## 1. Summary Statistics for the LMA Variable

## Source details

**Canonical URL:** [1. Summary Statistics for the LMA Variable](https://www.imf.org/-/media/files/publications/wp/2017/wp17246.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2017/wp17246.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2017/wp17246.pdf.json)

---

### Role and risks of Loss of Market Access (LMA)
- Market access enables sovereigns to finance budget deficits and service debt; LMA can trigger capital flight, reserve depletion, exchange rate pressures, and a balance of payments (BOP) crisis.
- Severe LMA manifestations:
  - Gradual or abrupt stop of a country’s financing.
  - Increase in borrowing costs to levels that preclude debt sustainability.
- In complete debt‑distress with LMA, needed actions may include IMF financing under an IMF‑supported program, new financing from other multilateral donors, Paris Club treatment, and debt restructuring (pre or post arrears).

### IMF policy context and implications for lending
- Fund‑supported programs must resolve a member’s BOP problem subject to adequate safeguards and achieve medium‑term viability (restore debt sustainability and reestablish access to capital markets).
- Use of IMF resources requires a joint assessment of market access and public debt sustainability.
- Exceptional access provisions:
  - IMF can provide funding in exceptional access cases even if a member has lost market access if debt is assessed sustainable with “high probability” (expectation: LMA is temporary and IMF financing helps regain access).
  - If debt is unsustainable, exceptional access conditional on a debt restructuring operation sufficiently deep to restore sustainability with high probability.
  - If debt is sustainable but not with high probability, exceptional access may be justified if non‑Fund financing improves sustainability and enhances safeguards; such financing may include financing obtained through any intended debt restructuring.
- Exceptional access quantitative definition: 435 percent of access on a cumulative basis, 145 on an annual basis.

### Conceptual and operational definition of LMA
- May 2013 Board definition cited: ability to tap international capital on a sustained basis through loans and/or issuance of securities across maturities and at reasonable interest rates.
- Operational clarifications:
  - Distinguish inability versus voluntary choice not to tap markets (e.g., fiscal surpluses and low refinancing needs do not constitute LMA).
  - If lack of issuance reflects rationing or deteriorating financing terms (rising interest rates incompatible with debt sustainability), treat as LMA.
  - Drawing down international reserves or using central bank swap lines instead of borrowing during distress is treated as LMA when absence from the market is driven by a deteriorating environment.
- Long‑term instruments for this paper: maturity of more than one year. Inability to borrow long term (only placing very short‑term instruments) → considered LMA.
- Domestic issuance qualifies as international market access only when accessible to nonresidents; country‑specific judgment required.

### Empirical and analytical contributions; seven‑step framework
- Main analytical contributions:
  - Framework to assess whether a country under distress has lost market access and whether loss is temporary or structural.
  - LMA definition accounting for normal‑times issuance behavior and links to debt sustainability assessment (DSA).
  - Examination of potential leading indicators and their predictive power; highest explanatory power found for sovereign bond spreads, nonresident holdings of general government debt, and sovereign credit ratings.
  - Framework to evaluate likely causes for LMA via market pricing, macro/fiscal fundamentals, financial sector health, and global risks.
  - Discussion of implications of temporary vs structural LMA for external assistance.
- Proposed seven‑step framework (summary):
  1. Consider potential leading indicators of LMA.
  2. Determine timing of past LMA episodes through a systematic decision rule.
  3. Evaluate whether selected indicators show systematic patterns prior to LMA.
  4. Test predictive power of indicators using signaling and risk‑zone approaches.
  5. Evaluate additional indicators (CDS spreads, yield term structure, bond price shifts).
  6. Distinguish temporary vs structural LMA.
  7. Assess implications of LMA nature on required degree of adjustment and policy advice.

### Practical observations on determination and use
- Historical IMF determinations relied on a mix of widening spreads, cancellations of planned issuances, rating downgrades, and foreign investor withdrawals without a consistent framework.
- The paper’s framework is intended to inform and structure IMF judgment; it is not formal guidance for lending decisions.

### Box 1 — Tapping markets in times of plenty (practical points)
- Rationale for continued market presence: preserve benchmark yield curve, provide investors safe alternatives, enable rapid scaling up of issuance; negative carry can be small when macro environment favorable.
- Treatment of domestically‑issued debt:
  - “Domestic debt” defined by jurisdiction of issuance.
  - Central bank purchases or liquidity facilities do not constitute true market access.
  - Most sovereigns can issue/rollover very short original maturities of domestic‑currency debt even in distress; inability to rollover short‑term domestic debt signals extreme LMA (examples: Cote d’Ivoire 2011, Russia 1998, Ukraine 1998).
- Operational timing criteria for empirical work (issuance‑based):
  - For regular issuers: start of LMA = (i) expected long‑term issuance/syndicated loan that did not happen or was cancelled; and/or (ii) announcement of default/restructuring; and/or (iii) request for a high‑access or exceptional access non‑precautionary IMF program. End = issuance of a long‑term security or syndicated loan in markets accessible to foreign investors.
  - For sporadic issuers: similar rules but rely on predicted issuance based on rollover/financing needs.
  - Regular issuers defined by pre‑announced schedule or identifiable issuance patterns; sporadic issuers tend to be developing countries.

### Leading indicators and empirical performance (Steps 1–4)
- Potential leading indicators (significant deterioration typically foreshadows LMA):
  - Sovereign spreads: changes well above historical levels across maturity spectrum.
  - Nonresident holdings of public debt: significant and sustained fall.
  - Term structure of bond prices: implied recovery rates and probabilities of default (PoDs).
  - Adverse deviations in primary bond issuance practices: volume, frequency, maturity, financing terms.
  - Government bond rollover rates: sustained falls not due to reduced financing needs.
  - Government cash balances: abnormal decline or greater reliance on central bank financing.
  - Sovereign credit ratings: downgrades.
  - Bond trading activity: thinner volumes, wider bid‑ask spreads.
  - Shape of the yield curve: flattening or inversion.
  - CDS spreads and upfront CDS costs.
- Sample and summary statistics (empirical exercise):
  - Sample: 57 advanced, emerging and frontier market economies, monthly data for 1990–2015.
  - Out of 57 countries, 49 had lost market access at least once over 1990–2015; 24 lost market access more than once.
  - Median duration of market access loss (1990–2015): 41 months.
    - Median duration for episodes that began during 1990s: 36 months.
    - Median duration for episodes that began during 2000–2009: 31 months.
    - Median duration for episodes that began during 2010–2015: 16 months.
  - Example: Ireland ceased issuing long‑term securities starting October 2010 despite gross financing needs of 43 percent of GDP (about 21 percent of GDP due to financial sector support). Ireland re‑accessed via exchanges in 2012 and new syndications in January and March 2013.
- Stylized facts on indicator timing:
  - Spreads start rising about 10 months prior to LMA for the median country; steepest increase occurs 2 months prior.
  - Credit ratings decline starting around 6 months prior to LMA, with more abrupt deterioration after initial loss of market access (analysis covers 22 emerging and advanced economies; ratings averaged across S&P, Moody’s and Fitch).
  - Nonresident holdings decline prior to LMA (Arslanalp and Tsuda database; starts in 2004 so subsample limited).
  - Long‑term local‑currency shares decline prior to LMA; long‑term domestic local‑currency fixed‑rate borrowing declines.

- Predictive power (signaling and risk‑zone approaches):
  - Spreads, credit ratings, nonresident debt share, and changes in borrowing composition (currency and maturity) display greatest signaling power.
  - Spreads have the best predictive power among indicators considered.
  - Risk‑zone findings:
    - Countries in elevated risk zone at time of LMA had been within this zone for a minimum of 3 months prior; high risk zone minimum was 1 month prior.
    - No instances of LMA in the sample where indicators were in low risk zone; few instances where indicators were in depressed risk zone at time of LMA.
  - Example table of hits with each risk zone (spreads and ratings):
    - Spreads:
      - High risk & LMA: 3
      - Elevated risk & LMA: 9
      - IQR & LMA: 5
      - Depressed risk & LMA: 1
      - Low risk & LMA: 0
    - Ratings:
      - High risk & LMA: 4
      - Elevated risk & LMA: 6
      - IQR & LMA: 5
      - Depressed risk & LMA: 3
      - Low risk & LMA: 0

### Additional market‑implied indicators (Step 5)
- CDS spreads and implied PoDs:
  - Example: Ukraine 5‑year CDS rose to more than 4,000 basis points by April 2015.
  - Simple PoD rule of thumb: PoD ≈ CDS spread / (1 − R). Example: using R = 40 percent, Ukraine 5‑year CDS implied PoD reached about 45 percent by start of February 2015. Caveats: formula relies on strong assumptions; CDS illiquidity can bias PoD.
- Bond prices and term structure:
  - Rapidly falling bond prices and convergence across tenors point to high probability of LMA and likely significant haircut. Example: Ukraine international bond prices fell from low 70s to low 60s in December 2014 and converged to low 40s by early May 2015.
  - Yield curve flattening/inversion observed prior to LMA (example: Greece prior to April 2010).
- Market‑implied recovery rates and exit yields:
  - Example Ukraine: estimated average recovery rate pre‑December 2014 about 75 percent; by March 2015 expected recovery rate fell to about 40 percent (Citibank market estimate ~73 percent end‑March 2014 cited for consistency).
  - Upfront CDS costs approximate (1 − R) as default becomes inevitable; convergence across tenors signals expected default timing.
- Use: implied recovery rates and exit yields help assess whether LMA likely short‑lived or will require restructuring and to gauge likely NPV haircut size.

### Distinguishing temporary vs structural LMA (Step 6)
- Temporary LMA: driven by market condition changes not directly related to debt/fiscal fundamentals (external shocks, contagion, political instability); typically short‑lived.
- Structural LMA: accompanied by significant deterioration in debt/fiscal fundamentals or a financial crisis with serious fiscal implications; unlikely to resolve without major fiscal adjustment and/or debt treatment.
- Historical patterns:
  - LMA duration depends on causes: contagion/global risk aversion typically short; episodes tied to deterioration in fiscal/debt sustainability and growth outlook are longer and often require fiscal adjustment and sometimes restructuring.
  - Governments with lower debt‑to‑GDP at time of LMA re‑access faster than those with higher debt‑to‑GDP.
  - Moody’s (2013) finding: sovereigns on average did not issue in international markets for 5.6 years after default and 4.4 years after final default resolution; 45 percent of defaults never regained market access during the studied period.
  - Cases with small NPV losses and prudent policies re‑accessed relatively quickly (examples: Uruguay, Dominican Republic); large nominal/NPV reductions led to longer re‑access times.

### Policy and operational implications (Step 7 and selected guidance)
- Continued market presence in times of plenty recommended for regular issuers to preserve market functioning and avoid costly re‑introduction of issuance infrastructure.
- IMF assessments of market access should:
  - Incorporate both issuance‑based criteria and market indicators (spreads, ratings, CDS, nonresident participation, issuance patterns).
  - Exercise country‑specific judgment on whether domestic issuance accessible to nonresidents constitutes international market access.
  - Use signaling and risk‑zone frameworks to identify elevated risk of LMA, with spreads typically the most powerful predictor.
- Distinguish temporary vs structural LMA by jointly assessing market access indicators and DSA:
  - Structural LMA may necessitate debt treatment and major fiscal adjustment.
  - Temporary LMA may be addressed through catalytic IMF lending and macroeconomic adjustment.

### Box 2 — Market access and debt sustainability: diagnostic logic and operational use
- Conceptual distinctions:
  - Sustainable debt: can be stabilized in baseline and realistic shock scenarios with a primary balance that is economically and politically feasible.
  - Market access: ability to raise funds at acceptable terms in international and, in some cases, domestic markets.
- Interaction and heat‑map logic:
  - If market access and DSA indicators both cross vulnerability thresholds → LMA likely structural → require major policy adjustment and/or debt treatment.
  - If only market access indicators cross thresholds while DSA does not → LMA likely temporary.
  - Forward‑looking indicators (recovery rates, implied exit yields) help indicate whether restructuring likely needed and likely haircut size.
- MAC DSA and heat map:
  - MAC DSA complements basic DSA for higher scrutiny countries with realism checks, heat‑map of risks, stochastic simulations; thresholds used are illustrative but the MAC DSA uses a wider set of indicators and scenarios.
  - Use cases: early warning, signal impending structural LMA, guide policy/preventive IMF engagement.
- Judgment required:
  - Even if debt thresholds not breached, LMA may be assessed structural because of factors such as major political regime changes with solvency risk.

### Annex III — Signaling approach and extracting term‑structure information
- Signaling approach vs multivariate probit:
  - Signaling approach: quick, better out‑of‑sample performance, tolerant of data gaps, cannot test conditional statistical significance.
  - Threshold selection methods: Minimization of total misclassified errors (TME) or maximization of signal‑to‑noise ratio; TME gives greater weight to misclassifying crisis events.
- Risk‑zone classification example for spreads:
  - Percentiles computed at τ = {5th, 25th, 50th, 75th, 95th} denoted a, b, c, d, e.
  - Zones: High risk: X ≥ e; Elevated risk: e > X ≥ d; IQR: d > X ≥ b; Low/Depressed risk: a > X ≥ b and X ≤ a.
  - Choice of the “high risk” percentile τ can be tuned to historical LMA episode concentration.
- Term‑structure extraction:
  - Joint estimation of recovery rate R and default intensity λ by minimizing weighted fitting errors between model prices and observed bond prices using a risk‑free discount curve (Svensson curves) and weighted by inverse Macauley duration.
  - Estimated recovery rates robust to survival function assumptions; robustness checks include fixing λ between 0.5 and 0.9.
- Restructuring scenarios conditional on market expected recovery rate:
  - Market‑priced recovery rate R̄ implies upper bound on haircut = 1 − R̄.
  - Haircut definitions and sensitivity to exit yield S̈ emphasized; derive S̈ consistent with market bounds by quadratic minimization.
  - Illustrative scenarios and recovery values used in charts: Rec: 40%, 42%, 44%, 46%, 48%, 50% (Chart A); Rec: 40%, 50%, 60%, 70% (Chart B).

### Annex VIII — IMF lending policies (summary highlights)
- Exceptional access framework:
  - Exceptional access may be provided where rigorous analysis indicates a high probability public debt is sustainable in the medium term.
  - If debt sustainable but not with high probability, exceptional access can be justified if non‑Fund financing improves sustainability and enhances safeguards; non‑Fund financing may include financing obtained through intended debt restructuring.
  - If debt unsustainable ex ante, exceptional access only where non‑Fund financing restores sustainability with high probability.
  - Exceptional access defined quantitatively as 435 percent cumulative, 145 annual.
- Recent reforms (Press Release No. 16/3, January 29, 2017):
  - Elimination of the “systemic exemption” (introduced 2010).
  - Increased flexibility where debt assessed sustainable but not with high probability.
  - Clarified market access criterion.
- Application across debt situations:
  - If debt sustainable with high probability: provide financing in support of strong adjustment and payment of obligations as they fall due (including temporary LMA).
  - If debt clearly unsustainable: prompt and definitive debt restructuring to restore sustainability with high probability is least‑cost approach.
  - If sustainable but not with high probability: case‑by‑case options include retained market access, reprofiling for significant private claims falling due, or reliance on official financing with favorable terms.
- Timing and flexibility:
  - Ideally reprofiling undertaken before Fund arrangement approval; flexibility exists to conclude debt operation later if warranted.
- Exceptional access and spillovers/tail risks:
  - For rare tail‑event cases where restructuring poses unmanageable domestic or cross‑border risks, Fund approval of exceptional access without restructuring permitted if official sector partners supply financing on sufficiently favorable terms to restore sustainability with high probability; such use expected to be rare and decisions evenhanded.
- Role of DSA and market‑access criterion:
  - DSA central and forward‑looking, considering policy implementation prospects, growth, contingent liabilities, creditor base, investor confidence, global outlook.
  - Market access condition must be met even with open‑ended official support beyond program period; Fund expects regaining market access within a timeframe facilitating repayment of all obligations to the Fund.

*Source: wp17246 (excerpt).*

### 1. Summary Statistics for the LMA Variable ..........................................................................18

### 1. Summary Statistics for the LMA Variable

### H3: Role and Risks of Loss of Market Access (LMA)
- Market access is critical for sovereigns to finance budget deficits and service debt; loss of market access (LMA) can trigger capital flight, depletion of reserves, and exchange rate pressures, and can lead to a balance of payments (BOP) crisis.
- Severe LMA may manifest as:
  - a gradual or abrupt stop of a country’s financing, or
  - an increase in borrowing costs to levels that preclude debt sustainability.
- In complete debt-distress with LMA, a country may need IMF financing, often under an IMF-supported program, combined with new financing from other multilateral donors, Paris Club treatment, and debt restructuring (pre or post arrears).

### H3: IMF Policy Context and Implications for Lending
- Under current policy, a Fund-supported program must resolve a member’s balance of payments problem subject to adequate safeguards and achieve the member’s medium term viability (i.e., restore debt sustainability and reestablish access to capital markets).
- The use of IMF resources requires a joint assessment of market access and the degree of sustainability of the member’s public debt.
- The IMF can provide funding in exceptional access cases even if a member has lost market access, if its debt is assessed to be sustainable with “high probability.” In such cases, the expectation is that LMA is temporary and IMF financing can help the country regain market access.
- If debt is assessed as unsustainable, exceptional access would be conditional on a debt restructuring operation sufficiently deep to restore debt sustainability with high probability.
- If the member’s debt is considered sustainable but not with a high probability, exceptional access may be justified if financing from sources other than the Fund improves debt sustainability and sufficiently enhances safeguards for Fund resources; such financing may include financing obtained through any intended debt restructuring.

### H3: Conceptual and Operational Definition of LMA
- The May 2013 IMF Board paper definition cited: “the ability to tap international capital on a sustained basis through the contracting of loans and/or issuance of securities across a range of maturities, regardless of the currency denomination of the instruments, and at reasonable interest rates.”
- Operational challenges:
  - The comprehensive definition is not operationally precise; under the most demanding interpretation only a handful of advanced and emerging market countries would meet the standard of issuing in international markets on a sustained basis.
  - Distinction between regular and sporadic issuers is key: infrequent international issuers would not necessarily be in continuous LMA.
- Important clarifications for operationalizing LMA:
  - Distinguish between inability versus voluntary choice not to tap markets (e.g., running fiscal surpluses and very low refinancing needs would not constitute LMA).
  - If lack of issuance reflects debt amount rationing or deteriorating financing terms (rising interest rates incompatible with debt sustainability), the lack of issuance should be considered LMA.
  - Drawing down international reserves or using central bank swap lines instead of borrowing during distress is treated as LMA when absence from the market is driven by a deteriorating environment.

### H3: Empirical and Analytical Contributions of the Paper
- Main contributions:
  - Provides an analytical framework to assess whether a country under distress has lost market access, and whether such loss is temporary or structural.
  - Defines LMA taking into account the extent to which sovereigns tap international capital in normal times and outlines links to the debt sustainability assessment (DSA).
  - Examines potential leading indicators and evaluates their predictive power in signaling past LMA cases.
    - Findings indicate the indicators with highest explanatory power in past cases are sovereign bond spreads, nonresident holdings of general government debt, and sovereign credit ratings.
  - Presents a framework to evaluate likely causes for LMA by analyzing market pricing, macroeconomic and fiscal fundamentals, financial sector health, and global risks.
  - Discusses implications of temporary versus structural LMA for appropriate external assistance and highlights application of the framework for identifying emerging risks to market access.
- The paper proposes a seven-step framework for assessing LMA.

### H3: Practical Observations on Determination and Use
- In practice, determinations of LMA in IMF-supported programs have relied on a range of considerations (e.g., widening spreads, cancellations of planned bond issuances, rating downgrades, withdrawal of foreign investors), but no explicit decision-making framework had been consistently applied.
- The framework developed by this paper is intended to inform, complement, and provide structure to IMF judgment when assessing market access.
- The paper is not intended as a formal guidance note for IMF lending decisions.

*Source: wp17246 - 1. Summary Statistics for the LMA Variable (excerpt).*

### Box 1. Tapping Markets in Times of Plenty

### Box 1. Tapping Markets in Times of Plenty

### Rationale for continued market presence
- Maintain benchmark yield curve for corporate debt, provide investors with relatively safe investment alternatives, and enable quick scaling up of issuance if financing needs arise.  
- Continued issuance when no fiscal financing need exists implies a fiscal cost (negative carry), especially for EMs; however, with favorable macroeconomic environment and no financing need, net interest cost likely to be low and may pay off when financing needs arise.  
- Recent crisis experience: Ireland (pre-crisis budget surplus, small debt) had to reintroduce issuance procedures and instruments when gross funding needs surged; Italy (pre-crisis deficit and high debt) absorbed additional issuance through limited adjustments to established mechanisms and instruments.

### B. Accessing international capital—treatment of domestically‑issued debt
- “Domestic debt” is defined by jurisdiction of issuance, i.e. “domestically‑issued debt.”  
- Fund’s mandate emphasizes international dimension (ability to tap “international capital”); domestic issuance may not constitute true market access if authorities can force domestic institutions to buy sovereign debt.  
- Central banks may purchase domestic debt or provide liquidity facilities using sovereign debt as collateral; such purchases do not constitute true market access.  
- Most sovereigns can issue/rollover very short original maturities of domestic‑currency debt even in distress; exceptions include Cote d’Ivoire (2011), Russia (1998), Ukraine (1998) which defaulted on domestic T‑bills. Inability to rollover short‑term domestic debt signals an extreme case of LMA.  
- Distinction between domestic and international issuance is less important in advanced economies and some EMs that borrow abroad in own currency or have high foreign participation in domestic markets; domestic issuance accessible to nonresidents can constitute international market access. IMF staff must exercise country‑specific judgment.

### C. Contracting loans and issuance across maturities
- Market access requires ability to contract long‑term loans or issue long‑term securities. For this paper, long‑term instruments are those with maturity of more than one year.  
- If sovereign can only place very short‑term instruments and cannot borrow long term, it is considered to be in LMA.

### D. Taking into account degree of market access in normal times
- To assess LMA one must first evaluate a country’s pre‑crisis market access: interest rates, issuance frequency, amounts relative to funding needs, tenors, and currency denomination.  
- Initial conditions differ for regular vs. infrequent (sporadic) issuers; significant deviation from established issuance patterns indicates possible LMA.  
- For this empirical work, the end of LMA is defined as issuance of a new long‑term security (or syndicated loan) in international market (or domestic market accessible to foreign investors). A more stringent standard would require resumption of normal issuance patterns including pre‑crisis spreads.

### E. Traditional LMA definition in empirical literature
- Conventional definition: period between announcement of default/restructuring and next international bond issuance. Empirical studies use primary gross issuance in international bonds, equity, and syndicated loan markets.  
- Conventional definition does not capture LMA episodes not associated with debt distress (e.g., Brazil, Ireland, Portugal) and does not account for whether absence of issuance is explained by benign factors (e.g., fiscal surpluses).

### F. Expanded LMA definition used in this paper
- Real‑time assessment: LMA if primary or secondary market spreads rise to levels that, if permanent, would render borrowing unsustainable (consistent with IMF Board Paper May 2013 emphasis on “reasonable interest rates”).  
- Empirical work defines dependent variable (timing of LMA) using issuance‑based criteria and treats spreads/yields only as explanatory variables. Requests for non‑precautionary IMF‑supported programs are attributed to LMA (operationalized below).  
- Operational timing criteria used in empirical work:
  - For governments that are regular issuers:
    - Start of LMA coincides with either (i) an expected issuance (based on auction calendar, past patterns, or rollover needs) of long‑term international or domestic securities (accessible to nonresidents), or a syndicated loan with long‑term maturity, that did not happen or was cancelled; and/or (ii) announcement of a default or restructuring; and/or (iii) request for a high‑access or exceptional access non‑precautionary IMF‑supported program.
    - End of LMA marked by issuance of a long‑term security or syndicated loan in markets accessible to foreign investors.
  - For governments that are sporadic issuers:
    - Start of LMA marked by either (i) a predicted issuance (based on rollover and financing needs) of a long‑term international security or syndicated loan that did not happen; and/or (ii) announcement of a default or restructuring; and/or (iii) request for a high‑access non‑precautionary IMF‑supported program.
    - End of LMA marked by issuance of a long‑term security or syndicated loan in markets accessible to foreign investors.
  - Regular issuers are those that issue according to a pre‑announced schedule or have clearly identified issuance patterns (they do not have to be frequent issuers). Sporadic issuers tend to be developing countries.

### III. An operational methodology for assessing LMA—seven steps (overview highlights)
- LMA defined as inability to tap international capital markets; includes restructuring/rescheduling and unexplained cessation of primary issuance. The seven‑step approach evaluates leading indicators, tests predictive power of common indicators (spreads and credit ratings), and identifies supplementary indicators. It also examines relationship between LMA and DSA to distinguish temporary vs. structural LMA.

H3: Step 1: Potential leading indicators of LMA
- Indicators whose significant deterioration typically foreshadows LMA:
  - Sovereign spreads: changes well above historical levels across maturity spectrum.  
  - Nonresident holding of public debt: significant and sustained fall.  
  - Term structure of bond prices: implied recovery rates and probabilities of default (PoDs) signaling distress.  
  - Adverse deviations in primary bond issuance practices (volume, frequency, maturity, financing terms) relative to “normal” circumstances:
    - Volume: compare with (i) total financing needs; and (ii) announced bond auction schedule.  
    - Frequency: compare with (i) average frequency of issuances; and (ii) auction schedule (auctions cancelled/delayed).  
    - Maturity: compare with average original maturity of instruments. Inability to issue even short‑term domestic debt may signal extreme LMA.  
    - Financing terms: compare recent financing terms with past placements (e.g., shift from fixed to variable rates).  
  - Government bond rollover rates: sustained falls not attributable to reduced financing needs.  
  - Government cash balances: abnormal decline in fiscal balances or cash balances, or greater reliance on direct central bank financing (note: seasonal/idiosyncratic patterns caution).  
  - Sovereign credit ratings: downgrades indicating lost creditworthiness.  
  - Bond trading activity: thinner trading volumes, wider bid‑ask spreads.  
  - Shape of the yield curve: flattening or inversion may signal sovereign debt problems and LMA.  
  - CDS spreads: steep increases signal higher default risk.  
  - Upfront CDS contracts: increases and convergence in upfront costs.

H3: Step 2: Determine timing of past LMA episodes (dependent variable)
- Decision rule focuses on objective criteria—announcement of default/restructuring, request for high‑access non‑precautionary IMF program, and complete cessation of primary issuances by regular issuers that cannot be explained by benign factors (e.g., prefunding, fiscal surplus).  
- For practical analysis, complete cessation of primary issuance is used as an objective indicator of LMA; in real time IMF staff can also determine “partial LMA” from indicator behavior.

Key sample and summary statistics (empirical exercise)
- Sample: 57 advanced, emerging and frontier market economies, monthly data for 1990–2015. Data sources: BEL and Dealogic databases, de Broeck‑Guscina (2012) for euro area, fiscal fundamentals from WEO and VEE.  
- Out of 57 countries, 49 had lost market access at least once over 1990–2015; 24 lost market access more than once.  
- Median duration of market access loss (1990–2015): 41 months.  
  - Median duration for episodes that began during 1990s: 36 months.  
  - Median duration for episodes that began during 2000–2009: 31 months.  
  - Median duration for episodes that began during 2010–2015: 16 months.  
- Example: Ireland ceased issuing long‑term securities starting October 2010 despite gross financing needs of 43 percent of GDP (half of gross financing need due to financial sector support amounting to about 21 percent of GDP). Ireland’s re‑access proceeded via exchanges in 2012 and new syndications in January and March 2013.

H3: Step 3: Past performance of leading LMA indicators (stylized facts)
- Spreads: start rising about 10 months prior to LMA for the median country; steepest increase occurs 2 months prior. Spreads higher when markets price in probability of face‑value cut restructuring/reprofiling; spreads of countries that avoid face‑value cuts are lower prior to LMA and decline quicker after adjustment/IMF catalytic lending.  
- Credit ratings: decline starting around 6 months prior to LMA, with more abrupt deterioration after initial loss of market access. (Analysis covers 22 emerging and advanced economies; ratings averaged across S&P, Moody’s and Fitch.)  
- Nonresident holdings of government debt decline prior to LMA (Arslanalp and Tsuda database; database starts in 2004 so subsample limited). Examples: Greece (2010), Mexico (1994), Portugal (2011), Ireland (2010) showed reductions.  
- Currency/maturity composition: long‑term local‑currency shares decline prior to LMA as sovereigns find it difficult to place longer‑term local‑currency instruments even domestically; long‑term domestic local‑currency fixed‑rate borrowing declines (examples: Greece 2010, Portugal 2011, Ireland 2010).

H3: Step 4: Test predictive properties of selected indicators (signaling and risk‑zone approaches)
- Signaling approach chosen for out‑of‑sample performance and to accommodate data gaps; used to minimize Type I and Type II errors. It cannot test conditional statistical significance of variables. Risk‑zone approach complements signaling by accounting for country heterogeneity.  
- Preliminary results: spreads, credit ratings, nonresident debt share, and changes in composition of borrowing (currency and maturity) display greatest signaling power. Spreads have the best predictive power among indicators considered.  
- Risk‑zone findings:
  - Countries in elevated risk zone at time of LMA had been within this zone for minimum of 3 months prior to LMA; high risk zone minimum was 1 month prior.  
  - No instances of LMA in the sample where indicators were in low risk zone; few instances where indicators were in depressed risk zone at time of LMA.  
- Table 2: Number of hits with each risk zone (instances where LMA occurred while indicator was in zone)
  - Spreads:
    - High risk & LMA: 3
    - Elevated risk & LMA: 9
    - IQR & LMA: 5
    - Depressed risk & LMA: 1
    - Low risk & LMA: 0
  - Ratings:
    - High risk & LMA: 4
    - Elevated risk & LMA: 6
    - IQR & LMA: 5
    - Depressed risk & LMA: 3
    - Low risk & LMA: 0

H3: Step 5: Additional indicators and market‑implied information
- CDS spreads: steep increases signal higher default risk; example: Ukraine 5‑year CDS rose to more than 4,000 basis points by April 2015.  
- CDS‑implied PoDs: simple rule of thumb PoD = CDS spread / (1 − R). Example: using R = 40 percent, Ukraine 5‑year CDS implied PoD reached about 45 percent by start of February 2015. Caveats: formula relies on unrealistic assumptions; CDS market illiquidity in distress can bias PoD.  
- Bond prices and term structure:
  - Rapidly falling bond prices and convergence across tenors point to high probability of LMA and likely significant haircut. Example: Ukraine international bond prices fell from low 70s to low 60s in December 2014 and converged to low 40s by early May 2015. Flattening of price curve supports view bonds would be dealt on equal terms regardless of maturity in restructuring/default.  
  - Yield curve behavior: flattening and eventual inversion prior to LMA (example: Greece prior to April 2010). Monitoring yield curve shape is recommended as leading indicator.  
- Implied recovery rates and exit yields from term structure:
  - Using an implied default probability model for bond valuation, one can estimate expected recovery rates from market prices to assess whether LMA is likely short‑lived or will require restructuring and to gauge NPV haircut size. Example: Ukraine estimated average recovery rate pre‑December 2014 about 75 percent (consistent with market estimates such as Citibank’s ~73 percent end‑March 2014). By March 2015 expected recovery rate fell to about 40 percent.  
- Upfront CDS costs:
  - When CDS protection is quoted as an upfront payment, upfront cost V approximates (1 − R) as default becomes inevitable; R can thus be inferred. Convergence of 1‑year and 5‑year upfront costs signals market moving expected default closer in time (example: Ukraine). Upfront costs may display discontinuities during heightened uncertainty—useful signal for LMA.  
  - Recovery rates implied by upfront CDS and bond pricing tend to be consistent (Ukraine estimate around 42 percent at end of sample).

H3: Step 6: Distinguish temporary vs. structural LMA
- Temporary LMA: loss of funding access driven by changes in market conditions not directly related to debt/fiscal fundamentals (external shocks, contagion, political instability); typically short‑lived.  
- Structural LMA: loss of funding access accompanied by significant deterioration in debt/fiscal fundamentals or a financial crisis with serious fiscal implications; unlikely to resolve without major fiscal policy adjustments and/or debt treatment.  
- Historical patterns:
  - LMA duration depends on underlying causes: contagion/global risk aversion episodes typically short; episodes tied to deterioration in fiscal/debt sustainability and growth outlook are longer and often require fiscal adjustment and sometimes debt restructuring.  
  - Governments with lower debt‑to‑GDP at time of LMA re‑access faster than those with higher debt‑to‑GDP.  
  - Duration after default correlated with size of NPV loss: Moody’s (2013) found sovereigns on average did not issue in international markets for 5.6 years after default and 4.4 years after final default resolution; 45 percent of defaults never regained market access during the studied period. Cases with small NPV losses and prudent policies re‑accessed relatively quickly (examples: Uruguay, Dominican Republic); large nominal/NPV reductions led to longer re‑access times. Figure 17 (in source) illustrates time to market re‑access vs NPV reduction for restructuring/reprofiling cases.

H3: Policy and operational implications (selected)
- Continued market presence in times of plenty recommended for regular issuers to preserve market functioning and avoid costly re‑introduction of issuance infrastructure.  
- IMF assessments of market access should:
  - Incorporate both issuance‑based criteria and market indicators (spreads, ratings, CDS, nonresident participation, issuance patterns).  
  - Exercise country‑specific judgment on whether domestic issuance accessible to nonresidents constitutes international market access.  
  - Use signaling and risk‑zone frameworks to identify elevated risk of LMA, with spreads typically the most powerful predictor.  
- Distinguish temporary vs. structural LMA by jointly assessing market access indicators and debt sustainability; structural LMA may necessitate debt treatment and major fiscal adjustment whereas temporary LMA may be addressed through catalytic IMF lending and macroeconomic adjustment.

*Source: IMF staff estimates and analysis as presented in Box 1 of the source content.*

### Box 2. Market Access and Debt Sustainability—Differences and Similarities

### Box 2. Market Access and Debt Sustainability—Differences and Similarities

### Conceptual distinctions and linkages
- Sustainable debt: debt that "can be at least stabilized in baseline and realistic shock scenarios with a primary balance that is economically and politically feasible."
- Market access: a country's ability "to raise necessary funds at acceptable terms (cost and risk) in the international and, in some cases, domestic markets."
- Interactions:
  - Market access affects debt sustainability via the interest rate channel and the resulting structure of government borrowing.
  - Debt sustainability analysis can influence market access assessments.
- Key determinants of market access beyond debt burden and fiscal indicators:
  - Investors’ expectations about ability and willingness to service debt.
  - Global shocks, sudden increases in risk aversion, herd behavior, or domestic shocks can cause loss of market access even with low debt ratios.
  - Some countries retain market access despite "an immense debt burden" when they: have reserve currency status, can borrow abroad in their own currencies, and have a large share of domestically-held debt.
- Sovereigns may temporarily borrow at rates "incompatible with debt sustainability" to:
  - Preserve sustained market access if a spike in rates is believed temporary.
  - Meet severe funding needs that cannot be postponed.

### Distinguishing temporary vs. structural Loss of Market Access (LMA)
- Use of combined indicators: market access indicators (spreads and credit ratings) together with DSA/fiscal indicators helps distinguish temporary vs. permanent LMA.
- Heat map logic (illustrative):
  - When both market access and DSA indicators cross vulnerability thresholds → LMA likely permanent/structural, requiring major policy adjustment and/or debt treatment.
  - When only market access indicators cross thresholds while DSA indicators do not → LMA likely temporary.
  - Forward-looking indicators (e.g., recovery rates, implied exit yields) provide information on whether LMA is expected to be short-lived or to require debt restructuring and the likely size of restructuring.
- Empirical examples:
  - Brazil: two LMA episodes were temporary; market-based LMA indicators crossed into elevated risk zone (especially 2002) while primary balance remained in low risk zone.
  - Argentina: 1995 LMA episode relatively short lived as fiscal fundamentals remained below critical thresholds; 2001 episode was structural and resulted in a debt restructuring operation because market-based and debt sustainability indicators deteriorated together.

### MAC DSA, heat map, and operational use
- MAC DSA complements basic DSA for higher scrutiny countries with:
  - Assessment of realism of baseline assumptions.
  - A heat-map presenting risks to debt levels, GFNs, and profile of debt.
  - Stochastic simulations of debt paths.
- Note: "The heat map presents only two DSA indicators for illustrative purposes. In reality, the MAC DSA uses a much wider set of indicators as well as a set of shock scenarios, stochastic projections, and staff judgement to assess debt sustainability. The thresholds for both fiscal indicators used are the same as in the MAC DSA."
- Use cases:
  - Forward-looking tool/early warning exercise to identify country-specific vulnerabilities to LMA.
  - Signal impending structural LMA so that a country can act to prevent or mitigate it, or approach the IMF for support.
  - Annex VII and April 2016 GFSR referenced as applications (not reproduced here).

### Step 7 and implications for required degree of adjustment
- Integrated assessment outcome and policy implications:
  - If debt sustainability indicators are below indicative thresholds but market access indicators are not → loss of market access likely temporary; "just funding support is needed."
  - If both debt sustainability and market access thresholds are breached → loss of market access may be structural; sovereign is a potential candidate for "a strong fiscal adjustment and/or debt treatment (reprofiling or face-value cut restructuring) as a condition for an exceptional access program."
  - If market-based indicators appear safe but DSA indicators signal problems → assess financial sector for contingent liabilities that could rapidly erode investor confidence and cause LMA.
- Judgment required: even if debt thresholds are not breached, LMA may be assessed structural due to major political regime changes that could lead to solvency problems.

### Seven-step methodological framework for analyzing LMA (summary)
- Steps outlined in the paper:
  1. Consider potential leading indicators of LMA.
  2. Determine timing of past LMA episodes through systematic decision rule.
  3. Evaluate whether past performances of selected indicators follow a systematic pattern prior to LMA.
  4. Test predictive power of selected LMA indicators using signaling and risk zone approaches.
  5. Evaluate additional indicators of LMA (e.g., CDS spreads, shifts in yield term structure, falling/converging bond prices).
  6. Distinguish between temporary and permanent (structural) LMA.
  7. Assess implications of LMA and its nature on required degree of adjustment and policy advice.
- LMA definition used for this analysis:
  - Broad coverage: includes periods of unexpected lack of primary debt issuance that cannot be explained by benign factors (lack of funding needs, prefunding), in addition to traditional definition (time between announcement of default/restructuring and next international bond issuance).
- Classification rules:
  - LMA can be judged temporary if caused by external shocks such as contagion.
  - LMA can be judged structural if accompanied by serious deterioration in debt sustainability indicators or by a financial crisis likely to have significant fiscal costs.

### Annex I — Decision rule highlights for identifying LMA empirically
- Stage 1: Identify LMA periods associated with default or restructuring.
  - a. Code as LMA all time periods following default or announcement of restructuring/reprofiling (whichever is earlier).
  - b. A country remains in LMA until the next issuance of long-term debt instrument.
- Stage 2: Identify additional LMA periods not associated with defaults/restructurings.
  - 2A. Identify suspect cases for LMA by evaluating gross issuance data of long-term securities and/or syndicated loans for deviations from established patterns.
    - Requirement: "At least 3 years-worth of data are necessary for countries that issue on an annual basis (a shorter period is sufficient for countries that issue at a monthly or quarterly frequency)."
    - If no issuance patterns can be identified (sporadic issuers), define LMA as in Stage 1.
    - For regular issuers, suspect LMA if they deviate from established patterns:
      - Monthly/quarterly issuers: suspect any period when sovereign does not issue (ignore 1 month blips).
      - Annual issuers: suspect the year the sovereign does not issue.
  - 2B. Narrow suspects by checking for benign explanations:
    1. Lack of issuance explained by lack of funding needs?
      - a. Is the country running fiscal surpluses?
      - b. Is gross financing need smaller than 15 percent of GDP?
      - If both answers "Yes" → clear suspect (not LMA).
      - Exceptions exist (e.g., very small deficits like Ireland in 2003 with -0.1 percent of GDP).
    2. Lack of issuance explained by prefunding?
      - Prefunding rules (either):
        - a. Prefunding large enough to cover primary deficit and amortization needs over suspect period.
        - b. Prefunding 2 standard deviations higher than the 12-month historical average.
        - c. Prefunding occurs prior to suspect LMA period and can span a few months (e.g., two consecutive bonds whose sum meets (a) or (b)).
      - Exception: near-prefunding that coincides with decreasing funding needs may be considered legitimate prefunding.

### Empirical application and concluding remarks
- The framework was applied to past LMA cases and shown useful for:
  - Identifying risks to LMA.
  - Determining likely nature (temporary vs structural).
  - Informing required degree of possible adjustment and policy advice.
- Limitations and further work:
  - Distinguishing temporary versus structural episodes is not fully resolved; statistical analysis to infer the distinction with high certainty is left for future research.
  - The approach "does not perform as well for advanced countries" because they tolerate higher debt burdens and have alternative financing sources.
- Practical value: combining market-access and DSA indicators helps form a sound judgment on LMA occurrence and supports early detection of sovereign debt distress to guide fiscal adjustment, debt treatment, and IMF involvement.

*Source: Box 2. Market Access and Debt Sustainability—Differences and Similarities, wp17246*

### Annex III. Signaling Approach for Identifying Vulnerability Indicators

### Annex III. Signaling Approach for Identifying Vulnerability Indicators

### Signaling approach versus multivariate probit
- Two approaches to assess the probability of an imminent LMA:
  - Conduct multivariate probit econometric analysis linking probability of LMA to spreads, macroeconomic and fiscal fundamentals, investor sentiment, institutional characteristics, global factors, and interactions.
    - Advantages: decent in-sample and cross-country predictions
    - Disadvantages: takes time, data availability concerns, major data gaps, overfitting and/or poorer out-of-sample performance
  - Evaluate vulnerability to market access loss using the “signaling approach”
    - Advantages: can be done quickly, better out-of-sample performance compared to the multivariable probit, easy to understand and interpret, data gaps are not a constraint
    - Disadvantages: individual predictive variables cannot be tested for their conditional statistical significance

- Prior uses of the signaling approach cited:
  - Kaminsky, Lizondo, and Reinhart (1998) for currency crises
  - Hemming, Kell, and Schimmelpfennig (2003) for fiscal vulnerability indicators predicting financial crises
  - IMF (2007) for assessing the risk of sudden stops

### Operational mechanics of the signaling approach
- Method: use each potential indicator separately, identify critical thresholds that signal crisis events with the lowest prediction error.
- Two common methods to determine an optimal threshold:
  - Minimization of total misclassified errors (TME)
  - Maximization of the signal-to-noise ratio
- TME is computed from false negatives (FN), false positives (FP), total crises (Nc), and total non-crises (N). The optimal threshold X* is the value that minimizes TME(X).
- Note: Because crises are rare relative to non-crisis events, the TME methodology places greater weight on misclassifying crisis events, yielding relatively conservative thresholds.

### Practical considerations
- Signaling approach is well-suited when quick, interpretable out-of-sample detection is desired and when data gaps hinder multivariate modeling.
- Trade-off: inability to test conditional statistical significance of individual predictors.

---

### Risk Zone Classification — Example for Spreads

- Construction:
  - Build empirical distribution of the time series X. If X ~ F, then F(x) = Pr(X < x).
  - Empirical distribution function F̂(x) places weight 1/n at each data point xi: F̂(x) = 1/n ∑_{i=1}^n I(xi ≤ x).
  - The τ–th sample percentile Sτ corresponds to Sτ = F̂^{-1}(τ).

- Example percentiles and notation:
  - Compute percentiles at τ = {5th, 25th, 50th, 75th and 95th} and denote these as a, b, c, d and e respectively.
  - Risk zone classification scheme:
    - High risk: X ≥ e
    - Elevated risk: e > X ≥ d
    - IQR (inter-quartile range): d > X ≥ b
    - Low risk / Depressed risk: a > X ≥ b and X ≤ a (note: source denotes IQR64 as zone of ‘normal’ variation)

- Example: evolution of risk zones over time
  - Illustration described where the series entered a zone of elevated risk around February 2010 and a high risk zone by January 2011.

- Selecting an “optimal” high risk percentile
  - Choice of percentiles depends on historical data properties and tolerance levels (configuration for τ).
  - If tolerance for spread widening is lower than the 95th percentile, one would set S_small such that S_small < S.
  - Adaptation: exploit information on past LMA episodes across countries by recording the number of LMA episodes historically within each risk zone for different τ values across the cross-section.
  - Selecting τ to maximize episodes of LMA within the high risk zone will reduce episodes observed in the elevated risk zone.
  - Example statements from the framework:
    - S_high risk = F̂^{-1}(0.8).
    - F̂^{-1}(0.775) would correspond to a percentile such that no recorded LMA episodes fall in a zone of elevated risk, for instance.

---

### Extracting Information from the Term Structure of Distressed Bond Prices

- Pricing framework (market pricing to recovery):
  - Under recovery value R for one dollar of principal, market price P of a coupon-paying bond is given by a present-value expression that weights cash flows and recovery by survival probabilities and discounting using spot rates.
  - Key parameter λ is the unconditional default intensity per unit of time (hazard rate).
    - Probability of no default from time 0 until time t is e^{-λ t}.
    - Annualized probability of default is 1 − e^{-λ}, which is approximately λ when λ is not too large.
  - r is the risk free discount rate.

- Estimation approach:
  - Jointly estimate R and λ by minimizing the sum of squared fitting errors between P and observed market prices of outstanding bonds at a point in time.
  - Each period’s cash flow is discounted using appropriate points on the relevant spot curve.
  - Generate Svensson curves using observed US spot yields to obtain continuous risk-free discount curves; a new Svensson curve is generated for each date.
  - Weighted minimization: due to heteroscedastic bond prices, minimize a weighted sum weighted by the inverse of Macauley duration (strategy adopted in the estimations).

- Advantages and robustness:
  - Uses information from the entire term structure of outstanding bonds (advantage over CDS-based measures).
  - Estimated recovery rates are robust to different distributional assumptions on the survival function (e.g., Weibull, Gumbel, Merrick (2001), Vrught (2011)).
  - Robustness checks include holding λ fixed and experimenting with values between 0.5 and 0.9 (i.e., assuming more than a 50 percent chance of default).

---

### Restructuring Scenarios Conditional on the Market’s Expected Recovery Rate

- Concept:
  - Consider market-priced recovery rate as a lower bound R: bond holders will not agree to restructurings consistent with recovery rates lower than priced.
  - Under that conjecture, bond holders’ upper bound on haircut equals 1 − R.

- Haircut calculation formulas (as presented):
  - Definition (S S): H_{...} = 1 − NPV_{new}(S̈) / NPV_{old}(S̈)
  - Alternative earlier definition (b): H_{...} = 1 − NPV_{new}(S̈) / 100

- Important sensitivities:
  - Haircut calculations are sensitive to the choice of discount rate or “exit yield” S̈.
  - Results depend on whether haircut is calculated as (a) or (b).

- Deriving exit yields consistent with market bounds:
  - Let R̄ be the expected lower bound on recovery rate. Then implied upper bound on haircut is 1 − R̄.
  - For a given restructuring scenario, derive S̈ by minimizing the quadratic distance between the haircut implied by S̈ (using formula (a)) and the haircut bound 1 − R̄. That S̈ is consistent with the market’s upper bound for the haircut.

- Illustrative scenario:
  - Maturity extension on all bonds by 10 years, accompanied by a coupon reduction by 25 percent and principal reduction by 25 percent.
  - Exit yield estimates corresponding to haircut definitions (a) and (b) are provided in Charts A and B (source).
  - Observed behavior:
    - When using definition (b), implied exit yield declines linearly with higher recovery rates.
    - When using definition (a), the implied exit yield shows a relatively more non-linear decline.
  - Robustness: experiments with fixed λ values between 0.5 and 0.9 (more than 50 percent default chance) reported.

- Example recovery values mentioned in charts and notes:
  - Rec: 40%, Rec: 42%, Rec: 44%, Rec: 46%, Rec: 48%, Rec: 50% (Chart A)
  - Rec: 40%, Rec: 50%, Rec: 60%, Rec: 70% (Chart B)

---

### Heat Map (notes and interpretation)
- Heat map based on EMBI Global sovereign spreads.
  - Color coding: red = high risk, orange = elevated risk, yellow = medium risk, green = low risk.
- Critical thresholds are time-varying and depend on country-specific characteristics.
- Example country notes: Cote d’Ivoire is rated by Moody’s.
- Table/entries include Sovereign Dollar Bond Spreads (bps) and Sovereign Credit Ratings with sample values and averages for 2015 and 2016 and changes (Δ Jun 2014 - Dec 2015) for a cross-section of countries (Asia, Central and Eastern Europe, Middle East, Africa, Latin America) as presented in the source.
- Sources for heat map and table: Bloomberg, Dealogic, S&P Capital, Fitch, Moody's, and authors' calculations.

*Source: Annex III–VII, "wp17246 - Annex III. Signaling Approach for Identifying Vulnerability Indicators" (source PDF content as provided).*

### Annex VIII. Summary of IMF Lending Policies

### Annex VIII. Summary of IMF Lending Policies

### A. Exceptional Access — framework and criteria
- Exceptional access may be provided where a rigorous and systematic analysis indicates a high probability that the member’s public debt is sustainable in the medium term.
- If debt is considered sustainable but not with a high probability, exceptional access can be justified if financing from sources other than the Fund, although it may not restore sustainability with high probability, improves debt sustainability and sufficiently enhances the safeguards for Fund resources.
- Financing from sources other than the Fund may include, inter alia, financing obtained through any intended debt restructuring.
- Where the member’s debt is assessed to be unsustainable ex ante, exceptional access will only be made available where the financing being provided from sources other than the Fund restores debt sustainability with a high probability.
- Exceptional access is defined as 435 percent of access on a cumulative basis, 145 on an annual basis.

### A. Exceptional Access — recent reforms and decisions
- IMF Press Release No. 16/3, January 29, 2017: Executive Board approved reforms to the exceptional access lending framework to make it more calibrated to members’ debt situations and to avoid unnecessary costs for members, creditors, and the financial system.
- Key changes included:
  - Elimination of the “systemic exemption” introduced in 2010.
  - Increased flexibility for members where debt is assessed to be sustainable but not with high probability.
  - Clarification to the criterion related to market access.
- Background work and sequencing:
  - May 2013 Executive Board endorsement of a four-pronged work program (Public Information Notice No. 13/61).
  - Two components concluded earlier: (i) strengthening the contractual framework to address collective actions problems (Press Release No. 14/459); and (ii) reforming the IMF’s policy on the non-toleration of arrears to official creditors (Press Release No. 15/555).
  - Additional work on private sector involvement and lending-into-arrears policy to begin shortly.

### A. Exceptional Access — Executive Board assessment and guidance
- General support for reforms to allow responses better calibrated to a member’s debt vulnerabilities while enabling the Fund to provide financing to address balance of payments problems, including where spillover and contagion risks exist.
- Rationale for removing the systemic exemption:
  - Use of the systemic exemption can delay remedial measures and impair prospects for success and safeguards for Fund resources.
  - Replacement of maturing private claims with official claims (including Fund credit) can result in effective subordination of remaining private claims in a restructuring.
  - The systemic exemption aggravates moral hazard and may exacerbate market uncertainty in sovereign stress periods.
  - Invoking the systemic exemption may not address the underlying market concerns that drive contagion.

### A. Exceptional Access — application across debt situations
- If the Fund is confident debt is sustainable with high probability:
  - Continue providing financing in support of a strong adjustment program that envisages payment of outstanding obligations as they fall due (including cases of temporary loss of market access).
- If debt is clearly unsustainable:
  - Prompt and definitive debt restructuring to restore sustainability with high probability remains the least-cost approach.
- If debt is sustainable but not with high probability:
  - Requiring a definitive debt restructuring could incur unnecessary costs; exceptional access may be granted where additional financing from other sources improves debt sustainability and enhances safeguards for Fund resources.
  - There is a range of options; no presumption for any particular option — case-by-case justification required.
  - Options include:
    - Retained market access or small volumes of private claims falling due: maintain sufficient private exposure without restructuring.
    - Lost market access and significant private claims falling due: reprofiling of existing claims (short extension of maturities, normally no reduction in principal or coupons) typically appropriate to allow a less stringent adjustment path and reduce needed access.
      - Reprofiling is a form of debt restructuring but may be less costly than definitive restructuring.
      - Scope of debt to be reprofiled determined case-by-case; avoid reprofiling if costs (including risks to domestic financial stability) outweigh benefits.
      - Notably, short-term debt instruments (by original maturity), trade credits, and local currency-denominated debt had not been included in most past restructurings.
    - Financing from official bilateral creditors could be via extensions of maturities on existing claims and/or new financing commitments.

### A. Exceptional Access — timing and flexibility
- Ideally, a reprofiling should be undertaken before approval of the Fund arrangement; however, circumstances may warrant flexibility so that debt operation conclusion may occur later.
- It would not be necessary to hold up Fund support until complete clarity regarding the terms of the other financing exists.

### A. Exceptional Access — spillovers, tail risks, and official financing
- Fund staff and Directors recognized cross-border spillovers can occur from restructurings; some spillovers that reflect repricing of risk in line with fundamentals should be accommodated and complementary policy actions taken where fluctuations are not rooted in fundamentals.
- For rare tail-event cases where any restructuring of private claims, even a reprofiling, poses unmanageable risks to domestic financial stability or cross-border spillovers:
  - The reformed framework allows Fund approval of exceptional access without such a restructuring provided official sector partners supply necessary financing.
  - Such official financing must be on terms sufficiently favorable to improve sustainability and enhance safeguards for Fund resources, with assurances that terms could be modified in future if debt sustainability outlook deteriorates.
  - If official creditors are unwilling to provide such assurances, upfront terms would need to be sufficiently generous to restore debt sustainability with high probability.
  - In circumstances where debt is unsustainable, official bilateral financing terms would similarly need to be sufficiently favorable to restore debt sustainability with high probability (e.g., loans with long tenors and concessional rates, grants, or other instruments).
  - Requirements implemented flexibly; Fund could rely on political commitments to backstop debt sustainability without specifying all modalities.
  - Directors expect the approach to be used rarely and decisions should be made evenhandedly; some Directors saw potential use in less extreme circumstances.

### A. Exceptional Access — role of debt sustainability assessments and market-access criterion
- Debt sustainability assessment remains central and involves significant judgment; determination is forward-looking and considers:
  - Country-specific information on prospects for policy implementation.
  - Growth opportunities.
  - Contingent liabilities.
  - Nature of the creditor base.
  - Indicators of investor confidence.
  - Outlook for the global economic environment.
- Levels of debt consistent with sustainability can vary significantly across programs.
- Market access criterion:
  - Directors supported that the market access condition needs to be met even with open-ended official support beyond the program period.
  - Resolution of balance of payments problems and achievement of medium-term external viability is a key objective; ability (distinct from need) to access private capital markets is inherent to this resolution.
  - Staff should account, case-by-case, for positive impacts of official support commitments on a member’s ability to access markets when assessing the market access criterion.
  - Clarification on timeframe: Fund expects a member to gain or regain market access within a timeframe that facilitates repayment of all of its obligations to the Fund—not just the last one that is due.

### A. Exceptional Access — implementation
- Changes to the exceptional access framework enter into effect immediately and apply to all future completion of reviews under existing arrangements or approval of new Fund arrangements.
- Directors asked staff to continue work on the Fund’s lending toolkit for systemic crises and contagion and on debtor-creditor engagement issues, including the lending-into-arrears policy, to complete the program aimed at facilitating timely and orderly resolution of sovereign debt problems.

### B. Normal Access — policy statement
- The IMF needs to assess debt to be sustainable in the medium term to provide financing under normal access.
- To the extent debt is assessed to be unsustainable, the IMF is precluded from providing financing unless steps are taken by the member to restore sustainability.

*Annex VIII. Summary of IMF Lending Policies — wp17246*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2017/wp17246.pdf_