for destabilizing runs on the banking system.
- Preconditions and complementary policies:
  - Well-designed legal and operational framework (could be de facto implemented without explicit legal reference to deposit restructuring).
  - Adequate prudential policies, including a bank resolution framework.
- Activation design options and trade-offs:
  - Solution 1: Purely systemic trigger (applies to all banks when systemic liquidity hits a threshold) — halts runs early but may penalize conservative banks and invite litigation absent verifiable indicators.
  - Solution 2: Purely idiosyncratic trigger (automatic when any individual bank runs out of liquidity) — removes disciplining benefit on bank management and may be politically difficult.
  - Solution 3: Middle-of-the-road: requires both a bank’s liquidity to fall below a threshold and supervisor consent that cause is systemic — limits moral hazard but introduces discretion and possible delay.
  - Solution 4: Milder, CBR-like option (Chilean-type deposit insurance preserving transaction deposit liquidity while allowing the rest of the bank to be “temporarily closed”) — prompt consent more likely, postpones liquidation.
  - Policy preference: non-discretionary, automatic triggers determined by transparent, legally uncontestable criteria (natural mechanism: link trigger to failure to comply with a minimum level of liquidity so LARs backing non-CBR deposits are not compromised).
- Post-activation considerations (“the day after”):
  - Effectiveness depends on whether introduction creates expectations of forthcoming losses on restructured deposits.
  - Rules should be well specified and transparent: time frame of restructuring, procedures if liquidity problems persist, efficient bank resolution procedure to minimize loss in deposit value.
  - Dollarized-economy-specific options:
    - Give peso depositors an official guarantee backed by preferential claims on residual bank value.
    - Limit bank’s liability on dollar losses to its dollar capital.
    - Full “de-coupling”: separate peso and dollar components of bank balance sheet upon CBR activation (legal complexity noted).

### Box 2 — The Mexican Contingent Credit Line (highlights)
- Facility overview:
  - In November 1997, Mexico subscribed to a contingent credit line with 31 private international financial institutions.
  - Facility size: US$2.5 billion (later extended to US$2.66 billion).
  - Availability: up to 18 months.
  - Pricing: 3M LIBOR plus 50 bps during the first semester, increasing by 25 bps each subsequent semester.
  - Cost: annual commitment fee of 30 bps (approximately US$7.6 million).
- Practical problems and shortcomings:
  - Difficulty defining the trigger event and insurers’ reluctance to assume costs produced controversial interpretations, reducing timeliness and effectiveness.
  - Moral hazard and hedging incentives: insurer banks may hedge exposure (e.g., shorting country assets), accelerating collapse.
  - Correlated risks and size limits: insurers’ scope to diversify highly correlated regional risks narrows.
  - Reverse moral hazard: insurers may take on excessive risk for commissions and then refuse to provide resources once a crisis begins.
- Public insurance and IMF experience:
  - Contracting insurance from IFIs or regional arrangements can overcome some private insurance drawbacks, but IMF experience shows difficulties isolating liquidity from solvency concerns.
  - The IMF’s contingent credit line (CCL) required discretionary pre-qualification, had limited size, was never requested, and was ultimately phased out.
- Implication:
  - Private liquidity insurance is likely limited and possibly unreliable in substantive shocks; realistic option: self-insurance via LARs or central bank reserves.

### Box 3 — Introducing CBR-Type Measures Under Crises (highlights)
- Deposit freezes versus compulsory reprogrammings:
  - Deposit freezes: depositors lose access, uncertainty about restoration and interest earned; typically temporary until longer-term solution found.
  - Compulsory reprogrammings: depositors receive a new instrument with known maturity, tradable in secondary and repo markets; a freeze should evolve into full reprogramming if conditions persist.
- The Argentine “corralito”: features and drawbacks
  - Allowed intra-system transfers and check usage but prohibited cash withdrawals and restricted transfers abroad; deepened currency mismatch and flight to quality; cash hoarding disrupted payments system.
  - Empirical indicator: liquidity premium on informal market check discounts reached up to 12% on sample dates in 2002 (dates listed).
- Historical precedent and acceptability:
  - Option clauses/suspension of convertibility existed historically and were sometimes effective (European option clause; Bank of Scotland 1730–1765).
  - U.S. National Banking Era suspensions documented; worst depositor loss estimated at 2.1 cents per dollar.
  - Collective action clause analogy: initial reluctance did not lead to market penalties when included.
- Which deposits and currencies to subject to CBR:
  - Two key questions: deposit type/maturity and currency treatment.
  - Approaches: differentiate pesos from dollars; currency-blind; intermediate (differentiate but protect peso convertibility).
  - Natural operational choice: apply CBRs only to time deposits while ensuring full convertibility of transactional deposits (into cash or dollars).

### Box 5 — Chilean “Narrow Bank” Safety Net (pre-programmed CBR example)
- Mechanics:
  - Sight deposits (and term deposits < 30 days or term to maturity < 10 days) fully guaranteed by BCCh.
  - BCCh requires banks hold liquid assets (central bank debt) against sight deposits in excess of 2.5 times their capital.
  - If guarantee triggered, BCCh becomes most senior claimant on bank assets; bank’s non-sight liabilities frozen while sight deposits remain accessible; sight deposits are decoupled with corresponding liquid assets plus BCCh guarantee.
  - Five-day advance notice required by banks for transferring term deposits into sight deposits limits unwarranted last-minute guarantee expansion.
  - Bank resolution conditioned on creditor agreement ratified by majority of non-sight creditors and the Superintendency; when agreement cannot be reached, outright liquidation of non-narrow bank is the only outcome.
- CBR options and currency-regime implications:
  - Option 1: Let exchange rate float (differentiate pesos and dollars) — limits need for foreign reserves but large depreciation can undermine banking solvency in highly dollarized economies.
  - Option 2: Currency board — requires very large reserves; without them not credible and may unravel.
  - Option 3: Currency-specific regulatory framework — viable for highly dollarized countries not pursuing rapid de-dollarization; ensure dollar deposits are not converted into peso deposits; apply CBRs only to time deposits while ensuring full convertibility of transactional deposits.
  - Policy design: differentiated LARs on dollar deposits; reduce LARs for long-maturity instruments if issued as non-redeemable tradable CDs; joint determination of LARs and CBRs.
- Activation and sequencing:
  - Activation design must balance early halting of runs and avoidance of penalizing well-managed banks; preferred mechanisms link triggers to verifiable liquidity breach relative to assigned LARs so that banks deplete backing before automatic reprogramming/CBR kicks in.

### Key Statistics and Contractual Figures (preserved exactly)
- US$6.7 billion
- US$1.77 billion
- US$4.75 billion
- US$1.35 billion
- 80 percent
- 5 percent
- 32 basis points
- LIBOR plus 205 basis points
- 940 bps
- 570 bps
- Roughly US$380 million per year
- September 30, 1998

### Conclusions and Policy Implications (summary)
- Main conclusions:
  1. The current menu of liquidity insurance options is mostly limited to forms of self-insurance (e.g., holding your own liquidity). Market-based (or official) insurance faces obstacles in achieving the right combination of size, cost, and guaranteed access.
  2. Accumulating reserves centrally induces moral hazard as banks, particularly weaker ones, free ride on the central bank and hold suboptimally low levels of dollar liquidity. Centralized reserve holdings can penalize conservative banks and subsidize the dollar at the expense of the peso.
  3. Decentralized LARs are generally preferable. Once introduced, the central bank may still maintain some international reserves and a dollar LOLR capacity to limit runs on individual banks.
  4. The optimal level of LAR—and associated cost—can be reduced by complementing LARs with circuit breakers (CBRs): automatic and early suspension of convertibility of (dollar) deposits to prevent exhausting dollar liquidity and compromising the payments system.
  5. To be successful (stabilizing rather than destabilizing), CBRs should be accompanied by a sound and transparent prudential and bank resolution framework that limits the scope for losses to depositors.
  6. Introduction of CBRs is likely to be smoother if emphasis is put on deposit protection and efficient bank closures instead of deposit restructurings.

*Italic: IMF Working Paper _wp05188 - References (content excerpt).*

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

### _wp05188 - References

### I. Introduction
- Dollarized economies have experienced severe banking and currency crises driven largely by runs on bank deposits, mainly dollar deposits.
- The Argentine crisis (2002) and its spillovers to Uruguay and Paraguay illustrate that dollarization can greatly complicate crisis management and create major financial vulnerabilities.
- Systemic runs in dollarized economies have an important self-fulfilling component, consistent with Diamond and Dybvig (1983) and related literature (Calvo (1988); Cole and Kehoe (1998)).
- Key dynamics:
  - Dollar deposits can only be paid off to the extent banks have sufficient dollar liquidity or access (through lender of last resort, or LOLR) to dollar liquidity on-lent by the central bank.
  - In highly dollarized fractional reserve systems, only a limited proportion of deposits can be backed by central bank dollar liquidity.
  - Runs are exacerbated by firesales of bank assets and by the threat of last-resort measures such as deposit freezes.
  - In mixed-currency economies, additional incentives to run include threats of forced conversions into pesos and large exchange rate depreciations (as occurred during the Argentine crisis of 2002).
  - Floating the exchange rate to stem peso runs can further harm dollar assets and worsen banks’ positions.
- Paper objectives:
  - Review recent experiences with liquidity insurance.
  - Argue in favor of liquid asset requirements (LARs).
  - Present the concept of circuit breakers (CBR) — temporary, efficient, pre-programmed suspension of convertibility — combined with limited earmarked dollar liquidity and automatic reprogramming of term dollar deposits.
  - Conclude that CBR, if well designed and accompanied by prudential policies and a bank resolution framework, could limit the need and cost of dollar liquidity buffers and reduce destabilizing runs.

### II. Can Liquidity Be Borrowed?
- Two broad approaches to insure a dollarized financial system against dollar liquidity shortages:
  - Self-insurance: holding a substantial stock of foreign-currency-denominated liquid assets, either by the central bank or by individual banks (LARs).
  - External insurance: contracts with private providers of dollar liquidity (consortium of financial institutions) or with international financial institutions (IFIs) to ensure access to dollar liquidity at reasonable cost.

- Cost and practical considerations of self-insurance:
  - Self-insurance entails a non-trivial cost equal to the excess financing cost over return on liquid foreign assets, combining a maturity and sovereign risk premia.
  - A back-of-the-envelope estimate: cost ≈ difference between average yield of external debt (e.g., J.P. Morgan’s EMBI index) and return on central bank reserves — a non-negligible number for most emerging economies.
  - If runs were purely self-fulfilling, external insurance would be less expensive than holding liquidity, and a perfectly insured country would be immune to non-fundamental runs.
  - Practical complications:
    - Required insurance packages are typically large and risks difficult to diversify.
    - Liquidity insurance is structured as a loan; solvency of the insured is a precondition, but runs are often accompanied by expectations of losses.
    - Defining the insurable event and verifiable losses is difficult, creating moral hazard.
  - Consequently, insurance premia are likely prohibitively expensive unless the facility is collateralized; collateral availability in emerging economies is generally limited and of dubious value.

### II.B Private Insurance — Recent Experiences and Lessons
- Argentine contingent credit line (late 1990s):
  - Contract allowed central bank and participating local banks to engage in repurchase agreements against Argentine sovereign securities for up to US$6.7 billion (see Box 1).
  - Operational experience:
    - Coverage was relatively limited.
    - Execution was delayed until August 2001, when the run was well underway.
    - Ultimately provided US$1.77 billion (out of US$4.75 billion available at the beginning of 2001).
    - Bond price declines used as collateral implied reductions in the size of the line; line dropped to US$1.35 billion at the first three-month renewal.
    - Result: financing gap exactly when more support was needed.
  - Contract features (Box 1):
    - Maturity: three years with an ever-greening clause extending life every three months.
    - Argentine dollar-denominated bonds taken at 80 percent of market value (if bond price fell by more than 5 percent, more bonds had to be delivered as margin).
    - Annual commitment fee: 32 basis points (bps).
    - Interest on withdrawn funds: roughly LIBOR plus 205 basis points.
    - In 1998, average government bond yield was around 940 bps.
    - Illustrative cost comparison:
      - Increasing reserves by the committed amount would entail a “fee” cost of about 570 bps (alternatively, a flow cost of roughly US$380 million per year), well above the 32 bps commitment fee under the contract.
  - Execution constraints:
    - Decline in collateral value and legal/budgetary restrictions prevented timely issuance of additional collateral.
    - Issuing bonds in private markets was not an option during the run.

- Mexican contingent credit line (Box 2 summary in main text):
  - The line was withdrawn in its entirety by the Mexican government on September 30, 1998 due to deterioration in international capital market access and a decline in oil prices that reduced fiscal resources.
  - Private insurers disputed the government’s decision but ultimately extended the loan; subsequent refusal to extend financing illustrates insurer actions to limit downside exposure under rising risk.

- General lessons from private insurance experiences:
  - Insurers facing poor collateral and rising risk will use contractual rights to limit downside exposure, reducing the effectiveness of the facility when needed most.
  - Insurance premia and contractual terms (margins, valuation rules, frequent renewals) can undermine the intended stabilizing role of private liquidity lines.
  - The combination of market-driven collateral deterioration and contract design can lead to a paradoxical reduction in available liquidity precisely during runs.

### III. Centralized versus Decentralized Liquidity Buffers (overview from introduction)
- Centralizing reserves at the central bank introduces agency costs leading to suboptimal dollar liquidity holdings and implicit subsidies to dollar intermediation.
- Decentralized holdings via LARs internalize currency-risk externalities but forgo diversification benefits of a common liquidity fund.
- For largely systemic shocks, benefits of decentralized LARs outweigh costs; a positive LAR (and one greater than the one corresponding to peso deposits) is optimal from a prudential perspective.
- Once LARs are introduced, a limited dollar liquidity recycling facility can be justified to address idiosyncratic liquidity risk and enhance banking system resilience.

### IV. Circuit Breakers (CBR) — Conceptual Proposal (overview from introduction and concluding remarks)
- CBR: a temporary, efficient, and pre-programmed suspension of convertibility designed to economize on reserves and stop runs with less trauma than severe last-resort measures.
- Core elements of the proposed scheme:
  - Limited dollar liquidity earmarked to ensure full convertibility of transactional dollar deposits.
  - A mechanism that automatically reprograms dollar-term deposits once triggered by a sufficient decline in banks’ liquidity.
- Potential benefits:
  - Reduce the need for a large dollar liquidity buffer and its cost.
  - Narrow scope for destabilizing runs on the banking system.
- Preconditions and complementary policies:
  - Well-designed legal and operational framework (could be de facto implemented without explicit legal reference to deposit restructuring).
  - Adequate prudential policies, including a bank resolution framework.

### Key Statistics and Contractual Figures (preserved exactly)
- US$6.7 billion (Argentine contingent credit line maximum)
- US$1.77 billion (amount ultimately provided by the Argentine line)
- US$4.75 billion (available at the beginning of 2001 for the Argentine line)
- US$1.35 billion (size of the Argentine line at first three-month renewal)
- 80 percent (valuation of Argentine dollar-denominated bonds under the agreement)
- 5 percent (bond price fall threshold triggering additional margin)
- 32 basis points (annual commitment fee under the Argentine contract)
- LIBOR plus 205 basis points (interest rate on withdrawn funds under the Argentine contract)
- 940 bps (average government bond yield in 1998, illustrative)
- 570 bps (illustrative “fee” cost of holding equivalent reserves)
- Roughly US$380 million per year (illustrative flow cost of holding equivalent reserves)
- September 30, 1998 (date Mexican line was withdrawn)

*Italic: IMF Working Paper _wp05188 - References (content excerpt).*

### Box 2. The Mexican Contingent Credit Line

### Box 2. The Mexican Contingent Credit Line

### Description of the Mexican contingent credit line
- In November 1997, the Mexican government subscribed to a contingent credit line with 31 major private international financial institutions.
- Purpose: provide resources to meet external debt service in the event of a closure of the country’s access to international capital markets.
- Facility size: US$2.5 billion (later extended to US$2.66 billion with the addition of two new institutions).
- Availability: for up to 18 months.
- Pricing: at a rate equal to 3M LIBOR plus 50 bps during the first semester, increasing by 25 bps each subsequent semester.
- Cost: an annual commitment fee of 30 bps (or, approximately, US$7.6 million).

### Practical problems and controversial interpretation
- Main practical difficulties that affected timeliness and effectiveness:
  - Difficulty in defining unambiguously the event that triggers the contingency clause.
  - Reluctance of the insurer to assume the costs of the contract.
- Consequence: the contract was subject to controversial interpretations that may have detracted from its timeliness and effectiveness.

### Shortcomings of private insurance (empirical and theoretical)
- Moral hazard and hedging incentives:
  - Insurer banks have a stake in the value of the country and may hedge exposure as crisis probability rises.
  - Hedging by insurer banks (e.g., shorting the country’s assets or the domestic currency) can accelerate collapse or deepen a currency crisis.
- Limitations due to correlated risks and size:
  - If contracts extend to several emerging markets, insurers’ scope to diversify highly correlated regional risks narrows, limiting coverage size.
  - Where currency attacks follow a common pattern across insured countries, there is a limit to private institutions’ willingness to extend coverage.
- Reverse moral hazard / excessive risk-taking by insurers:
  - Insurers attracted by commissions may take on more risk than they can handle and may become reluctant to provide committed resources once a crisis is underway.
- Overall assessment:
  - Private insurance may be useful as a partial complement, but is not a promising alternative for countrywide coverage of systemic liquidity risk.

### Public insurance and IMF experience
- Contracting insurance from nonprofit official organizations (IFIs) or regional arrangements can overcome some private insurance drawbacks (moral hazard, hedging).
- IMF experience:
  - IMF-led packages have provided liquidity insurance imperfectly and reluctantly, aiming broadly to stabilize the capital account rather than specifically to assist local banks.
  - Difficulties in isolating liquidity from solvency concerns have hampered smooth, timely provision of liquidity support.
  - The IMF’s contingent credit line (CCL) required a discretionary pre-qualification process and had limited size; it was never requested and was ultimately phased out.

### Taking stock — implications for policy choice
- Two preliminary conclusions:
  - Private liquidity insurance is likely to be limited and possibly unreliable in substantive shocks due to insurer risk-transfer capacity and limited diversification of country risk.
  - Launching an ad hoc country insurance facility by IFIs appears unrealistic in the near future.
- Resulting realistic option: self-insurance (costly but reliable).
  - Two main self-insurance forms in a liquidity run on the banking sector: liquid assets requirements (LARs) on dollar deposits, or accumulation of central bank reserves.

### Evidence on international reserves (selected figures reproduced)
- Table 1: International Reserves (As percentage of GDP) — selected entries:
  - Mexico: 1992 = 5.1 ; 2002 = 7.7
  - Average EM: 1992 = 8.9 ; 2002 = 18.1
  - India: 1992 = 2.0 ; 2002 = 13.5
  - China, P.R.: Mainland: 1992 = 4.0 ; 2002 = 22.5
  - Argentina: 1992 = 4.2 ; 2002 = 10.2
  - United States: 1992 = 0.6 ; 2002 = 0.3
  - Average Industrials: 1992 = 5.4 ; 2002 = 6.7

### Lender-of-last-resort (LOLR) vs. liquid assets requirements (LAR) — model insights
- Two main model insights:
  - Presence of a LOLR reduces banks’ demand for precautionary liquidity because borrowing cost from LOLR replaces the marginal cost of falling short.
  - Fundamental asymmetry in central bank cost: peso LOLR can be provided at virtually no cost via issuance of peso liabilities, whereas a peso central bank can provide dollar LOLR only to the extent it holds dollar reserves, and dollar LOLR is costly to the central bank.
- Distortions and risk consequences of dollar LOLR:
  - Dollar LOLR acts as a subsidy to commercial banks (at taxpayer expense), lowering cost of dollar deposits and promoting excessive dollar intermediation.
  - Reliable dollar LOLR may subsidize risk-taking: risky banks rely more on LOLR, hold fewer reserves, compete unfairly with conservative banks, and central bank reserves—if binding—tend to support risky banks disproportionately.
- Trade-offs:
  - Eliminating dollar LOLR reduces distortions and moves banks toward optimal dollar liquidity demand, but LOLR remains useful for resolving idiosyncratic crises and limiting contagion due to central bank seniority and coordination role.
  - A central bank that recycles dollar liquidity from liquidity-rich to liquidity-poor banks can improve interbank market functioning and economize on international reserves.

### Policy implications and recommended framework
- Because restricting LOLR use strictly to idiosyncratic events is difficult in practice (contagion can convert idiosyncratic shocks into systemic crises), the second-best policy recommended is:
  - Require banks to hold minimum liquid dollar reserves in proportion to their liquid dollar liabilities (introduce LAR).
- Expected effects of LAR:
  - Ensure banks do not “free ride” on central bank LOLR and international reserves, avoiding shifting liquidity buffer costs to the public sector and undermining central bank solvency.
  - Prevent weaker/riskier banks from unduly benefiting from LOLR at the expense of stronger banks, reducing system fragility.
- Design features:
  - LARs should induce banks to hold socially optimal liquidity levels.
  - If liquidation externalities exist (additional bank liquidation costs not internalized), LAR may need to be higher than private optimal.
  - Eligible assets for LAR should be restricted to foreign assets that benefit from a deep, liquid market to avoid shifting the buffer burden to the public sector.
- Role of central bank dollar LOLR after LAR:
  - Can be credibly limited to exceptional additional support for isolated banks facing very large liquidity shocks.
  - Official international reserves are needed only to the extent central bank obligations to liquidity-supplying banks may need partial backing by foreign reserves, particularly for very short maturities.
  - If a market for short-term dollar treasury bills exists, that market’s liquidity may need minimum backing of international reserves to allow repo operations; under systemic crises, interest rates on these instruments should be allowed to rise to limit their liquidity and contribute to an interest rate defense.

### Preview of circuit breakers (CBR) proposal
- Problem: full backing of all deposits with liquid foreign assets is generally infeasible or too costly; hence need mechanisms for very large, not optimally insurable, dollar liquidity runs.
- Circuit breaker concept:
  - A pre-specified regulatory scheme that limits in an organized and predictable way the convertibility of dollar deposits during pronounced runs.
  - Intended to short-circuit runs at an early stage.
- Motivation:
  - Historical ad hoc measures in dollarized economies (Ecuador, Argentina, Uruguay) were implemented late, improvised, introduced significant uncertainty, and generated litigation and costly fiscal outlays.
  - Institutionalizing CBRs within the regulatory framework and tying them to bank resolution procedures offers advantages:
    - Depositors know rules in advance, limiting ex-post legal action.
    - Creates market incentives for risks to be fully priced and internalized ex ante.

*Source: Box 2. The Mexican Contingent Credit Line (extracted from the supplied content).*

### Box 3:  Introducing CBR-Type Measures Under Crises

### Box 3:  Introducing CBR-Type Measures Under Crises

### Deposit freezes versus compulsory reprogrammings
- Deposit freezes:
  - Depositors lost access to the deposits included in the freeze and did not know how and when they would regain access (no clear mechanism or date were provided in that regard).
  - Depositors did not know the interest rate earned by the frozen deposits during the freeze.
  - In most cases, the freeze was introduced as a temporary measure until a longer-term solution could be found.
- Compulsory reprogrammings:
  - Depositors receive a new financial instrument from the bank with a known maturity, which could be securitized as a CD or medium-term note, to be traded in the secondary and repo markets increasing its liquidity for cash-strapped depositors.
  - If the conditions that led to the freeze are not reverted in the short run, a freeze should eventually evolve into a full reprogramming.

### The Argentine “corralito”: features and drawbacks
- Features:
  - Allowed depositors to maintain full access to their funds as long as they remained within the (local) banking system.
  - Agents could issue checks and transfer funds between bank accounts, from one bank to another, from peso to dollars, and from time to sight deposits.
  - Cash withdrawals were prohibited and transfers abroad were restricted to commercial account transactions (capital controls were introduced).
- Crucial drawbacks:
  - Did not prevent further dollarization of bank deposits as agents anticipated a large devaluation; this deepened banks’ currency mismatch because banks were unable to adjust the currency denomination of their loans.
  - Worsened flight to quality, with nothing preventing runs on the weakest individual banks; the central bank was pressed into providing extensive lending of last resort.
  - Individual bank failures reinforced panic.
  - Failed to shield the real side of the economy: large part of transactions were conducted in cash, which agents started to hoard, making cash increasingly scarce and affecting the payments system.
- Empirical indicator:
  - Figure 1 illustrates the price of liquidity in Argentina 2002, as measured by the cash discount of checks in the informal market, with the liquidity premium ranging up to 12% over the sample dates: 2/11/2002, 2/25/2002, 3/11/2002, 3/25/2002, 4/8/2002, 4/22/2002, 5/6/2002, 5/20/2002, 6/3/2002, 6/17/2002, 7/1/2002, 7/15/2002, 7/29/2002.

### Circuit breakers (CBRs), historical precedent, and acceptance
- Potential benefits of well-designed CBRs:
  - Reduce the cost of a crisis by ensuring that the payment system continues to function, banks remain open, and scarce dollar liquidity is allocated to its best use.
  - Limit vulnerability of the banking system by allowing “most suitable players” to have comparative advantage in banking intermediation.
- Political economy and acceptability:
  - Governments reluctant to introduce CBRs during “tranquil” times for fear of “scaring away” depositors and reducing financial intermediation.
  - Historical precedents from the free banking era suggest option clauses/suspension of convertibility were accepted and sometimes effective (see Box 4).
  - Chile provides a modern example of a bank resolution mechanism similar to a CBR in place since after the 1982 banking crisis; aimed at idiosyncratic risk and so far untested.
  - Analogy with collective action clauses (CACs) in sovereign debt: initial reluctance was not realized in market penalties when CACs were included in issues (examples cited: Mexico, Brazil, Uruguay).
- When CBRs may matter:
  - If runs are triggered by anticipation of inevitable bank losses, presence of the CBR should not significantly affect likelihood of the run.
  - If liquidity crises are purely self-fulfilling and fueled by memory of prior bank-run episodes, the CBR can help prevent banking panics and ensure stability—analogous to CACs reducing sunk costs of protracted debt crises.

### Historical evidence on suspension of convertibility and CBR-like measures (from Box 4)
- European option clause:
  - Allowed banks to “defer the redemption of their notes provided they pay interest for the period of deferment” and printed statements promising higher interest for the period of deferment.
  - Bank of Scotland introduced an option clause in 1730 that remained in place until 1765, when it was outlawed (Shah, 1997).
- United States (National Banking Era) suspensions of convertibility:
  - Calomiris and Gorton describe seven major episodes of suspension (including 1873, 1893, 1907, and 1914).
  - Estimates: the worse loss to depositors during the National Banking Era was 2.1 cents per dollar of deposits; the worse case in terms of banks failing was 1.28 percent during the Panic of 1893 (p. 114).
- Comparative evidence (Dwyer and Hasan, 1999):
  - During the 1861 bank panic in Wisconsin and Illinois, 87 percent of banks in Illinois finally closed, while around 44 percent of the banks closed in Wisconsin.
  - Evidence indicates that suspension of payments decreased noteholders’ losses by about 20 percent.

### Which deposits should be subjected to a CBR? (key considerations)
- Two key questions:
  1. What type and maturity of deposits (transactions, savings, or term deposits; short-term or long-term) should be subject to the CBR?
  2. Which currency should be subject to the CBR?
- Approaches to currency treatment:
  - Differentiate pesos from dollars: in a crisis (if not before), monetary authorities will protect the peso by letting the exchange rate float.
  - Currency-blind approach: treat pesos and dollars in exactly the same way under the regulatory framework.
  - Intermediate approach: pesos and dollars are subjected to different regulations but the central bank will maintain at all times (through its international reserve holdings, monetary policy, and/or peso LOLR) the full convertibility of peso deposits into cash or dollars.

*Source: _wp05188 - Box 3:  Introducing CBR-Type Measures Under Crises*

### Box 5.  A More Recent Example of Pre-Programmed CBR: The Chilean

### Box 5.  A More Recent Example of Pre-Programmed CBR: The Chilean “Narrow Bank” Safety Net

### The Chilean “Narrow Bank” design and mechanics
- Sight deposits (and term deposits of less than 30 days or whose term to maturity is less than 10 days) are fully guaranteed by the Central Bank of Chile (BCCh).
- The BCCh requires that banks hold liquid assets (in the form of central bank debt) against sight deposits in excess of 2.5 times their capital.
- If the guarantee is triggered, the BCCh becomes the most senior claimant on the bank’s assets.
- When a bank is unable to meet its commitments (including as regards its liquid asset requirement) or severe solvency or managerial shortfalls emerge (as defined in the banking law):
  - The bank’s non-sight liabilities are frozen while sight deposits remain fully accessible.
  - Sight deposits are “decoupled” from the rest of the bank together with the corresponding liquid assets plus the BCCh guarantee.
  - This arrangement protects the payments system, mitigates contagion risk of a bank closure, and provides breathing space for an efficient resolution of the non-narrow part of the bank.
- The risk of an unwarranted, last minute expansion of the guarantee is limited by a five-day advance notice required by banks for transferring term deposits into sight deposits.
- The bank resolution system is conditioned to (and shaped by) a creditor agreement ratified by the majority (in terms of claims) of the bank’s non-sight creditors (and the Superintendency of Banks and Financial Institutions).
  - When an agreement cannot be reached, outright liquidation on the entire non-narrow bank is the only possible outcome.

### CBR (Circuit Breaker) options and currency-regime implications
- Option 1: Let the exchange rate go (differentiate pesos and dollars).
  - Fully consistent with a medium-term de-dollarization strategy.
  - Limits the need for foreign reserves.
  - Drawback: a large exchange rate depreciation needed to keep pesos in place will probably undermine solvency of the banking system in a highly dollarized economy, exacerbating preventive runs and complicating crisis management.
  - Unless peso components of banks’ balance sheets can be protected, limiting CBRs to dollars may fail to stabilize peso deposits.
- Option 2: Currency board (local currency legally convertible to a foreign currency at a fixed exchange rate).
  - Main drawback: requires very large international reserves; without them it may not be credible and may unravel.
  - Risk: currency blindness could promote dollarization and produce severe crisis outcomes (as exemplified in recent Argentine experience).
- Option 3: Currency-specific regulatory framework (viable for highly dollarized countries not pursuing rapid de-dollarization).
  - Provided the CBR ensures that dollar deposits are not converted into peso deposits to siphon off dollar reserves, the limited size of peso deposits allows them to be credibly backed and the exchange rate maintained.
  - Pesos need not be subjected to LARs or CBRs; central bank can provide support through conventional means.
  - Regulatory differentiation: apply CBRs only to time deposits while ensuring full convertibility of transactional deposits (into cash or dollars).23
- Currency-specific prudential architecture implications:
  - Differentiated LARs on dollar deposits: banks should hold enough liquid assets to back dollar sight deposits, similar to a narrow banking system.
  - For dollar term deposits, backing could be reduced and banks would rely on CBRs to cope with substantial withdrawals.24
  - Banks/clients could be offered a choice between instruments not subjected to CBRs but with high LARs, and instruments subjected to CBRs but benefiting from low LARs.25
  - The substitutability between LARs and CBRs suggests they should be determined jointly.

### Which deposits to subject to CBRs
- Aim: maintain access to transaction deposits (checking and savings accounts) while restricting access to the most liquid savings instruments that can be withdrawn at short notice.
- Natural choice: apply CBRs only to time deposits, while ensuring full convertibility of transactional deposits (into cash or dollars).23
- Rationale: maintaining access to transactional balances minimizes impact on the payment system; time deposits may be temporary stores of transactional balances but are the best available option during a crisis.23
- Debate on LAR level:
  - Whether LAR needs to be close to, or exactly equal to, 100 percent of deposits is debatable.24
  - Important that the level be significantly higher than encountered in real life cases.24
- To promote long-maturity instruments, LARs for those instruments could be reduced if issued as non-redeemable CDs tradable on the stock exchange (Bolivia’s experience with over two-year maturity savings instruments noted as precedent).25

### Activation: Who should “pull the trigger” and when?
- Two key policy issues:
  1. The point at which the CBR should be activated.
  2. Whether activation should be central (system-wide) or bank-specific.
- Four possible trigger designs:
  - Solution 1: Purely systemic trigger (applies to all banks when systemic liquidity hits a threshold).
    - Advantage: brings run to an early halt across-the-board.
    - Shortcomings:
      - Imposes CBR on banks that may not need it (penalizes conservative banks, erases competitive advantage of large reputable international banks).
      - Questionable on legal grounds absent a verifiable indicator of systemic nature—can lead to litigation costs and delayed action.
  - Solution 2: Purely idiosyncratic trigger (automatic when any individual bank runs out of liquidity).
    - Problem: removes disciplining benefit on bank management and lending policies; discourages depositors; may raise cost of funds and political difficulty of introducing CBRs.
  - Solution 3: Middle-of-the-road: requires both a bank’s liquidity to fall below a threshold and supervisor consent that the underlying cause is systemic.
    - Limits moral hazard and prevents banks from gaming the system.
    - Drawbacks: introduces discretion in defining systemic crises; supervisors may be reluctant to consent due to contagion risk and litigation—likely to bias against activation and limit usefulness.
  - Solution 4: Milder, CBR-like option (Chilean-type deposit insurance preserving transaction deposit liquidity while allowing the rest of the bank to be “temporarily closed”).
    - If prudential requirements are met, the bank is reopened once crisis subsides; otherwise liquidated.
    - Supervisory discretion still involved, but prompt consent more likely as it postpones liquidation rather than restructures deposits—stresses deposit protection and may be easier to sell.
- Policy preference: non-discretionary, automatic triggers determined by transparent, legally uncontestable criteria to avoid costly delays and agency problems.
  - Natural mechanism: link trigger to failure of the bank to comply with a minimum level of liquidity so that LARs assigned to back non-CBR deposits are not compromised.
  - Example logic for dollar time deposits: as dollar time deposits leave the bank or move to sight/peso to elude the CBR, the stock of LARs backing dollar time deposits declines; initially banks cushion declines with excess liquidity or borrowing, but ultimately liquidity falls below the level needed to fully back transaction deposits and the bank is forced to introduce the CBR.

### What happens after CBR activation (“the day after”)
- Effectiveness of CBRs in stopping/preventing runs and restoring confidence depends on whether their introduction creates expectations of forthcoming losses on restructured deposits.
  - If expectations of forthcoming losses are widespread, CBRs may encourage preventive runs.
  - Whether this occurs depends largely on the broader prudential framework and post-CBR actions.
- Key post-CBR issues to address:
  - Rules of the game defining what happens after deposits have been restructured should be well specified and transparent.
  - Specify time frame of restructuring and procedures if liquidity problems are not resolved within that period.
  - Ensure a well-defined, efficient banking resolution procedure that minimizes loss in deposit value if the bank ultimately cannot meet obligations.
- Dollarized-economy-specific concerns and options:
  - Limit scope for cross-currency risk contamination and maintain public confidence in peso banks when dollar liabilities have been restructured.
  - Option resembling Chilean scheme: give peso depositors an official guarantee backed by preferential claims on the residual value of the bank—effectively granting seniority to transaction deposits.
  - Alternative: limit bank’s liability on dollar losses to its dollar capital (i.e., if the dollar bank loses its capital, the capital of the peso bank is not called up).
  - Full “de-coupling” option: separate peso and dollar components of the bank’s balance sheet as soon as CBRs are introduced, transferring dollar assets and dollar capital to a separate shadow balance sheet (potentially a trust fund if not re-incorporated). This would immediately reduce incentives for dollarization.31
  - Note: legal underpinnings of full de-coupling could be complicated.

### Conclusions (summary findings and policy implications)
- The paper addresses management of scarce dollar systemic liquidity in highly dollarized environments and considers: (1) holding vs borrowing liquidity, (2) centralization vs decentralization of liquidity, and (3) the use of pre-wired, transparent circuit breakers (CBRs).
- Key preliminary conclusions:
  1. The current menu of liquidity insurance options is mostly limited to forms of self-insurance (e.g., holding your own liquidity). Market-based (or official) insurance faces obstacles in achieving the right combination of size, cost, and guaranteed access.
  2. Accumulating reserves centrally induces moral hazard as banks, particularly weaker ones, free ride on the central bank and hold suboptimally low levels of dollar liquidity. Centralized reserve holdings can penalize conservative banks and subsidize the dollar at the expense of the peso.
  3. Decentralized LARs are generally preferable. Once introduced, the central bank may still maintain some international reserves and a dollar LOLR capacity to limit runs on individual banks.
  4. The optimal level of LAR—and the associated cost in terms of reduced loanable funds and wider intermediation margins—can be reduced by complementing LARs with circuit breakers (CBRs): the automatic and early suspension of convertibility of (dollar) deposits to prevent exhausting dollar liquidity and compromising the payments system.
  5. To be successful (stabilizing rather than destabilizing), CBRs should be accompanied by a sound and transparent prudential and bank resolution framework that limits the scope for losses to depositors.
  6. The introduction of CBRs is likely to be smoother if emphasis is put on deposit protection and efficient bank closures instead of deposit restructurings.

*Source: Chile FSSA; content as provided in the IMF working paper excerpt.*

### REFERENCES

### _wp05188 - REFERENCES

### Dollarization, currency, and lender-of-last-resort issues
- Broda, C., and E. Levy Yeyati, 2003a, “Dollarization and the Lender of Last Resort,” in Dollarization: Debates and Policy Alternatives, ed. by E. Levy Yeyati and F. Sturzenegger (Cambridge, Massachusetts: MIT Press).
- ________, 2003b, “Endogenous Deposit Dollarization,” Staff Report 160 (New York: Federal Reserve Bank of New York), also forthcoming in Journal of Money, Credit and Banking.
- Gulde, A.M., D. Hoelscher, A. Ize, D. Marston, and G. De Nicoló, 2004, “Financial Stability in Dollarized Economies,” IMF Occasional Paper 230 (Washington: International Monetary Fund).
- Ingves, S., and M. Moretti, 2003, “Banking Failures in Countries Dependent on a Foreign Currency,” Paper presented at the workshop on “Individual Failures of Large Banks: How To Avoid Systemic Crises?” Buenos Aires, unpublished.

### Banking crises, liquidity, and lender-of-last-resort theory
- Calomiris, C., and G. Gorton, 1991, “The Origin of Banking Panics: Models, Facts, and Bank Regulation,” in Financial Markets and Financial Crisis, ed. by Glenn Hubbard (Chicago: The Chicago University Press), pp. 109–73.
- Diamond, D., and P. Dybvig, 1983, “Bank Runs, Deposit Insurance, and Liquidity,” Journal of Political Economy, Vol. 91, No. 3, pp. 401–19.
- ________, and R. Rajan, 2001, “Liquidity Risk, Liquidity Creation, and Financial Fragility: A Theory of Banking,” Journal of Political Economy, Vol. 109, No. 2, pp. 287–327.
- Freixas, X., B. Parigi, and J.C. Rochet, 2003, “The Lender of Last Resort: A 21st Century Approach,” ECB Working Paper 298 (Frankfurt: European Central Bank).
- Dwyer, G., and I. Hasan, 1999, “Suspension of Payments, Bank Failures and the Nonbank Public’s Losses,” unpublished.
- Cole, H., and T. Kehoe, 1996, “A Self-fulfilling Model of Mexico’s 1994–1995 Debt Crisis,” Journal of International Economics, Vol. 41, pp. 309–30.
- Chang, R., and A. Velasco, 1998, “Financial Crises in Emerging Markets,” NBER Working Paper No. 6606 (Cambridge, Massachusetts: National Bureau of Economic Research).

### Public debt, expectations, and fiscal issues
- Calvo, G., 1988, “Servicing the Public Debt: The Role of Expectations,” American Economic Review, Vol. 78, pp. 647–61.
- Ize, A., and G. Ortiz, 1987, “Fiscal Rigidities, Public Debt, and Capital Flight,” Staff Papers, International Monetary Fund, Vol. 23, No. 2, pp. 311–32.
- Zettelmeyer, J., 2000, “Can Official Crisis Lending be Counterproductive in the Short Run?” Economic Notes, Vol. 29, No. 1, pp. 13–29.

### Capital flows, market discipline, and contingent management
- Caballero, R., S. Panageas, 2004, “Contingent Reserves Management: An Applied Framework,” NBER Working Paper No. 10786 (Cambridge, Massachusetts: National Bureau of Economic Research).
- Golfajn, I. and R. Valdés, 1997, “Balance of Payments Crises and Capital Flows: The Role of Liquidity,” Central Bank of Chile Working Paper No. 11 (Santiago).
- Kaminski, G., and C. Reinhart, 1999, “The Twin Crises: The Causes of Banking and Balance-of-Payments Problems,” American Economic Review, Vol. 3, pp. 473–500.
- Levy Yeyati, E., S. Martínez Pería, and S. Schmukler, 2004, “Market Discipline in Emerging Economies: Beyond Bank Fundamentals,” in Market Discipline across Countries and Industries, ed. by W. Hunter, G. Kaufman, C. Borio, and K. Tsatsaronis, (Cambridge, Massachusetts: MIT Press).

### IMF-related work and central bank capitalization
- Cordella, T., and E. Levy Yeyati, 2005, “A (New) Country Insurance Facility,” IMF Working Paper No. 05/23 (Washington: International Monetary Fund).
- Ize, A., 2005, “Capitalizing Central Banks: A Net Worth Approach,” forthcoming in Staff Papers, International Monetary Fund.

### Historical and theoretical examinations of banking systems
- Cowan, T., and R. Kroszner, 1989, “Scottish Free Banking before 1845: A Model for Laissez-Faire?” Journal of Money, Credit, and Banking, Vol. 21, No. 2 (May).
- Shah, P., 1997, “The Option Clause in Free-Banking, Theory and Practice: A Reappraisal,” Review of Austrian Economics, Vol. 10, No. 2, pp. 1–25.

*Source: _wp05188 - REFERENCES*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2005/_wp05188.pdf_