## EXECUTIVE SUMMARY

## Source details

**Canonical URL:** [EXECUTIVE SUMMARY](https://www.imf.org/-/media/files/publications/pp/2025/english/ppea2025001.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/pp/2025/english/ppea2025001.pdf.md)
- [Structured JSON version](/-/media/files/publications/pp/2025/english/ppea2025001.pdf.json)

---

### Purpose and context
- Provides background for an informal discussion with Executive Directors on the Comprehensive Review of GRA Access Limits.
- GRA access limits balance: (i) members’ confidence in availability of Fund financing; and (ii) preservation of liquidity and the revolving nature of Fund resources.
- Last comprehensive review: February 2016; time elapsed since then described as "nearly 9 years".
- Temporary increase in overall annual and cumulative access limits originally raised in March 2023 and extended in March 2024.
- Publication date: November 14, 2024.

### Key findings and analysis
- Notable erosion of access limits relative to macroeconomic aggregates since the 2016 review; the March 2023 temporary increase offset much of this erosion by around 40 percent.
- No strong evidence that increases in access limits have inadvertently overburdened members with senior debt or worsened their capacity to repay.
- Elevated uncertainty in a shock-prone environment is pertinent for calibrating access limits and risk tolerance in Fund lending.
- The impending effectiveness of the quota increase under the 16th General Review of Quotas (GRQ) adds urgency: absent adjustments, all limits would automatically increase by 50 percent in SDR terms when the quota increase becomes effective.
- The comprehensive review reassesses overall (annual and cumulative) GRA access limits, instrument-specific limits, and quota-based thresholds, taking into account erosion, changing global conditions, available Fund resources, and necessary safeguards.

### Staff recommendations (landing zone and policy stance)
- Set overall access limits at the temporarily increased levels:
  - Annual access limit (AAL): 200 percent of current quota.
  - Cumulative access limit (CAL): 600 percent of current quota.
- Other instrument-specific limits and thresholds are viewed as broadly appropriate.
- Continuation of existing access policy would:
  - Maintain stability and predictability.
  - Offset much of the erosion against macroeconomic aggregates since 2016.
  - Maintain alignment with PRGT access limits.

### Alternative option considered
- Modest further increase to more fully offset 2016–24 erosion:
  - Annual and cumulative access limits up to 225 and 675 percent of current quota, respectively.
- Drawbacks:
  - Would break alignment between GRA and PRGT overall access levels.
  - Would complicate implementation and communication.
- Any such increase should be judged against maintaining appropriate safeguards.

### Transitional proposals pending completion of the review
- To avoid disruptive swings in nominal access limits at end-2024, staff proposes:
  - Extend the current temporary access limits until the Executive Board completes the Comprehensive Review of GRA Access Limits (expected early 2025).
  - Maintain overall GRA access limits at 200 and 600 percent of quota until the review is completed.
- To manage the effect of the 16th GRQ quota increase and preserve nominal SDR safeguards:
  - Unless the Board decides otherwise, access limits and quota-based thresholds would mechanically increase by 50 percent in nominal terms upon effectiveness of the 16th GRQ quota increase.
  - Staff proposes dividing by 1.5 (that is, reducing by one-third) overall GRA access limits, instrument-specific limits, quota-based thresholds, and the RST quota-based limits when the general effectiveness conditions for the 16th GRQ quota increases are met.
- If Directors support these transitional measures, staff proposes to circulate illustrative decisions for the Board’s consideration and approval before end-2024 on a lapse-of-time basis.

### Operational and safeguard considerations
- Decisions require judgment across: erosion, evolving global conditions, available Fund resources, necessary safeguards, and elevated uncertainty in a shock-prone environment.
- Annual and cumulative overall limits are intended to trigger heightened scrutiny and safeguards under the exceptional access (EA) policy rather than impose a hard ceiling.
- Instrument-specific caps and safeguards remain important to align access with instrument objectives (e.g., modest caps for emergency financing instruments like the RFI; higher limits for precautionary instruments subject to stringent ex ante conditionality).

*Source: ppea2025001 - EXECUTIVE SUMMARY*

---

### 8. Access limits are reviewed periodically, with the last comprehensive review in February

### Overview of recent reviews and temporary modifications
- 2016 comprehensive review set AAL/CAL at 145/435 percent of current quota.
- Since 2016, several adjustments to overall and instrument-specific limits in response to economic challenges, notably COVID-19.
- 2023 temporary increase raised overall limits to 200/600 percent of quota, with the increase expiring at end-2024.
- Access under the SLL is currently capped at 200 percent of quota.
- RFI and LND window temporary increases during the pandemic reverted to pre-pandemic levels in June 2024:
  - Regular RFI: 50/100 percent of quota.
  - LND: 80/133.33 percent of quota, with additional 25 percent in cumulative limits if the Food Shock Window was used.

### Quota-based thresholds and procedural safeguards
- FCL liquidity impact assessment requirement: a staff note is required at the time of a request if proposed access exceeds 575 percent of quota or SDR 10 billion, whichever is lower.
- Post Financing Assessment (PFA): countries without a Fund-supported program or SMP that have outstanding credit above 200 percent of quota or above specified absolute amounts are expected to engage in PFA discussions.
- Article IV Consultation Cycles: countries with outstanding credit above 145 percent of the member’s quota may not be placed on an extended Article IV consultation cycle.

### Historical context and recent temporary changes (Box 1 highlights)
- 1980s: annual and cumulative access limits were gradually reduced in percent of quotas following the 7th and 8th GRQs.
- 1994: temporary increase in annual access limits following the 9th GRQ; retained after the 1999 11th GRQ.
- 1994–2008: AAL/CAL at 100 and 300 percent of quota, respectively.
- 2009 ad-hoc review (GFC): AAL to 200 and CAL to 600 percent of quota.
- 2016 comprehensive review: AAL/CAL at 145/435 percent of quota.
- COVID-19 response:
  - July 2020: AAL temporarily raised from 145 to 245 percent of quota (initially 12 months; extended to end-2021).
  - March 2023: AAL and CAL increased to 200 and 600 percent of quota (initially 12 months; extended in March 2024 to end-2024).

### Factors informing calibration of access limits
- Calibration balances erosion and global environment against Fund risk management and UFR-related trends.
- Erosion considerations:
  - Access limits set as a percentage of quota lead to erosion in real value as quotas change.
  - Access limits in nominal terms have eroded by about one-third when assessed against various macro metrics since the February 2016 review.
  - Table 1 (AAL/CAL to Offset 2016–24 Erosion) includes metric-based median/aggregate results such as 223 / 668, 217 / 652, 239 / 718, 234 / 702, 227 / 682, 201 / 603, 204 / 612, 223 / 669, with median results reported as 225 / 675 and 220 / 661 (depending on methodology). For EMDEs, examples include 224 / 672, 230 / 691, 240 / 721, 247 / 741, 233 / 699, 214 / 642, 199 / 598, 207 / 620, with median results 229 / 686 and 222 / 667.
- Changing global economic environment:
  - Medium-term global growth prospects are the weakest in decades and global economic policy uncertainty remains notably elevated.
  - Staff finds no clear evidence that increases in access limits have overburdened members with senior or non-restructurable debt or worsened their capacity to repay.

### Use and availability of GRA resources; Fund liquidity and credit developments
- Demand and arrangements:
  - Demand for GRA resources surged during COVID-19 and shifted toward GRA arrangements supporting UCT programs, particularly EFF arrangements.
  - As of 2024:Q3, there were 38 active GRA arrangements.
  - Median access to GRA resources has been around 250 percent of quota for normal access arrangements in recent years, equivalent to approximately SDR 1.1 billion.
- GRA credit outstanding and lending capacity:
  - Nominal GRA credit outstanding hovered around SDR 90 billion since end-2022.
  - As of 2024:Q3, 52 members had GRA credit outstanding; two members were under current EA arrangements; six members had less than 200 percent of their quota available for further financing without surpassing the current CAL.
  - Without the temporary increase during 2023–24, four members would have been subject to the EA policy under their current arrangements.
  - Credit outstanding relative to the Fund’s lending capacity was 11½ percent at end-September 2024.
  - Total commitments: SDR 199 bn in March 2022; SDR 215 bn in 2012; fell to SDR 162.6 billion at end-September 2024; commitments remained about 23½ percent of the Fund’s lending capacity since the pandemic (compared to 50–60 percent in recent decades and about 80 percent in 1998).
- Capacity to repay and borrower outcomes:
  - Total GRA credit outstanding as a share of external debt has remained relatively stable since the GFC and trended downward as a share of GDP.
  - Repayment obligations increased slightly but remain below historical peaks and are expected to decline with falling credit outstanding.
  - An event study suggests borrowers’ capacity to repay in normal access programs did not experience major changes following the 2016 adjustments.
  - Since the pandemic, capacity to repay of members with large exposures has shown some deterioration, primarily among EA arrangements.
- 16th GRQ implications:
  - The 16th GRQ will not increase the Fund’s lending capacity but will increase the share of quota-based resources, replacing Bilateral Borrowing Agreements and rolling back NAB credit arrangements.
  - Improved quality of resources—more quickly mobilized and predictable—will enhance the Fund’s agility.

### Key statistics and exact figures (selected)
- AAL/CAL from 2016 comprehensive review: 145/435 percent of current quota.
- Temporary increases: 2023 raised limits to 200/600 percent of quota; expiration at end-2024.
- SLL cap: 200 percent of quota.
- FCL liquidity-impact assessment threshold: 575 percent of quota or SDR 10 billion, whichever is lower.
- PFA threshold for outstanding credit: 200 percent of quota.
- Active GRA arrangements as of 2024:Q3: 38.
- Median access for normal access arrangements: around 250 percent of quota ≈ SDR 1.1 billion.
- Nominal GRA credit outstanding since end-2022: around SDR 90 billion.
- Number of members with GRA credit outstanding as of 2024:Q3: 52.
- GRA credit outstanding as percent of Fund lending capacity at end-September 2024: 11½ percent.
- Total commitments: SDR 199 bn (March 2022); SDR 215 bn (2012); SDR 162.6 billion (end-September 2024).
- Commitments as share of Fund’s lending capacity since the pandemic: about 23½ percent; historical comparisons: 50–60 percent in recent decades; about 80 percent in 1998.

*Source: ppea2025001 - 8.      Access limits are reviewed periodically, with the last comprehensive review in February*

---

### 18. A decision on access limits involves judgment, considering the lending capacity of the Fund, erosion, evolving global conditions, available Fund resources, and necessary safeguards

### Staff proposal to make temporary increases permanent and maintain instrument-specific limits
- Staff recommends making permanent the current temporarily higher overall GRA access limits and keeping instrument-specific limits unchanged.
- Rationale: maintains stability, predictability, and alignment with the PRGT overall access limits; balances addressing erosion with maintaining appropriate risk mitigation safeguards.
- Key elements:
  - Maintain GRA AAL/CAL at 200/600 in percent of current quota (a 38 percent increase relative to 145/435 set in February 2016).
  - Maintain SLL and PLL access limits, and the threshold of the FCL without articulation of exit expectation, at levels established during the 2023 Review of the FCL, SLL, and PLL.
  - Maintain RFI access limits, with the 2019 one-third increase broadly offsetting estimated erosion during 2016–24.

### Alternative: raise overall GRA access limits further
- Directors could consider raising overall GRA access limits up to 225/675 in percent of current quota (up to a 12.5 percent increase from current 200/600 levels).
- Trade-offs:
  - Further offsets erosion since 2016.
  - Would result in GRA and PRGT overall access levels no longer being aligned and complicate implementation and communication.
  - Increasing RFI access limits would create a wedge with the Rapid Credit Facility (RCF).

### Transitional arrangements and extending temporary limits
- Comprehensive review expected early 2025; temporary increase expires end-2024 absent Board action.
  - Absent action, overall access limits would fall back to 145/435 on January 1, 2025, before likely being raised at completion of the comprehensive review.
- Staff proposes extending the current temporary access limits until the comprehensive review is completed.
  - No specific calendar expiration date; limits remain in effect until Board completes the review.
  - Staff proposes issuing an illustrative Executive Board decision for consideration and approval on a lapse-of-time basis in advance of December 31, 2024.

### Implications of the 16th GRQ for access limits
- Staff proposes dividing by 1.5 all limits and thresholds (defined in percent of current quota) discussed in this paper when the quota increase under the 16th GRQ becomes effective.
  - This maintains unchanged in nominal (SDR) terms the value of the access limits and other quota-related limits and thresholds.
  - Avoids an uncalibrated relaxation of lending safeguards that would otherwise occur because access limits and thresholds would increase by 50 percent in SDR terms with the 16th GRQ.
- Specific proposal for rounding:
  - Staff would propose rounding up current overall GRA annual and cumulative access limits for simplicity to 135 and 405 percent of (16th GRQ) quotas, respectively, to align with overall PRGT access limits following the 16th GRQ.

### Safeguards-related thresholds (exact proposals preserved)
- Re-align requirement for the FCL liquidity impact assessment with the CAL while keeping nominal threshold of SDR 10 billion:
  - Current trigger: access under the arrangement exceeds the lower of 575 percent of quota or SDR 10 billion.
  - Proposal: SDR 10 billion or 600 percent.
  - Staff notes erosion considerations would suggest raising SDR 10 billion to around SDR 15 billion, but proposes to keep SDR 10 billion.
- Maintain other safeguards-related thresholds:
  - PFA trigger remains SDR 1.5 billion or 200 percent of quota.
  - Quota-based threshold for extended Article IV cycle:
    - Current: 145 percent of quota.
    - Proposal: round up to 150 percent of current quota (which would translate to 100 percent of the 16th GRQ quota).
- Table 2 summary (figures preserved):
  - Commitment fee thresholds (lower/upper thresholds): Current 200/600; Proposal 200/600.
  - Requirement for the FCL liquidity impact assessment (CAL): Current lower of SDR 10 billion or 575 percent; Proposal SDR 10 billion or 600 percent.
  - PFA—Outstanding credit trigger: Current SDR 1.5 billion or 200 percent; Proposal SDR 1.5 billion or 200 percent.
  - Extended Article IV cycle—outstanding Fund credit does not exceed: Current 145; Proposal 150.

### Resource implications and enterprise risks
- Impact on Fund resources:
  - Experience with temporary increase indicates limited effect on demand for Fund resources.
  - Effect on demand during temporary 200/600 limits maintained since March 2023 had been limited; few arrangements utilized higher access limits.
  - Since then, one new arrangement and two augmentations approved with access within normal access limits that would have exceeded previously established lower limits (145/435).
- Enterprise risk considerations:
  - Risks without Board decisions:
    - If temporary increase expires at end-2024, limits fall to 145/435, resulting in de facto tightening of IMF’s risk tolerance; could lead to higher incidence of EA requests and weaken Fund’s ability to respond swiftly; later increase would add volatility (“seesaw”).
    - If no action on the 16th GRQ, access limits and thresholds would see a 50 percent nominal increase, resulting in de facto sizable relaxation of Fund’s risk tolerance.
  - Risks to member engagement: undermined predictability and potential loss of trust.
  - Strategic and reputational risks: constrained response ability could trigger member disengagement, increased reliance on self-insurance, excessive or disorderly adjustment, and criticism.
  - Risks to adequacy and liquidity of lending resources:
    - Access limits expected to have limited impact on Fund liquidity; estimated effects on demand compatible with staff’s risk tolerance benchmarks but subject to uncertainty and warrant close monitoring.
  - Credit risk:
    - Generally, credit risks remain broadly unchanged under staff’s proposal; higher access limits raise thresholds for triggering enhanced scrutiny which could potentially raise debt sustainability risks for some countries.
    - Access decisions continue to be determined by rigorous assessments informed by standard access policy criteria (¶6), including BoP need, strength of program policies, country’s record, debt sustainability, and capacity to repay the Fund.
    - Staff’s analysis suggests the 2016 review increase did not have a statistically significant impact on overall CtR of borrowers (Annex III).
  - Operational risks:
    - Operational risks from adopting and applying policy changes are limited given accumulated experience; early action would mitigate operational risks.

### Issues for discussion
- Directors may wish to comment on:
  - Whether staff’s proposal to make permanent current access limits and thresholds strikes a good balance between addressing erosion and maintaining appropriate risk mitigation safeguards.
  - Whether to extend the current temporary overall GRA access limits until the Executive Board completes the Comprehensive Review of GRA Access Limits.

*Source: ppea2025001 - 18.*

---

### Annex I — Measuring erosion of access limits (summary) and Table 1 (methodologies)

### Erosion definition and dimensions
- Erosion defined as decrease in the ratio of nominal access limits (in SDR terms) to relevant nominal macroeconomic variables (also in SDR).
- Estimation requires multidimensional analysis across:
  - Macroeconomic variables in the denominator (e.g., GDP, measures of cross-border flows).
  - Country coverage (world, EMDEs, or past users of Fund lending).
  - Aggregation method (median across countries vs. aggregate for sample).
  - Time horizon (start point for erosion calculations).

### Staff preference and methodology for the Review
- Use same metrics as 2023/24 interim reviews; country coverage comparable to 2016 review; median erosion proposed as preferred measure of central tendency.
- Advantages of median erosion: reflects erosion experienced by representative member; avoids bias toward large economies inherent in aggregate erosion.
- For continuity and comparability, staff complements median-based estimates with aggregate erosion.
- End point for analyzing erosion updated to 2024 to incorporate October 2024 WEO forecasts.

### Annex I. Table 1 — Time periods and concepts (selected)
- Time periods compared:
  - 2016 Review: 2009–2015
  - 2023/24 Interim Reviews: 2015–2023
  - 2025 Review: 2016–2024
- Economic concepts and metrics used as proxies:
  - Capacity to repay: GDP (all three reviews)
  - Current account / Current payments:
    - 2016 Review: Trade = Exports + Imports
    - 2023/24 Interim Reviews and 2025 Review: Current payments = Imports of goods and services + Primary income debit + Secondary income debit
  - Capital account / Capital flows:
    - 2016 Review: Non-FDI External Liabilities
    - 2023/24 Interim Reviews and 2025 Review: Capital flows = FDI + Portfolio liabilities + Other investment liabilities
  - Aggregate BoP needs / External financing needs (EFN):
    - 2016 Review: EFN = – Current account balance + Amortization falling due in the next 12 months
    - 2023/24 Interim Reviews and 2025 Review: EFN = – Current account balance + Amortization – Change in reserves (Change in reserves subtracted only for low-reserves countries: reserves covering less than 100 percent of short-term external debt on remaining maturity basis)

### Country coverage and proxies for likely borrowers
- 2016 Review: Broadest coverage.
- 2023/24 Interim Reviews: All IMF members (World) with restrictions: for capital flows and EFN only countries with positive values; additionally for EFN: program countries (Program countries are countries that had at least one purchase under a GRA arrangement since 1990).
- 2025 Review: All IMF members (World) with restrictions for capital flows and EFN: only countries with positive values; Program countries.
- Proxy for likely borrowers (EMDE focus and exclusions) varied across reviews (see source for exact specifications).

### Central tendency measures
- 2016 Review: Erosion of aggregate for country sample.
- 2023/24 Interim Reviews: Erosion of aggregate; preferred measure: Median erosion within country sample.
- 2025 Review: Median erosion within country sample (preferred) and erosion of aggregate for comparability.

*Italic footnotes from source preserved verbatim where present.*

*Source: ppea2025001 - Annex I. Table 1. Erosion Estimation Methodologies Across Recent Reviews (selected tables and annexes excerpt).*

---

### Annex II — Analysis of Changes in Credit Risks Since the 2016 Comprehensive Review of Access

### Key findings and context
- Measures of credit risk in the Fund’s lending portfolio generally point to some weakening since 2016, mitigated by an increase in financial buffers, limiting the impact on residual credit risk.
- Drivers of weaker credit risk indicators:
  - More challenging economic and financial landscape following global shocks.
  - Significant expansion in total commitments during 2016-23.
  - Expansion in actual and projected total credit outstanding and tripled exposure to the largest borrower.
  - Commitments under precautionary arrangements in 2023 were substantially lower than in 2016.
  - Sovereign credit ratings signal rising stress for some borrowers, particularly the large ones.
  - Share of lending to countries rated CCC or below increased substantially.
- Mitigating factors:
  - Precautionary balances reached the current medium-term target of SDR 25 billio  n by end-FY2024; burden sharing capacity has rapidly expanded mainly on the back of the higher SDR interest rate and the absence of arrears cases recently.

### Annex II. Table 1 — Selected credit risk indicators (In billions of SDRs, end December 2023 unless otherwise indicated)
- Risks (2016 / 2023 / Change where shown in table):
  - Total commitments: 159.0 / 174.0
  - Commitments under precautionary arrangements 1/: 89.0 / 58.0
    - of which FCL and PLL arrangements: 86.1 / 56.6
  - Credit outstanding:
    - Actual: 49.3 / 92.1
    - Projected peak 2/: 49.3 / 97.6
  - Largest individual exposure:
    - Actual: 10.1 / 30.4
    - Projected peak 2/: 22.9 / 34.2
  - Credit concentration:
    - Top 5 (in percent of total): 80.2 / 68.0
    - Top 1 (in percent of total): 26.0 / 33.0
  - Share of largest regional exposure in total commitments (percent): 45.9 / 58.7
  - Share of largest regional exposure in total credit outstanding (percent): 55.8 / 48.0
  - Share of RFI in the credit portfolio (percent): 2.4 / 15.4
  - Weighted sovereign credit rating of Fund credit exposures (S&P) 3/: 12.5 / 15.7
  - Weighted sovereign spreads of largest five borrowers (basis points): 480 / 1962
  - Share of Fund credit to members rated CCC or lower (percent): 0.05 / 2.9
  - Arrears (SDR millions): 254.1 / 0.0
- Buffers:
  - Precautionary balances 4/: 15.2 / 24.4
  - Precautionary balances (in percent of):
    - Credit outstanding: 31.8 / 26.5
    - Total commitments: 12.1 / 14.0
    - Lending capacity: 2.2 / 3.5
  - Burden sharing capacity (SDR millions): 30.5 / 1329.5
- Other:
  - Lending capacity 4/: 684.5 / 695.6
- Notes from table preserved verbatim:
  - 1/ Undrawn balances under arrangements treated as precautionary by the authorities.
  - 2/ Data for 2016 are actual and not projected.
  - 3/ Sovereign credit ratings are assigned numerical scores. The weaker the rating the higher the score. A higher weighted average rating of Fund exposures reflects a deterioration in ratings.
  - 4/ Data for 2016 as at end-Fiscal Year

*Source: ppea2025001 - Annex II.*

---

### Annex III — Capacity to Repay (CtR) and Access Limits

### A. CtR trends and interpretation
- Borrowers’ capacity to meet debt obligations to the Fund has remained broadly stable over the past 15 years, despite upticks among debt service indicators in recent years.
- CtR indicators worsened recently among large Fund borrowers, including in the context of Exceptional Access programs.
- An event study around the 2016 increase (45 percent on average in SDR terms) finds no statistically significant differences in the distribution of CtR indicators for IMF-supported normal access programs approved before and after the 2016 increase.

### B. Empirical findings on CtR indicators
- Stock indicators:
  - For most countries, credit outstanding as a share of GDP, reserves, and external debt has remained low and stable.
  - Notable increase in the top deciles of the distribution, indicating disproportionate deterioration for certain members, especially large borrowers.
  - Normal access borrowers’ CtR remained relatively stable, with recent improvement in average and median CtR stock indicators.
- Flow indicators:
  - Obligations to the Fund relative to exports, fiscal revenues, and external debt service show a slight upward shift in overall distribution.
  - Deterioration primarily driven by borrowers with exceptional access; some deterioration also observed for normal access countries.

### C. Event study of the 2016 access limit increase (design and results)
- Design:
  - 2016 increase treated as exogenous (triggered by 14th GRQ implementation).
  - Compares CtR indicators (projected at time of approval) for normal access programs approved three years before and three years after the 2016 increase, tracking indicators over five years post-approval.
  - Sensitivity check: results robust when narrowing the window to 18 months before and after the 2016 increase.
- Results:
  - Distributions of CtR indicators for programs approved before and after the 2016 increase show considerable overlap.
  - Expansion in the upper deciles of distributions for programs approved after the 2016 increase, concentrated among large borrowers.
  - For stock variables, widening of upper distributions observed throughout the five years following program approval.
  - For flow indicators, widening becomes apparent from the third year after program approval.
  - Statistical tests (Two-sample t-test and Mann-Whitney U-test) indicate no statistically significant differences at the 95 percent confidence level in mean or distribution of CtR indicators between programs approved before and after the 2016 increase.
- Conclusion: Despite deterioration in upper deciles for some indicators, mean and overall distribution comparisons suggest a limited impact of the 2016 access limit increase on borrowers’ capacity to repay.

### Figures and statistical presentation (selected)
- Figures show interquartile ranges (25th and 75th percentiles) with medians indicated by dashed lines for indicators including:
  - Credit Outstanding as Percent of GDP
  - Credit Outstanding as Percent of Reserves
  - Credit Outstanding as Percent of Public External Debt
  - Obligations to the Fund as Percent of Exports of Goods and Services
  - Obligations to the Fund as Percent of Fiscal Revenues
  - Obligations to the Fund as Percent of Public External Debt Service
- Statistical tests presented with 95 percent confidence intervals for differences in means across time horizons t to t+5.

### Key textual finding
- "Overall, drawing from the 2016 experience, making permanent the current temporary access limits would be expected to have a limited effect on the overall distribution of IMF borrowers’ capacity to repay."
- Decisions on the level of access of an individual member should continue to be guided by:
  - the member’s actual, prospective, or potential balance of payments needs,
  - the strength of its economic reform program,
  - the EA framework when applicable,
  - the amount of its outstanding use of Fund resources,
  - its capacity to repay the Fund.

*Source: ppea2025001 - Annex III. Figure 3. Interquartile range of the projected member’s Capacity to Repay by*

### EXECUTIVE SUMMARY

### EXECUTIVE SUMMARY

### Purpose and context
- Provides background for an informal discussion with Executive Directors on the Comprehensive Review of GRA Access Limits.
- GRA access limits are part of the Fund’s risk management framework balancing: (i) members’ confidence in availability of Fund financing; and (ii) preservation of liquidity and the revolving nature of Fund resources.
- Last comprehensive review: February 2016. Time elapsed since then described as "nearly 9 years".
- Temporary increase in overall annual and cumulative access limits originally raised in March 2023 and extended in March 2024.
- Publication date: November 14, 2024.

### Key findings and analysis
- There has been a notable erosion of access limits relative to macroeconomic aggregates since the 2016 review; the March 2023 temporary increase offset much of this erosion by around 40 percent.
- Staff finds that there is no strong evidence that increases in access limits have inadvertently overburdened members with senior debt or worsened their capacity to repay.
- Elevated uncertainty in a shock-prone environment is pertinent for calibrating access limits and risk tolerance in Fund lending.
- The impending effectiveness of the quota increase under the 16th General Review of Quotas (GRQ) adds urgency: absent adjustments, all limits would automatically increase by 50 percent in SDR terms when the quota increase becomes effective.
- The comprehensive review aims to reassess overall (annual and cumulative) GRA access limits, instrument-specific limits, and quota-based thresholds, taking into account erosion, changing global conditions, available Fund resources, and necessary safeguards.

### Staff recommendations (landing zone and policy stance)
- Staff considers that setting overall access limits at the temporarily increased levels would strike an appropriate balance:
  - Annual access limit (AAL): 200 percent of current quota.
  - Cumulative access limit (CAL): 600 percent of current quota.
- Other instrument-specific limits and thresholds are viewed as broadly appropriate.
- Continuation of existing access policy would:
  - Maintain stability and predictability.
  - Offset much of the erosion against macroeconomic aggregates since 2016.
  - Maintain alignment with PRGT access limits.

### Alternative option considered
- Modest further increase to more fully offset 2016–24 erosion:
  - Annual and cumulative access limits up to 225 and 675 percent of current quota, respectively.
- Drawbacks of the alternative:
  - Would break alignment between GRA and PRGT overall access levels.
  - Would complicate implementation and communication.
- Any such increase should be judged against maintaining appropriate safeguards.

### Transitional proposals pending completion of the review
- To avoid disruptive swings in nominal access limits at end-2024, staff proposes:
  - Extend the current temporary access limits until the Executive Board completes the Comprehensive Review of GRA Access Limits (expected early 2025).
  - Maintain overall GRA access limits at 200 and 600 percent of quota until the review is completed.
- To manage the effect of the 16th GRQ quota increase and preserve nominal SDR safeguards:
  - Unless the Board decides otherwise, access limits and quota-based thresholds would mechanically increase by 50 percent in nominal terms upon effectiveness of the 16th GRQ quota increase.
  - Staff proposes that overall GRA access limits, instrument-specific limits, quota-based thresholds, and the RST quota-based limits be divided by 1.5 (that is, reduced by one-third) when the general effectiveness conditions for the 16th GRQ quota increases are met—consistent with approaches used in the recent Review of Charges and the Surcharge Policy and the Review of the PRGT Facilities and Financing.
- If Directors support these transitional measures, staff proposes to circulate illustrative decisions for the Board’s consideration and approval before end-2024 on a lapse-of-time basis.

### Operational and safeguard considerations
- Decisions on access limits require judgment across multiple considerations: erosion, evolving global conditions, available Fund resources, necessary safeguards, and elevated uncertainty in a shock-prone environment.
- Annual and cumulative overall limits are intended to trigger heightened scrutiny and safeguards under the exceptional access (EA) policy rather than impose a hard ceiling.
- Instrument-specific caps and safeguards remain important to align access with instrument objectives (e.g., modest caps for emergency financing instruments like the RFI; higher limits for precautionary instruments subject to stringent ex ante conditionality).

*Source: ppea2025001 - EXECUTIVE SUMMARY*

### 8.      Access limits are reviewed periodically, with the last comprehensive review in February

### 8.      Access limits are reviewed periodically, with the last comprehensive review in February

### Overview of recent reviews and temporary modifications
- The most recent comprehensive review took place in February 2016 and set the AAL/CAL at 145/435 percent of current quota.
- Since 2016, there have been several adjustments to both overall and instrument-specific limits in response to economic challenges, notably COVID-19.
- The 2023 temporary increase raised overall limits to 200/600 percent of quota, with the increase expiring at end-2024.
- Access under the SLL is currently capped at 200 percent of quota.
- The Rapid Financing Instrument (RFI) and its Large Natural Disaster (LND) window experienced several temporary increases during the pandemic and reverted to pre-pandemic levels in June 2024 (regular: 50/100; LND: 80/133.33 percent of quota, with additional 25 percent in cumulative limits if the Food Shock Window was used).

### Quota-based thresholds and procedural safeguards
- Assessment requirement for FCL liquidity impact: a staff note is required at the time of a request if proposed access exceeds 575 percent of quota or SDR 10 billion, whichever is lower.
- Post Financing Assessment (PFA): countries without a Fund-supported program or Staff Monitored Program (SMP) that have outstanding credit from the GRA, the PRGT, or the RST above 200 percent of quota or above specified absolute amounts are expected to engage in PFA discussions.
- Article IV Consultation Cycles: countries with outstanding credit to the Fund above 145 percent of the member’s quota may not be placed on an extended Article IV consultation cycle.

### Box 1 — Historical context and recent temporary changes
- 1980s: annual and cumulative access limits were gradually reduced in percent of quotas following the 7th and 8th GRQs to prevent sharp SDR increases.
- 1994: temporary increase in annual access limits following the 9th GRQ to correct erosion relative to GDP and trade; these temporary increases were retained after the 1999 11th GRQ.
- 1994–2008: annual and cumulative access limits remained at 100 and 300 percent of quota, respectively.
- 2009 ad-hoc review: access limits were doubled in percent of quota amid the Global Financial Crisis (GFC), raising AAL to 200 and CAL to 600 percent of quota; these levels were sustained until the 14th GRQ effectiveness in 2016.
- 2016 comprehensive review: access raised by 45 percent in SDR terms on average, resulting in AAL/CAL at 145/435 percent of quota (shorthand “145/435”).
- COVID-19 response:
  - July 2020: AAL temporarily raised from 145 to 245 percent of quota (initially for 12 months; extended to end-2021). No change to the cumulative GRA access limit at that time.
  - March 2023: AAL and CAL increased to 200 and 600 percent of quota, respectively (initially for 12 months; extended in March 2024 to end-2024).

### Factors informing the calibration of access limits
- Calibration balances erosion of access limits and implications of the global economic environment against the Fund’s risk management role and UFR-related trends.
- Erosion considerations:
  - Access limits set as a percentage of quota lead to erosion in real value as quotas change, effectively tightening the EA framework by triggering it at lower levels of Fund exposure relative to macro aggregates.
  - Access limits in nominal terms have eroded by about one-third when assessed against various global and country-group macro metrics since the February 2016 review.
  - Table 1 (AAL/CAL to Offset 2016–24 Erosion, In percent of current quota) lists multiple metric-based estimates; medians and aggregates include values such as 223 / 668, 217 / 652, 239 / 718, 234 / 702, 227 / 682, 201 / 603, 204 / 612, 223 / 669, with median results reported as 225 / 675 and 220 / 661 (depending on methodology). For EMDEs, examples include 224 / 672, 230 / 691, 240 / 721, 247 / 741, 233 / 699, 214 / 642, 199 / 598, 207 / 620, with median results 229 / 686 and 222 / 667.
- Changing global economic environment:
  - Medium-term global growth prospects are the weakest in decades and global economic policy uncertainty remains notably elevated.
  - Staff finds no clear evidence that increases in access limits have overburdened members with senior or non-restructurable debt or worsened their capacity to repay.

### Use and availability of GRA resources; Fund liquidity and credit developments
- Demand and arrangements:
  - Demand for GRA resources surged during COVID-19 and shifted toward GRA arrangements supporting UCT programs, particularly EFF arrangements.
  - As of 2024:Q3, there were 38 active GRA arrangements, comparable to the peak in 2010.
  - Median access to GRA resources has been around 250 percent of quota for normal access arrangements in recent years, equivalent to approximately SDR 1.1 billion.
- GRA credit outstanding and lending capacity:
  - Nominal GRA credit outstanding hovered around SDR 90 billion since end-2022.
  - As of 2024:Q3, 52 members had GRA credit outstanding; two members were under current EA arrangements; six members had less than 200 percent of their quota available for further financing without surpassing the current CAL.
  - Without the temporary increase during 2023–24, four members would have been subject to the EA policy under their current arrangements.
  - Credit outstanding relative to the Fund’s lending capacity was 11½ percent at end-September 2024, well below the 1998 peak (55 percent) and the GFC (19 percent), owing to expansion of the Fund’s borrowed resources since the GFC.
  - Total commitments: SDR 199 bn in March 2022 vs. SDR 215 bn in 2012; fell to SDR 162.6 billion at end-September 2024; remained moderate at about 23½ percent of the Fund’s lending capacity since the pandemic (compared to 50–60 percent in recent decades and about 80 percent in 1998).
- Capacity to repay and borrower outcomes:
  - Total GRA credit outstanding as a share of external debt has remained relatively stable since the GFC and trended downward as a share of GDP.
  - Repayment obligations to the Fund increased slightly but remain below historical peaks and are expected to decline with falling credit outstanding.
  - An event study suggests borrowers’ capacity to repay in normal access programs did not experience major changes following the 2016 adjustments to access limits.
  - Since the pandemic, capacity to repay of members with large exposures has shown some deterioration, primarily among EA arrangements.
- 16th GRQ implications:
  - The 16th GRQ will not increase the Fund’s lending capacity but will increase the share of quota-based resources, replacing Bilateral Borrowing Agreements and rolling back NAB credit arrangements.
  - The improved quality of resources—more quickly mobilized and predictable—will enhance the Fund’s agility in responding to higher demand for financing.

### Key statistics and exact figures (selected)
- AAL/CAL from 2016 comprehensive review: 145/435 percent of current quota.
- Temporary increases: 2023 raised limits to 200/600 percent of quota; expiration at end-2024.
- SLL cap: 200 percent of quota.
- FCL liquidity-impact assessment threshold: 575 percent of quota or SDR 10 billion, whichever is lower.
- PFA threshold for outstanding credit: 200 percent of quota (or above specified absolute amounts).
- Active GRA arrangements as of 2024:Q3: 38.
- Median access for normal access arrangements: around 250 percent of quota ≈ SDR 1.1 billion.
- Nominal GRA credit outstanding since end-2022: around SDR 90 billion.
- Number of members with GRA credit outstanding as of 2024:Q3: 52.
- GRA credit outstanding as percent of Fund lending capacity at end-September 2024: 11½ percent.
- Total commitments: SDR 199 bn (March 2022); SDR 215 bn (2012); SDR 162.6 billion (end-September 2024).
- Commitments as share of Fund’s lending capacity since the pandemic: about 23½ percent; historical comparisons: 50–60 percent in recent decades; about 80 percent in 1998.

*Source: ppea2025001 - 8.      Access limits are reviewed periodically, with the last comprehensive review in February*

### 18.      A decision on access limits involves judgment, considering the lending capacity of the

### A decision on access limits involves judgment, considering the lending capacity of the Fund, erosion, evolving global conditions, available Fund resources, and the necessary safeguards to protect the Fund’s resources, among others.

### Staff proposal to make temporary increases permanent and maintain instrument-specific limits
- Staff sees merit in making permanent the current temporarily higher overall GRA access limits and keeping unchanged the instrument-specific limits.
- Rationale: continuation of existing access limits maintains stability, predictability, and alignment with the PRGT overall access limits; balances addressing erosion with maintaining appropriate risk mitigation safeguards.
- Key elements:
  - Maintain GRA AAL/CAL at 200/600 in percent of current quota.
    - This is a 38 percent increase relative to the 145/435 set in the February 2016 review.
  - Maintain SLL and PLL access limits, and the threshold of the FCL without articulation of exit expectation, at levels established during the 2023 Review of the FCL, SLL, and PLL.
  - Maintain RFI access limits, considering the 2019 adjustment (one-third increase) broadly offset estimated erosion during 2016–24.

### Alternative: raise overall GRA access limits further
- Directors could consider raising overall GRA access limits up to 225/675 in percent of current quota.
  - This would be up to a 12.5 percent increase from current 200/600 levels.
- Trade-offs:
  - Would further offset erosion since 2016.
  - Would result in GRA and PRGT overall access levels no longer being aligned and would complicate implementation and communication.
  - RFI access limits could be increased slightly, but that would create a wedge with the Rapid Credit Facility (RCF).

### Transitional arrangements and extending temporary limits
- Comprehensive review expected to be completed in early 2025; temporary increase in overall GRA access limits expire at end-2024 absent Board action.
  - Absent action, overall access limits would fall back to 145/435 on January 1, 2025, before likely being raised at completion of the comprehensive review.
  - Such seesawing in overall access limits would undermine stability and predictability of Fund lending policies.
- Staff proposes extending the current temporary access limits until the comprehensive review is completed.
  - No specific calendar expiration date; limits remain in effect until Board completes the ongoing comprehensive review.
  - Staff proposes issuing an illustrative Executive Board decision for consideration and approval on a lapse of time basis in advance of December 31, 2024.

### Implications of the 16th GRQ for access limits
- Staff proposes dividing by 1.5 all limits and thresholds (defined in percent of current quota) discussed in this paper when the quota increase under the 16th GRQ becomes effective.
  - This approach maintains unchanged in nominal (SDR) terms the value of the access limits and other quota-related limits and thresholds.
  - It avoids an uncalibrated relaxation of lending safeguards that would otherwise occur because access limits and thresholds would increase by 50 percent in SDR terms with the 16th GRQ.
- Specific proposal for rounding:
  - Staff would propose rounding up current overall GRA annual and cumulative access limits for simplicity to 135 and 405 percent of (16th GRQ) quotas, respectively, to align with overall PRGT access limits following the 16th GRQ.

### Safeguards-related thresholds
- Proposal to re-align requirement for the FCL liquidity impact assessment (CAL) with the CAL while keeping unchanged the nominal threshold of SDR 10 billion.
  - Current trigger: access under the arrangement exceeds the lower of 575 percent of quota or SDR 10 billion.
  - Proposal: SDR 10 billion or 600 percent (align with proposed CAL); nominal threshold unchanged at SDR 10 billion (staff notes erosion considerations would suggest raising it to around SDR 15 billion, but proposes to keep SDR 10 billion).
- Maintain other safeguards-related thresholds at conservative levels:
  - Post Financing Assessment (PFA) trigger remains SDR 1.5 billion or 200 percent of quota.
  - Quota-based threshold of outstanding Fund credit above which a member may not be placed on an extended Article IV consultation cycle:
    - Current: 145 percent of quota.
    - Proposal: round up to 150 percent of current quota (which would translate to 100 percent of the 16th GRQ quota).
- Table 2 (summary of safeguards-related thresholds) — exact figures preserved:
  - Commitment fee thresholds (lower/upper thresholds): Current 200/600; Proposal 200/600.
  - Requirement for the FCL liquidity impact assessment (CAL): Current lower of SDR 10 billion or 575 percent; Proposal SDR 10 billion or 600 percent.
  - PFA—Outstanding credit trigger: Current SDR 1.5 billion or 200 percent; Proposal SDR 1.5 billion or 200 percent.
  - Extended Article IV cycle—outstanding Fund credit does not exceed: Current 145; Proposal 150.

### Resource implications and enterprise risks
- Impact on Fund resources
  - Experience with temporary increase indicates limited effect on demand for Fund resources.
  - Effect on demand during temporary 200/600 limits maintained since March 2023 had been limited; few arrangements utilized higher access limits.
  - Since then, one new arrangement and two augmentations approved with access within normal access limits that would have exceeded previously established lower limits (145/435).
- Enterprise risk considerations
  - Risks without Board decisions:
    - Absence of Board decisions on access limits: temporary increase would expire at end-2024, limits fall to 145/435, resulting in de facto tightening of IMF’s risk tolerance; could lead to higher incidence of EA requests and weaken Fund’s ability to respond swiftly; later increase would add volatility (“seesaw”).
    - Absence of action on effectiveness of the 16th GRQ: without adjustment to keep nominal SDR values unchanged, access limits and thresholds would see a 50 percent nominal increase, resulting in de facto sizable relaxation of Fund’s risk tolerance.
  - Risks to member engagement: undermined predictability and potential loss of trust in Fund’s ability to respond to financing needs.
  - Strategic and reputational risks: constrained ability to respond could trigger member disengagement, increased reliance on self-insurance, excessive or disorderly adjustment, and criticism regarding effectiveness.
  - Risks to adequacy and liquidity of lending resources:
    - Access limits expected to have limited impact on Fund liquidity; estimated effects on demand compatible with staff’s risk tolerance benchmarks but subject to uncertainty and warrant close monitoring.
  - Credit risk:
    - Generally, credit risks remain broadly unchanged under staff’s proposal; higher access limits raise thresholds for triggering enhanced scrutiny which could potentially raise debt sustainability risks for some countries.
    - Access decisions will continue to be determined by rigorous assessments informed by standard access policy criteria (¶6), including BoP need, strength of program policies, country’s record, debt sustainability, and capacity to repay the Fund.
    - Staff’s analysis suggests the 2016 review increase did not have a statistically significant impact on overall CtR of borrowers (Annex III).
  - Operational risks:
    - Operational risks from adopting and applying policy changes are limited given accumulated experience; early action would mitigate operational risks.

### Issues for discussion
- Directors may wish to comment on:
  - How Directors view staff’s proposal to make permanent current access limits and thresholds; whether it strikes a good balance between addressing erosion and maintaining appropriate risk mitigation safeguards.
  - Whether Directors agree to extend the current temporary overall GRA access limits until the Executive Board completes the Comprehensive Review of GRA Access Limits.

### Annex I — Measuring erosion of access limits (summary)
- Erosion definition: decrease in the ratio of nominal access limits (in SDR terms) to relevant nominal macroeconomic variables (also in SDR).
- Estimation requires multidimensional analysis across four dimensions:
  - Macroeconomic variables in the denominator (e.g., GDP, measures of cross-border flows).
  - Country coverage (world, EMDEs, or past users of Fund lending).
  - Aggregation method (median across countries vs. aggregate for sample).
  - Time horizon (start point for erosion calculations).
- Staff preference and methodology for the Review:
  - Use same metrics as 2023/24 interim reviews; country coverage comparable to 2016 review; median erosion proposed as preferred measure of central tendency.
  - Advantages of median erosion: reflects erosion experienced by representative member; avoids bias toward large economies inherent in aggregate erosion.
  - For continuity and comparability, staff complements median-based estimates with aggregate erosion.
  - End point for analyzing erosion updated to 2024 to incorporate October 2024 WEO forecasts.

*Source: ppea2025001 - 18.*

### Annex I. Table 1. Erosion Estimation Methodologies Across Recent Reviews

### Annex I. Table 1. Erosion Estimation Methodologies Across Recent Reviews

### Time period and concepts
- Time periods compared:
  - 2016 Review: 2009–2015
  - 2023/24 Interim Reviews: 2015–2023
  - 2025 Review: 2016–2024
- Economic concepts and metrics used as proxies:
  - Capacity to repay: GDP (all three reviews)
  - Current account / Current payments:
    - 2016 Review: Trade = Exports + Imports
    - 2023/24 Interim Reviews and 2025 Review: Current payments = Imports of goods and services + Primary income debit + Secondary income debit
  - Capital account / Capital flows:
    - 2016 Review: Non-FDI External Liabilities
    - 2023/24 Interim Reviews and 2025 Review: Capital flows = FDI + Portfolio liabilities + Other investment liabilities
  - Aggregate BoP needs / External financing needs (EFN):
    - 2016 Review: External financing needs (EFN) = – Current account balance + Amortization falling due in the next 12 months
    - 2023/24 Interim Reviews and 2025 Review: External financing needs = – Current account balance + Amortization – Change in reserves
      - Note: "Change in reserves is subtracted from EFN only for low-reserves countries. These are defined as countries with reserves covering less than 100 percent of short-term external debt on remaining maturity basis."

### Country coverage and proxies for likely borrowers
- Country coverage:
  - 2016 Review: Broadest coverage
  - 2023/24 Interim Reviews: All IMF members (World) with restrictions: for capital flows and EFN only countries with positive values; additionally for EFN: program countries (Program countries are countries that had at least one purchase under a GRA arrangement since 1990.)
  - 2025 Review: All IMF members (World) with restrictions for capital flows and EFN: only countries with positive values; Program countries
- Proxy for likely borrowers (EMDE focus and exclusions):
  - 2016 Review: EMDEs excluding India & China, with restriction for EFN: only countries with positive values
  - 2023/24 Interim Reviews: All EMDEs, with restrictions: for capital flows and EFN only EMDEs with positive values; additionally for EFN only: program countries
  - 2025 Review: All EMDEs, with restrictions for capital flows and EFN: only EMDEs with positive values; Program EMDEs with same restrictions; EMDEs excluding India & China with same restrictions (for improved comparability with 2016 Review)

### Central tendency measures
- 2016 Review: Erosion of aggregate for country sample
- 2023/24 Interim Reviews: Erosion of aggregate for country sample; preferred measure: Median erosion within country sample (Median of erosions observed in country sample)
- 2025 Review: Median erosion within country sample (preferred measure) and erosion of aggregate for country sample (for comparability)

*Italic footnotes from source preserved verbatim where present.*

---

### Annex II — Analysis of Changes in Credit Risks Since the 2016 Comprehensive Review of Access

### Key findings and context
- Overall: Measures of credit risk in the Fund’s lending portfolio generally point to some weakening since 2016, but this has been mitigated by an increase in financial buffers, limiting the impact on residual credit risk (Annex Table 1).
- Drivers of weaker credit risk indicators:
  - More challenging economic and financial landscape following global shocks.
  - Significant expansion in total commitments during 2016-23.
  - Expansion in actual and projected total credit outstanding and tripled exposure to the largest borrower.
  - Commitments under precautionary arrangements in 2023 were substantially lower than in 2016.
  - Sovereign credit ratings signal rising stress for some borrowers, particularly the large ones.
  - Share of lending to countries rated CCC or below increased substantially.
- Mitigating factors:
  - Precautionary balances reached the current medium-term target of SDR 25 billio  n by end-FY2024; and the burden sharing capacity has rapidly expanded mainly on the back of the higher SDR interest rate and the absence of arrears cases recently.

### Annex II. Table 1 — Selected credit risk indicators (In billions of SDRs, end December 2023 unless otherwise indicated)
- Indicators20162023Change
- Risks
  - Total commitments 159.0 174.0
  - Commitments under precautionary arrangements 1/ 89.0 58.0
    - of which FCL and PLL arrangements 86.1 56.6
  - Credit outstanding
    - Actual 49.3 92.1
    - Projected peak 2/ 49.3 97.6
  - Largest individual exposure
    - Actual 10.1 30.4
    - Projected peak 2/ 22.9 34.2
  - Credit concentration
    - Top 5 (in percent of total) 80.2 68.0
    - Top 1 (in percent of total) 26.0 33.0
  - Share of largest regional exposure in total commitments (percent) 45.9 58.7
  - Share of largest regional exposure in total credit outstanding (percent) 55.8 48.0
  - Share of RFI in the credit portfolio (percent) 2.4 15.4
  - Weighted sovereign credit rating of Fund credit exposures (S&P) 3/ 12.5 15.7
  - Weighted sovereign spreads of largest five borrowers (basis points) 480 1962
  - Share of Fund credit to members rated CCC or lower (percent) 0.05 2.9
  - Arrears (SDR millions) 254.1 0.0
- Buffers
  - Precautionary balances 4/ 15.2 24.4
  - Precautionary balances (in percent of) 4/
    - Credit outstanding 31.8 26.5
    - Total commitments 12.1 14.0
    - Lending capacity 2.2 3.5
  - Burden sharing capacity (SDR millions) 30.5 1329.5
- Other
  - Lending capacity 4/ 684.5 695.6

- Notes from table preserved verbatim:
  - 1/ Undrawn balances under arrangements treated as precautionary by the authorities.
  - 2/ Data for 2016 are actual and not projected.
  - 3/ Sovereign credit ratings are assigned numerical scores. The weaker the rating the higher the score. A higher weighted average rating of Fund exposures reflects a deterioration in ratings.
  - 4/ Data for 2016 as at end-Fiscal Year

---

### Annex III — Capacity to Repay (CtR) and Access Limits

### A. Fund borrowers’ CtR trends and interpretation
- Overall assessment:
  - Borrowers’ capacity to meet debt obligations to the Fund has remained broadly stable over the past 15 years, despite some upticks among debt service indicators in recent years.
  - CtR indicators worsened recently among large Fund borrowers, including in the context of Exceptional Access (EA) programs.
  - An event study around the 2016 increase in access limits (45 percent on average in SDR terms) finds no statistically significant differences in the distribution of CtR indicators for IMF-supported normal access programs approved before and after the 2016 increase, suggesting a limited impact of access limit increases on Fund borrowers’ CtR.
- CtR indicators considered:
  - Stock indicators: credit outstanding relative to GDP, international reserves, external debt.
  - Flow indicators: debt service (repurchases and charges) relative to exports, government revenues, external debt service.
- Interpretation of metrics:
  - Nominal GDP, exports, international reserves, external debt and debt service, and government revenues are used to capture various dimensions of repayment capacity and vulnerability to external shocks.
- Heterogeneity:
  - Analysis accounts for differences across borrower classifications (normal access vs EA).
  - Use of both unweighted average and median to avoid aggregation bias.

### B. Empirical findings on CtR indicators
- Stock indicators:
  - For most countries, credit outstanding as a share of GDP, reserves, and external debt has remained low and stable.
  - Notable increase in the top deciles of the distribution, indicating disproportionate deterioration for certain members, especially credit outstanding in percent of external debt.
  - Pattern primarily driven by large borrowers with exceptional access; CtR indicators for normal access borrowers remained relatively stable, with a recent improvement in average and median CtR stock indicators.
- Flow indicators:
  - Obligations to the Fund relative to exports, fiscal revenues, and external debt service show a slight upward shift in overall distribution.
  - Deterioration primarily driven by borrowers with exceptional access; some deterioration also observed for normal access countries (possibly linked to longer program lengths and the sharp increase in global interest rates after 2021).

### C. Event study of the 2016 access limit increase (design and results)
- Design:
  - The 2016 increase in GRA access limits produced an average increase in access limits of 45 percent in SDR terms.
  - Treated as an exogenous event because it was triggered by the implementation of the 14th GRQ rather than an economic shock.
  - Analysis compares CtR indicators (projected at time of approval) for normal access programs approved three years before and three years after the 2016 increase, tracking indicators over five years post-approval.
  - Sensitivity check: results robust when narrowing the window to 18 months before and after the 2016 increase.
- Results:
  - Distributions of CtR indicators for programs approved before and after the 2016 increase show considerable overlap.
  - Expansion in the upper deciles of the distributions is evident for programs approved after the 2016 increase, concentrated among large borrowers.
  - For stock variables (credit outstanding as a share of GDP and international reserves), widening of the upper distributions is observed throughout the five years following program approval.
  - For flow indicators (obligations to the Fund as a share of exports and fiscal revenues), widening of the upper distribution becomes apparent from the third year after program approval—possibly reflecting the shift toward EFF programs with longer grace and repayment periods.
  - Statistical tests:
    - Two-sample t-test (assuming normality) and Mann-Whitney U-test (non-parametric) performed.
    - Outcomes indicate no statistically significant differences at the 95 percent confidence level in mean or distribution of CtR indicators between programs approved before and after the 2016 increase.
  - Conclusion: Despite deterioration in upper deciles for some indicators, the mean and overall distribution comparisons suggest a limited impact of the 2016 access limit increase on borrowers’ capacity to repay.

*Italic: Source: ppea2025001 - Annex I. Table 1. Erosion Estimation Methodologies Across Recent Reviews (selected tables and annexes excerpt).*

### Annex III. Figure 3. Interquartile range of the projected member’s Capacity to Repay by

### Annex III. Figure 3. Interquartile range of the projected member’s Capacity to Repay by program approved before and after the 2016 CRAL

### Figures and statistical presentation
- Figures show the 25th and 75th percentile intervals.  
- The dashed line represent the medians.  
- Figure panels (indicators shown):
  - Credit Outstanding as Percent of GDP
  - Credit Outstanding as Percent of Reserves
  - Credit Outstanding as Percent of Public External Debt
  - Obligations to the Fund as Percent of Exports of Goods and Services
  - Obligations to the Fund as Percent of Fiscal Revenues
  - Obligations to the Fund as Percent of Public External Debt Service

### Annex III. Figure 4 — Statistical test description
- Figures show the 95 percent confidence interval of the statistics testing for the difference in means of both distributions.  
- The solid line is value of the statistics.  
- Time horizons presented in panels: t, t+1, t+2, t+3, t+4, t+5 (across multiple indicator panels).

### Key finding from text summary
- "Overall, drawing from the 2016 experience, making permanent the current temporary access limits would be expected to have a limited effect on the overall distribution of IMF borrowers’ capacity to repay."  
- The text highlights that the relatively wide range of CtR indicators suggests that higher access limits can be useful in some cases because they allow the Fund to remain nimble in its support to its members.  
- Decisions on the level of access of an individual member should continue to be guided by:
  - the member’s actual, prospective, or potential balance of payments needs,
  - the strength of its economic reform program,
  - the EA framework when applicable,
  - the amount of its outstanding use of Fund resources,
  - its capacity to repay the Fund.

*Source: ppea2025001 - Annex III. Figure 3. Interquartile range of the projected member’s Capacity to Repay by*

---


_Source: https://www.imf.org/-/media/files/publications/pp/2025/english/ppea2025001.pdf_