## SRDSF GUIDANCE NOTE — EXECUTIVE SUMMARY & SELECTED SECTIONS (ppea2022039) ## Source details **Canonical URL:** [SRDSF GUIDANCE NOTE — EXECUTIVE SUMMARY & SELECTED SECTIONS (ppea2022039)](https://www.imf.org/-/media/files/publications/pp/2022/english/ppea2022039.pdf) ## Other formats - [Markdown version](/-/media/files/publications/pp/2022/english/ppea2022039.pdf.md) - [Structured JSON version](/-/media/files/publications/pp/2022/english/ppea2022039.pdf.json) --- ### Purpose and adoption - Provides operational guidance for the use of the Sovereign Risk and Debt Sustainability Framework (SRDSF), which replaces the Debt Sustainability Framework for Market Access Countries. - Introduces improvements in organization, methodology, transparency, and communication for analyzing public debt in countries that mainly finance themselves with market-based debt. - Phased adoption begins [June 2022]; after adoption it will become the Fund’s principal tool for assessing public debt sustainability. - Document dated: July 19, 2022. ### Roles within IMF operations - Supports surveillance and lending: - Surveillance: early warning system gauging debt-related risks and identifying policy recommendations to prevent potential stress. - Lending / Fund-supported programs: assesses public debt sustainability (required for all IMF lending) and, where debt is unsustainable, provides methodology for setting targets to guide debt restructurings. ### Core outputs and definitions - Two main outputs: - Sovereign risk assessment (vulnerability to sovereign stress events). - Debt sustainability assessment (prospects for stabilizing debt in the baseline scenario). - Key definitions preserved from source: - Sovereign stress: an event where market and/or fiscal pressures related to public debt become acute; no presumption on resolution method. - Unsustainable debt: no politically and economically feasible policies stabilize debt-to-GDP and deliver acceptably low rollover risk without restructuring and/or exceptional bilateral support, even with Fund financing. - Debt non-stabilization under the baseline: debt/GDP ratio not expected to stabilize under best prediction of policies by the end of the projection horizon; not always equivalent to stress or unsustainable debt. ### Structure of the SRDSF and main tools - Time-horizon based analyses supported by modules: - Core subset for all countries informs near- and medium-term assessments. - Additional specialized analyses for broader medium- and long-term risks. - Major tool categories (by section): - Section III: Debt coverage and disclosure, debt structure, baseline scenario. - Section IV: Realism tools to detect/discourage overly optimistic projections. - Section V: Near-term Early Warning System (logit model) predicting sovereign stress over short (1-2 year) horizons. - Section VI: Medium-term analysis: - Debt Fanchart Module (solvency risks over next 5 years). - Gross Financing Needs (GFN) Module (liquidity/financeability risks). - Triggered stress tests. - Section VII: Optional long-term modules (climate, demographics, large amortizations, natural resources). - Section VIII: Synthesis into bottom-line assessments and restructuring targets when needed. - Section IX: DSA requirements and publication. ### Methodology and judgment - Standardized framework that requires informed judgment where models do not capture circumstances. - Realism tools include past projection diagnostics and distributional checks and specific checks on fiscal adjustments, REER gap, real GDP growth, fiscal multipliers, and financing terms. ### Application guidance and users - Applies to market access countries (MACs) — all advanced economies and most emerging market economies; some PRGT-eligible countries with substantial and durable market access may also use SRDSF. - Approved by IMF Executive Board in January 2021; replaces MAC DSA. - After SRDSF becomes effective for a country, Fund staff should include SRDSAs in all policy notes and staff reports sent for interdepartmental review. - Reading guidance for users: Sections II, III, IV, V, VI, VII.A, VIII.A essential for first-time users; program preparers should read Section VIII.B; full methodological review should include annexes and calibration details. ### Practical uses in program contexts - Precautionary arrangements: verify qualification requirements on debt sustainability. - Nonprecautionary programs: assess whether adjustment and new financing can resolve stress or whether exceptional measures (e.g., restructuring) are required. - When restructuring is required: SRDSF informs debt relief envelope and sets targets. ### Coverage choices, consolidation, and instrument treatment - Default perimeter: General Government (GG); broader coverage permitted if it anchors fiscal policy or captures material risks (e.g., NFPS, CPS) provided appropriate consolidation and disclosure. - Contingent liabilities included in baseline when likelihood of materialization and cost can be anticipated and estimated. - Central bank consolidation appropriate when central bank has large negative capital or is engaged in quasi-fiscal activities; otherwise central bank holdings are internalized via GG coverage and GFN module. - All debt instruments as in PSDS Guide included: debt securities, loans, currency and deposits, SDRs, accounts payables, IPSGSs. - IPSGSs omission excused if explained; large IPSGSs should be discussed and authorities urged to expand coverage. - Accounts payables should be estimated where cash accounting limits data. - Fund disbursements always included as public debt for SRDSA purposes. - Liquidity papers excluded only if three conditions met; otherwise included and memoed when excluded. - Central bank bilateral FX swap liabilities included when drawn unless exclusion conditions (liquidity support, central bank expected to extinguish without government support) satisfied; de-minimis exclusion if drawn swaps < 1 percent of GDP. ### Accounting and valuation - Fiscal accounts: accrual basis recommended; debt stock reported at nominal value per PSDS guide. - Deviations from recommended accounting must be disclosed and explained; if cash basis used, clarify whether arrears included. ### Realism tools — purpose and guidance - Suite of nine realism tools automatically produced to assess realism of baseline drivers and guard against optimism. - When flags appear, SRDSA must explain why they do not signal overoptimism; strong justification required if multiple tools flag issues. - Tools include forecast track record, output gap revisions, debt drivers decomposition, distribution of debt-to-GDP reductions, fiscal adjustment realism, REER gap realism, GDP growth realism, consistency with fiscal multipliers, and financing terms/issuance composition. ### Near-term assessment — Logit model (Section V) - Key output: Logit Stress Probability (LSP) — fitted probability that a stress event will materialize within 1-2 years; model uses realized values (no projections). - Explanatory variable groups: structural characteristics (institutional quality, stress history), cyclical position (current account/GDP, REER change, credit-to-GDP gap), debt burden and buffers (change in debt-to-GDP, debt/revenues, FX public debt/GDP, reserves/GDP), global conditions (VIX change; alternate for currency unions). - Criteria for sovereign stress (Table 2) include large IMF/IFI disbursements, defaults (external arrears ≥ 5 percent of public external debt and increasing ≥ 10 percent nominally), restructurings, chronic excessive inflation (doubling and >25 percent or >100 percent), market indicators/loss of market access (thresholds for spreads), and financial repression indicators. - Mechanical signal thresholds (Annex IV.A): - Low risk if LSP < 6.3 percent. - High risk if LSP > 19.5 percent. - Moderate otherwise. - Posterior probabilities by fitted signal (calibration sample): - High: 0.40 average probability (near-term tool high-risk zone). - Low: 0.02 average probability (near-term tool low-risk zone). - Use of judgment allowed for imminent events, temporary distortions, structural country-specific factors, timing issues, and factors outside the model. - Adjustments: augment international reserves with large, liquid government-controlled assets when appropriate; adjust credit gap discontinuities with clear documentation; alternate specification for currency union members. ### Medium-term assessment — Debt Fanchart and GFN Modules (Section VI) - Debt Fanchart Module: - Simulates many debt trajectories (standard 10,000 trajectories) via block bootstrapping of historical shocks; produces stochastic distributions and a debt fanchart. - Preliminary historical fanchart compares baseline to past behavior; realism adjustment triggered when projected debt-to-GDP falls below the 20th percentile in two or more years. - Final fanchart variants: - Baseline-centered (no realism issue): de-meaned shocks added to baseline. - Adjusted (realism activated): entire fanchart shifted upward so baseline corresponds to a comparator-group percentile. - Debt Fanchart Index (DFI) components: - Fanchart width = terminal 95th − 5th percentile. - Probability of debt non-stabilization = 1 − share of stabilizing trajectories (stabilizing defined using debt-stabilizing primary balance from Box 3). - Terminal debt level adjusted by an institutions index (higher institutions index → lower implied risk). - Liquid asset buffers override: if assets > 75 percent of GDP and 100 percent of public debt, module mechanical signal should be low regardless of DFI. - DFI mechanical signal thresholds (Annex IV.B): low risk if DFI < 1.13; high risk if DFI > 2.08; otherwise moderate. - GFN Module: - Assesses liquidity risks, constructs GFN Financeability Index (GFI) from: - Average baseline GFN-to-GDP, - Initial bank exposures to government (pct banks' assets), - Change in bank claims on the government in a generalized stress scenario (pct banks' assets). - Generalized stress scenario default shocks include: growth reduction (1 stdev for two years), nominal marginal interest rate rise by 300 bps upfront phased out over 5 years, one-off exchange rate depreciation equal to max in past 10 years, maturity shortening, and a debt-holder shock (foreign private rollover 67% for two years). - Mechanical signal thresholds for GFI: low risk if GFI < 7.6; high risk if GFI > 17.9; otherwise moderate. - Adjustments allowed in limited cases: central bank quantitative easing (match announced purchases), live precautionary Fund arrangements (add drawable resources to liquid assets), firm nonmarket financing commitments; exceptional adjustments require strong justification and interdepartmental/Management approval for staff SRDSAs. - Aggregation into Medium-Term Index (MTI): - MTI aggregates normalized DFI and GFI (weights and normalization described in Annex III/IV). - MTI mechanical signal thresholds: - Low risk if MTI < 0.257. - High risk if MTI > 0.395. - Moderate otherwise. - Posterior probabilities for medium-term index signals (calibration sample): - High: 0.43 average probability for high-risk MTI zone. - Low: 0.04 average probability for low-risk MTI zone. - Triggered stress tests: - Five standard tests: banking crisis, commodity price shocks, contingent liabilities from narrow coverage, REER correction, natural disasters. - Tests are activated based on triggers but may be manually applied; they inform judgment but have no standalone mechanical signal. - Banking crisis stress test calibration examples: - Trigger thresholds: credit-to-GDP gap = 10 percent; mispricing risk index percentile = 67. - First-round fiscal cost: Advanced economies 6.8 percent of GDP; Emerging markets 10.0 percent of GDP. - Commodity price, contingent liability, REER shock, and natural disaster calibrations detailed in source. ### Medium-term example (sample country "Ruritania" — selected numeric outputs) - Debt fanchart module outputs (Figure 23 sample): - Fanchart width (percent of GDP): 24.8 - Probability of debt non-stabilization (percent): 5.9 - Terminal debt-to-GDP x institutions index: 25.1 - Debt fanchart index (DFI): 0.8 → Risk signal: Low - GFN module outputs (Figure 23 sample): - Average baseline GFN (percent of GDP): 16.2 - Banks' claims on the gen. govt. (pct banks' assets): 7.4 - Chg. in banks' claims in stress (pct banks' assets): 0.7 - GFN financeability index (GFI): 8.2 → Risk signal: Moderate - Medium-term aggregation: - Weight on Debt fanchart index (normalized): 0.5 - Weight on GFN financeability index (normalized): 0.5 - Medium-term index value: 0.19 → Risk signal: Moderate - Probabilities reported: Prob. of missed crisis, 2022-2027: 9.1 pct; Prob. of false alarm, 2022-2027: 61.1 pct ### Long-term assessment and optional modules (Section VII) - Long-term assessment required element in all Fund SRDSAs; qualitative given high uncertainty; revise 10-year baseline if long-term analysis reveals material fiscal costs affecting 5-10 year horizon. - Optional long-term modules (voluntary except where mandatory): demographics (pensions, healthcare), natural resources, large debt amortizations, climate change (adaptation and mitigation). - Demographics modules: - Pension sub-module projects net pension expenditures to year 2100 and NPV of future financing needs; default real discount rate: 5 percent; real risk-free rate for pension assets: 3 percent. Mandatory use if pension expenditure growth over 2022-50 or old-age dependence growth over 2022-50 exceeds 75th percentile of sample. - Healthcare sub-module projects healthcare expenditures to 2100; default excess healthcare cost growth: 0.6 percent for EMs and LIDCs, 1.4 percent for AEs; real discount rate: 5 percent. - Natural resources module models t+6 to t+15 extraction scenarios and cash flows; include when extraction volumes deviate by more than 1 stdev from historical average. - Large Debt Amortizations module projects GFNs over a 25-year horizon and flags risk when variables exceed 10-year historical average by >1 stdev. - Climate change modules (adaptation and mitigation) project debt-to-GDP and GFN-to-GDP over a 30-year horizon: - Adaptation sub-module standard scenario uses IMF adaptation cost estimates (Aligishiev, Bellon and Massetti, 2022); customized scenario must be incorporated in 5–10 year baseline when relevant; adaptation sub-module compulsory for countries highly exposed to natural disasters and in some restructuring cases. - Mitigation sub-module uses EU-based proxies scaled to country emissions; required for countries with net-zero targets before 2050 and for 25 largest CO2 emitters per unit of output; default assumes fully government debt financing (upper bound). ### Sustainability assessments and probabilistic thresholds (Section VIII) - Debt sustainability definition (IMF Executive Board): public debt sustainable when the primary balance needed to stabilize debt under baseline and realistic shock scenarios is economically and politically feasible, preserving acceptable rollover risk and potential growth. - Probabilistic thresholds: - Any sustainable assessment associated with at least 50 percent probability. - "High probability" of sustainable debt: at least 80 percent probability. - Unsustainable debt event defined as probability < 20 percent. - Mechanical components for sustainability assessment: - Internal sustainability logit model (staff-only), DFI, and GFI combined into a numerical sustainability index compared to calibrated thresholds. - Staff judgment plays a central role and can override mechanical outcomes, especially for borderline cases, conflicting tool results, distorted variables, omitted factors (e.g., triggered stress tests or long-term module results), or clear signals of unmanageable public debt. - Additional requirement: in certain cases (exceptional access, drawing scenarios, or when shocks not captured by medium-term modules) sustainability must be assessed under an adverse/drawing scenario treating full Fund draw as part of public debt. ### Use of SRDSF tools in restructurings and program contexts - When debt is unsustainable and restructuring is undertaken, SRDSF tools inform debt relief envelope and targets: - Use 10-year horizon including essential adaptation investment costs where relevant. - GFN targets from GFN Module verify manageability post-restructuring; adjust debt relief envelope until both GFN and fanchart indicate sustainability. - Judgment and complementary targets address specific vulnerabilities. - SRDSAs required: - All IMF financing requests must be accompanied by an SRDSA. - For normal-access Fund arrangements, SRDSA included annually unless circumstances warrant updates. - Exceptional access arrangements require SRDSA updates at every program review. ### Reporting, governance, and publication - Standardized SRDSA reporting elements: - Summary table with mechanical and final risk signals at each horizon, mechanical signals from debt fanchart and GFN modules, triggered stress tests, sustainability assessment, and statement on baseline debt stabilization. - Summary chapeau paragraph with overall assessment and intuition (mechanical sustainability assessment not published). - Debt coverage and disclosure metadata, debt profile charts, baseline scenario table and decomposition, realism tools outputs, near-term and medium-term mechanical outputs, and long-term module outputs if used. - Governance: judgment in SRDSAs prepared by Fund staff is scrutinized in interdepartmental review; Management may adjudicate disagreements. - Publication constraints (Transparency Policy): - Certain elements deleted prior to publication: - Near-term assessment elements (signal, final assessment, commentary) not published prior to a review 12 months after SRDSF roll-out. - Mechanical signal on debt sustainability deleted before publication. - Probability qualifier ("with high probability" / "but not with high probability") removed except where required by lending policies (e.g., Exceptional Access). - Staff should prepare both Board document SRDSA and a version with required deletions for the Transparency Portal. *Source: IMF — SRDSF GUIDANCE NOTE (EXECUTIVE SUMMARY and selected sections), ppea2022039.* ### EXECUTIVE SUMMARY ### EXECUTIVE SUMMARY ### Purpose and adoption - Provides operational guidance for the use of the Sovereign Risk and Debt Sustainability Framework (SRDSF), which replaces the Debt Sustainability Framework for Market Access Countries. - Introduces improvements in organization, methodology, transparency, and communication for analyzing public debt in countries that mainly finance themselves with market-based debt. - Phased adoption begins [June 2022]; after adoption it will become the Fund’s principal tool for assessing public debt sustainability. - Document dated: July 19, 2022. ### Roles within IMF operations - Supports the Fund’s surveillance and lending functions: - Surveillance: acts as an early warning system gauging debt-related risks and helps identify policy recommendations to prevent potential stress. - Lending / Fund-supported programs: helps assess public debt sustainability (a requirement for all IMF lending) and, where debt is unsustainable, provides methodology for setting targets to guide debt restructurings. ### Core outputs and definitions - Two main outputs: - Sovereign risk assessment (vulnerability to sovereign stress events). - Debt sustainability assessment (prospects for stabilizing debt in the baseline scenario). - Key concepts defined exactly as in the source: - Sovereign stress: an event where market and/or fiscal pressures related to public debt become acute; no presumption on whether pressures are resolved by adjustment, financing, or exceptional measures. - Unsustainable debt: occurs when there are no politically and economically feasible policies that stabilize the debt-to-GDP ratio and deliver acceptably low rollover risk without restructuring and/or exceptional bilateral support, even in the presence of Fund financing. - Debt non-stabilization under the baseline: debt/GDP ratio is not expected to stabilize under the best prediction of policies by the end of the projection horizon; non-stabilization is not always equivalent to sovereign stress or unsustainable debt. ### Structure of the SRDSF and main tools - Time-horizon based analyses supported by modules: - Core subset applicable to all countries informs near- and medium-term assessments. - Additional specialized analyses for broader risks at medium and long-term horizons. - Major tool categories (organized by section in the note): - Section III: Debt coverage and disclosure, debt structure, baseline scenario. - Section IV: Realism tools to detect/discourage overly optimistic projections. - Section V: Near-term risk analysis — Early Warning System predicting sovereign stress over short (1-2 year) horizons using indicators of institutions, cyclical position, debt burden and buffers, and global conditions. - Section VI: Medium-term risk analysis comprising: - Debt Fanchart Module for solvency risks over the next 5 years. - Gross Financing Needs Module for liquidity/financeability risks over the medium term. - Triggered stress tests for specific vulnerabilities. - Section VII: Optional long-term modules for climate change, demographics (social security, pension, healthcare), large debt amortizations, and natural resources. - Section VIII: Synthesis into bottom-line assessments of sovereign risk and debt sustainability; informs restructuring targets when needed. - Section IX: DSA requirements and publication. ### Methodology and judgment - The SRDSF is standardized but emphasizes that informed judgment is integral: models cannot capture every circumstance and assessments should reflect additional considerations where relevant. - Realism tools cover: - Past projection diagnostics and distributional checks (e.g., debt drivers decomposition, distribution of debt-to-GDP reductions). - Specific realism checks on fiscal adjustments, REER gap, real GDP growth, fiscal multipliers, and financing terms. ### Application guidance and users - SRDSF applies to market access countries (MACs) — all advanced economies and most emerging market economies; some PRGT-eligible countries with substantial and durable market access may also use the SRDSF. - Approved by the IMF’s Executive Board in January 2021; replaces the MAC DSA. - After the SRDSF becomes effective for a country, Fund staff should include Sovereign Risk and Debt Sustainability Analyses (SRDSAs) in all policy notes and subsequent staff reports that include debt sustainability analyses and are sent for interdepartmental review. - Reading guidance for users: - First-time users: read Section II (key concepts), Sections III and IV (critical inputs), and the “standard application” and “interpreting the tool” parts of Sections V and VI; Sections VII.A and VIII.A are essential. - Users preparing SRDSAs for Fund-supported programs should read Section VIII.B. - Full methodological review (e.g., for training or review) should include Sections II, V, VI, relevant "considerations for special cases" and "using judgment", Annex III (procedures), and Annex IV (calibration details). ### Practical uses in program contexts - In precautionary arrangements: helps verify qualification requirements on debt sustainability. - In nonprecautionary Fund-supported programs: assesses whether adjustment and new financing can resolve stress or whether exceptional measures (e.g., restructuring) are required to restore medium-term debt sustainability. - When restructuring is required, the SRDSF can inform the envelope of debt relief and set targets for restructuring. ### Organizational and companion material - The SRDSF guidance note contains detailed methodological modules, figures, boxes, tables, and annexes (including a model SRDSA and technical notes) to operationalize the framework for country analyses. *Source: IMF — SRDSF GUIDANCE NOTE (EXECUTIVE SUMMARY), July 19, 2022.* ### SECTION II. OVERVIEW ### SECTION II. OVERVIEW ### Framework purpose and scope - The SRDSF is a collection of modules that guide assessments on sovereign stress risks and debt sustainability. - The overall framework aims to be comprehensive, looking at risks across a range of subjects and time horizons. ### Core framework applied to all MACs - The core framework consists of three modules with common design aspects: - (i) the near-term early warning system (logit model) - (ii) the Debt Fanchart Module - (iii) the GFN Module - Shared features of the three modules: - Numerical risk metrics: each module transforms input variables into a single index (a risk metric) where a higher index value represents a higher risk of sovereign stress. - Near-term module: index represents the probability of sovereign stress within the next two years; inputs are the independent variables of a logit model. - Debt Fanchart and GFN Modules: inputs are analytical measures derived from debt and GFN simulations based on macroeconomic projections, weighted by their power to predict past stress events. - Debt Fanchart and GFN Module Indices are combined into an aggregate Medium-Term Index (MTI) using simple averaging. - False alarms and missed crises: the framework contemplates binary decision rules defined by a threshold 휏 to partition index values into lower and upper segments, with two types of misclassification: - False alarm (Type I error): predict a crisis although no crisis materializes. - Missed crisis (Type II error): predict no crisis when a crisis materializes. - Misclassification probabilities: - Missed crisis probability for threshold 휏: share of sovereign stress events with an index value of 휏 or less (number of crisis events with index ≤ 휏 divided by number of all stress events). - False alarm probability for threshold 휏: share of non-stress events with an index value of 휏 or higher. - Thresholds and mechanical signals: - The framework divides risk index into “low”, “moderate” and “high risk” zones defined by two calibrated thresholds — lower (휏l) and upper (휏h). - Low-risk threshold 휏l is chosen so that a decision rule predicting “no sovereign stress” for any index value below 휏l implies a missed crisis probability of 10 percent. - High-risk threshold 휏h is calibrated so that a decision rule predicting “sovereign stress” for any index value above 휏h implies a false alarm probability of 10 percent. - Index values: below lower threshold = low risk; above higher threshold = high risk; middle = moderate risk. This is the “mechanical signal” of the module. ### Posterior probabilities of stress (from calibration sample) - Near-term tool: - Average probability of stress for an index value in the high-risk zone: 0.40 - Average probability of stress for an index value in the low-risk zone: 0.02 - Medium-term index: - Average probability of stress for an index value in the high-risk zone: 0.43 - Average probability of stress for an index value in the low-risk zone: 0.04 - The lowest stress probability associated with a high-risk signal in either of the tools is about 20 percent. - Interpretation note: “high risk” signals do not imply that a stress event is the most likely outcome, but indicate risk is sufficiently high to be taken seriously. ### Comparator groups - SRDSF outputs include comparisons against relevant comparator groups based on Fund engagement status and economic development: - (i) surveillance-only AEs (e.g., United States and Japan, 2022) - (ii) AEs with Fund programs (e.g., Greece, 2010-18) - (iii) surveillance-only EM commodity exporters (e.g., Oman, 2022) - (iv) surveillance-only EM non-commodity exporters (e.g., China, 2022) - (v) EM commodity exporters with Fund programs (e.g., Angola, 2018-21) - (vi) EM non-commodity exporters with Fund programs (e.g., Pakistan, 2019-22) - Note: The commodity/non-commodity exporter comparison is not available for advanced economies, given very few advanced economy commodity exporters. ### Additional modules, horizons, and aggregation - Additional modules for specialized analysis: - Scenario analyses triggered when country characteristics suggest further scrutiny. - Optional long-term modules to analyze issues materializing over a longer term. - Horizon-specific assessments and aggregation: - Near-term risk assessment informed by the Early Warning System (logit model). - Medium-term assessment synthesizes Debt Fanchart and GFN Modules; includes alternative scenario analyses when relevant. - Long-term assessment informed by optional long-term modules when included. - Final overall assessment: SRDSF users consider all results plus prospects for debt stabilization in the baseline to arrive at an overall assessment at their discretion. ### Debt sustainability assessments and relationship to SRDSF - Debt sustainability assessments (DSAs) can be added when needed, typically after stress materializes to inform resolution and program design. - Differences from stress framework: - DSAs parallel the core modules but are designed and calibrated differently. - DSA output is a single mechanical signal with three outcomes: debt is sustainable with a high probability; sustainable but not with high probability; unsustainable. - The stress framework outputs high/moderate/low risk signals. ### Use of judgment and governance - The framework governs and guides the use of judgment: - Judgment-based final assessments are required when mechanical signals are counterintuitive or when standard tools do not provide a mechanical signal (e.g., long-term, overall risk assessments). - Relevant considerations for each tool are elaborated throughout the guidance note. - Judgment in SRDSAs prepared by IMF staff is scrutinized in the interdepartmental review process, with IMF Management arbitrating disagreements across departments. - The review can involve scrutiny of both presentation of results and quantitative operation of the tools, operating with appropriate information sharing, including underlying SRDSA files. ### Reporting and standardized summary - Mechanical signals, final assessments, and uses of judgment are summarized in a standardized reporting table in every SRDSA. - The reporting table captures mechanical signals, final assessments, horizon-specific comments, and sustainability assessment outcomes to support interpretation and policy discussion. - Note on concept distinctions: - The risk of sovereign stress is broader than debt sustainability. Unsustainable debt typically requires exceptional measures (such as debt restructuring). A sovereign can face stress without debt being unsustainable, and non-restructuring measures (fiscal adjustment, new financing) can address stress. *Source: SECTION II. OVERVIEW (SRDSF GUIDANCE NOTE), ppea2022039* ### 20. SRDSA coverage may be broader than the GG if the broader definition anchors fiscal ### 20. SRDSA coverage may be broader than the GG if the broader definition anchors fiscal ### Coverage decisions and rationale - SRDSA coverage may be broader than the general government (GG) if the broader definition anchors fiscal policy discussions or if there is a conceptual argument for doing so. - Country authorities may choose NFPS level or CPS basis reporting where: - SOEs play an important part in public investment (including through PPPs); - legislative requirements or policy discussions anchor analysis at those broader perimeters. - Broader definitions may be used in the SRDSA provided that: - they support the analysis of a country’s fiscal situation (including government policies and contingent liability risks); - they are consistent with Fund surveillance needs; and - they are prepared with the appropriate statistical treatment, most notably consolidation. - Ad hoc adjustments are inappropriate for debt risk analysis and should be eschewed. - When broader coverage is used, it should be accompanied by an explanation in the commentary box of the Debt Coverage and Disclosures reporting. - In Fund-supported programs, debt conditionality may be measured according to a debt coverage that differs from the general government; however, a general government perimeter remains the default expectation for SRDSAs prepared for that program’s documents. ### Inclusion of contingent liabilities and specific entities - Broader-than-GG coverage may be necessary to fully capture sovereign risks and potential mitigants from entities outside the GG. - Full or partial NFPS coverage could be appropriate if it captures material fiscal risks from nonfinancial public corporations. - Contingent liabilities should be included in debt projections when users can anticipate and estimate: - the likelihood of materialization; and - the associated cost to be incurred by the government. - Central bank consolidation is appropriate in cases of central banks with large negative capital positions and/or where the user considers the central bank to be involved in significant direct monetary financing of the budget or quasi-fiscal activities. - Such consolidation would imply that: - central bank claims on the government are netted out; and - central bank debt liabilities (excluding currency and deposits held by residents) are added. - When central banks have healthy balance sheets, the framework internalizes mitigating characteristics of central bank holdings and future seigniorage revenues through GG coverage and the GFN module without consolidation. ### Instrument coverage (PSDS Guide instruments) - All debt instruments as defined in the Public Sector Debt Statistics: Guide For Compilers and Users (PSDS Guide; IMF 2013a) should be included in the gross public debt concept used for SRDSAs. - These instruments include: debt securities, loans, currency and deposits, SDRs, accounts payables, and insurance, pension and standardized guarantee schemes (IPSGSs). - Any omissions should be identified in the disclosure table and included in the contingent liabilities stress test based on available data. Specific instrument guidance: - IPSGSs: - Many countries currently do not report IPSGSs in fiscal accounts or debt stock; their exclusion can be excused if explained in the commentary box. - Where IPSGS liabilities may be large, commentary should discuss whether they constitute a material risk; if so, authorities should be urged to expand debt coverage for this item. - Accounts payables: - Data may not be available in countries with cash accounting; users should still seek to include an estimate of the total amount, especially where they appear to be a symptom of fiscal stress. - Debt-like financial derivatives: - Instruments that cannot be clearly distinguished from debt instruments as defined in the GFSM (e.g., off-market swaps) should be dealt with on a case-by-case basis, after consulting relevant departments. ### Fund disbursements, SDRs, and guarantees - Fund disbursements should always be included as public debt for SRDSA purposes. - Fund credit is an obligation of the member country and should always be included in SRDSA debt concepts even when disbursements constitute balance-of-payments support and are held at the central bank. - When IMF credit is on-lent to the budget through the central bank, users should ensure the debt is not doubly counted. - SDR inclusion: - Inclusion of SDRs in the SRDSA’s debt perimeter depends on the member’s institutional setup and whether the SDRs are being used; users should follow guidance in IMF 2021c. - Government guarantees: - Normally considered contingent liabilities and generally not included in government debt until they are called. - Pursuant to the PSDS Guide, where users assess a high likelihood of guarantees being called (e.g., guarantee to a corporation in financial distress), those guarantees should be treated as government debt. - Other contingent liabilities (likely legal settlements, SOE/bank recapitalization needs, commitments for PPPs, long-term leases, and other debt-like longer term financial liabilities) should be included in baseline debt projections when users assess materialization as both material and the most likely outcome, and when it is possible to forecast with relatively good accuracy. - Expected losses from standardized guarantee schemes should be reported as government debt under IPSGSs. ### Liquidity papers and central bank instruments - Liquidity papers (central bank securities for monetary policy) would normally be excluded from public debt in the SRDSA provided that all three conditions are met: - (i) A strong institutional framework exists to ringfence the proceeds so no financing to the government can be provided through their issuance; - (ii) The government is not responsible for paying interest on these securities; and - (iii) The securities do not represent a material fiscal risk (e.g., track record of central bank independence, or size of liquidity paper small relative to the capital position of the central bank). - When excluded, the stock of liquidity paper should be included as a memo item in standard reporting on the Baseline Scenario. - If any of the three conditions are not met, liquidity papers should be included in public debt and gross financing need measures used for the DSA. - If the chosen DSA perimeter includes the central bank, liquidity papers would be included in public debt for SRDSA purposes. ### Central bank FX swaps and official creditor deposits - Central bank bilateral FX swap liabilities should be included in public debt when drawn, unless certain exclusion conditions are met. - It would typically be inappropriate to include these liabilities in public debt when both conditions below are met: - The FX swap was drawn to support central bank liquidity operations designed to provide FX liquidity for financial stability purposes (as opposed to sovereign-to-sovereign medium-term balance of payments support); and - The central bank is expected to be able to extinguish the swap position without actions detrimental to government debt levels (e.g., outright government foreign borrowing to pay off the swap). This is typically the case if the borrower is: - a reserve currency issuer that pays its outstanding debt with its own currency with no consequence for public debt; or - a non-reserve currency issuing central bank that is financially strong and expected to discharge its obligations without sovereign support. - When exclusion criteria are not met (e.g., swap used for medium-term BOP support or central bank repayment capacity is in doubt), the gross drawn amount should be included in public debt. - Large undrawn bilateral FX swaps should be fully disclosed and built into projected debt stock to the extent drawing is expected (informed by BoP projections). - Official creditor deposits at the central bank should be included in SRDSA’s debt stock unless the central bank is expected to be able to repay them; assessment mirrors the FX swap repayment-capacity criteria. Box 2 — Decision tree highlights for central bank FX swaps: - Step 1: If the FX swap has not been drawn (no outstanding balance), it should not be included in public debt. - Step 2: A de-minimis threshold: the liability can be excluded from public debt if the total amount of all swaps drawn is less than 1 percent of GDP. - In borderline cases, users should consider including these liabilities to avoid year-to-year fluctuations. - Step 3: Include in public debt in full if the swap constitutes medium-term BOP support rather than provision of FX liquidity for financial stability. Indicators of medium-term BOP support include: - drawn to pay back another swap, finance government debt service, or sterilize monetary financing to the government; - maturity of the liability is 12 months or more (or strong reasons to believe a short-term liability will be rolled over repeatedly); - no matching claims against domestic commercial banks with corresponding maturities and conditions. - Step 4: If scheduled repayments of all FX swap liabilities lower central bank gross reserves below [60] percent of ARA metric, central bank capacity to respond to a BOP crisis is considered critically impaired and the swap should be included in public debt. - The guidance notes a cross-country sample (excluding reserve currency issuers) showing nearly 90 percent of countries with reserves below 60 percent of the ARA metric in 2021 exhibited some indication of a stress event; for countries between 60 and 100 percent of the ARA metric, only a bit more than half exhibited stress. ### Accounting principles and valuation - Fiscal accounts should be reported on an accrual basis and the debt stock for SRDSA purposes should be reported at its nominal value. - The GFSM (IMF 2014) uses the accrual basis as the standard for the Fund’s presentation of fiscal data; accrual provides comprehensive reporting of amounts a government owes to creditors, including arrears. - The PSDS guide advises valuing debt instruments at nominal values for vulnerability and sustainability analysis. - When national authorities do not prepare data in line with these recommended accounting principles, the authorities’ data may be used (which could include cash basis recording and/or face or market valuations). - Deviations from SRDSF recommendations must be disclosed in the table and explained in the accompanying commentary for Debt Coverage and Disclosures. - If a country reports debt on a cash basis, the SRDSF must clarify if any arrears are included in the debt stock. *SRDSF GUIDANCE NOTE — INTERNATIONAL MONETARY FUND* ### 29. Claims of government entities on other units within the same perimeter require ### 29. Claims of government entities on other units within the same perimeter require ### Consolidation rules and reporting requirements - Claims of government entities on other units within the same perimeter require consolidation: when computing total government debt within a perimeter, net out (do not include) liabilities among these entities because they do not constitute outside claims on the consolidated entity. - When public debt statistics are prepared on a consolidated basis: - Users should select "consolidated" in the standard reporting on Debt Coverage and Disclosures. - Users should fill out the accompanying consolidation table consistent with the SRDSA’s perimeter showing gross debt outstanding by each level of government and the crossholdings that are netted out in the final consolidated debt position. - Note on table entry: To be consistent with the last published public debt observation, only the cross holdings that were netted out of the public debt calculation should be entered in the table. If cross holdings are not netted (including because the sector is not in the SRDSA’s debt perimeter), users should enter zeros for that item. In the special case where coverage is at the narrowest perimeter (central government without consolidated social security funds), the table would be entirely composed of zeros as there are no claims to consolidate. ### Issues obscured by consolidation — disclosure and commentary - Users should describe any important issues that are obscured by consolidation in the debt coverage commentary. - SRDSF users should explore issues posed by consolidation and report relevant findings in the commentary accompanying the standardized reporting on Debt Coverage and Disclosures. - Potential interpretations when consolidated debt is much lower than non-consolidated debt: - Mitigating factor: cross-holdings reflect liquidity/sound asset management purposes (financially sound borrowing units). - Source of risk: cross-holdings issued by financially weak government units, or cross-holdings that contribute to policy imbalances (e.g., shoring up a public entity’s balance sheet with nonmarket public debt issuance without addressing underlying problems). ### Debt structure: standardized charts and risk interpretation - The SRDSF’s standardized reporting includes charts to illustrate vulnerabilities arising from the debt structure. Users must populate templates with underlying data and interpret the charts. Key chart themes and guidance: - Currency composition: - Populated with historical data on debt by currency entered by the user. - Projections are generated automatically by the template based on user-entered debt issuance assumptions for the baseline scenario. - A higher share of foreign currency denominated debt implies greater exchange rate risk and risk of sudden depreciations that can lead to debt spikes. - When denomination is spread across various currencies with diverging movements, valuation effects on debt-to-GDP ratios may be ambiguous and warrant commentary. - Debt holder profile: - Populate using data from the IMF’s Sovereign Debt Investor Base Datasets (updated semiannually by the Statistics Department). - Data allocate public debt among five creditor groups: (i) domestic central bank; (ii) domestic commercial banks; (iii) other domestic creditors; (iv) foreign official creditors (including regional central banks); and (v) foreign private creditors. - Risks are generally higher with greater dependence on foreign private creditors, especially when holdings have been volatile or risen rapidly (hot money flows). - Complements the GFN Module which quantifies the impact of financing shocks from riskier creditor groups. - If datasets lack breakdowns by holder, users should estimate allocations following Arslanalp-Tsuda methodology using best available information (allocate by residency, allocate domestic claims between central bank and commercial banks using balance sheet claims, residual to other domestic creditors; foreign official claims as sum of bilateral and multilateral; foreign private as external less foreign official). - Governing law: - Classify public debt by governing law (local law, foreign law, or multilateral debt) using the most recently available year. - Guiding principles: (i) debt issued in local markets considered subject to local law unless indenture specifies otherwise; (ii) external debt owed to bilateral creditors considered subject to foreign law; (iii) debt issued in international markets usually subject to foreign law, to be confirmed via prospectuses and documentation. - Governing law does not by itself constitute risk but may affect modalities and costs of debt restructurings. - Marketability: - Classify past issuance as marketable or nonmarketable; projections populated automatically from issuance assumptions. - Marketable instruments: debt securities such as bonds readily bought and sold in markets. - Marketability has no inherent risk implication but informs focus: marketable-dominated debt emphasizes rollover risks; nonmarketable-dominated cases emphasize solvency risks and debt service profile. - Maturity: - Enter historical data on debt by remaining maturity over the past five years. - Definitions: short-term = matures within a year; medium-term = 2–5 years; long-term = beyond five years. - Longer horizons typically imply lower rollover risk. - When detailed breakdowns are only available for a narrower perimeter than the SRDSA perimeter: - Users may populate charts with narrower-perimeter information but must explain this in the note beneath each chart. - Relevant data gaps should be noted in the commentary box of the standardized report. ### Baseline scenario for public debt and GFNs (framework inputs and outputs) - SRDSF tools produce projections based on data, projections, and financing assumptions entered by the user. - Required inputs for SRDSF debt reporting and core modules (applicable for all countries) include all macroeconomic, fiscal, and financial data, including 10-year projections, that drive debt-to-GDP and GFN-to-GDP evolution. These inputs include: - growth, inflation, interest, maturity, exchange rates, currency composition, the primary (noninterest) fiscal balance, and other transactions that create or extinguish public debt. - All SRDSAs include a standardized table showing baseline debt-to-GDP and GFN-to-GDP ratios and their main drivers over the framework’s 10-year horizon; the table is auto-populated after projections are entered and provides essential baseline information before applying key risk tools. - Baseline projections arise from user-entered data and projections; they should reflect users’ expectations of likely outcomes and are unconstrained ex-ante so long as realistic. - When constructing longer-term scenarios (years 6–10), users may extrapolate key debt drivers at the end of the medium-term horizon (e.g., year 5). The baseline also reflects projections for new debt stocks, interest, and amortization computed by the SRDSF template based on user-entered new issuance assumptions. - Decomposition of changes in public debt-to-GDP ratio reveals contributions from: - (i) the primary deficit (measure of fiscal effort); - (ii) automatic debt dynamics (real interest-growth differential and exchange rate movements); and - (iii) other factors including interest revenues, contingent liabilities materialization when recognized, arrears clearance, and asset transactions. - GFN-to-GDP projection is complemented by a projection of its debt service component, decomposed into local and foreign currency flows. ### Standard commentary expectations - Users should accompany standard reporting with commentary flagging relevant aspects of the baseline, including: - Main themes underpinning the baseline scenario, economic environment, and relevant policies. - Description of baseline debt and GFN trajectories. - If a debt or GFN trajectory is strongly upward or downward, indicate which drivers are responsible. - Explain any irregular movements in the debt path (jumps or drops). ### Debt dynamics and drivers (Box 3 — analytical formulation) - Public debt evolution relates current debt to previous year’s debt, interest bill, primary balance, stock-flow adjustments, and exchange rate revaluations for foreign currency debt. - Key symbolic definitions (as used in Box 3): - t, t-1: Time, t=current year, t-1=last year - D: Public debt - I: Interest - i: Implicit average interest rate - e: Nominal exchange rate - 1+ε: Nominal exchange rate change (et/et-1) - g: Real GDP growth - π: Inflation (GDP deflator) - ρ: 1+nominal GDP growth ((1+g)(1+π)) - PB: Primary balance - SFA: Stock-flow adjustment (all other effects) - Y: Nominal GDP (level) - Superscript f: Related to foreign currency debt or inflation - Superscript d: Related to domestic currency debt - Lower case: Variable expressed as a percent of GDP - Core debt identity and dynamic decomposition (as presented): - Dt = (et/et−1) Dt−1f + Dt−1d + It − PBt + SFA t - Expressed in ratios to GDP and rearranged to show changes: - Δdt = εt dt−1f/ρt + (it − [ρt − 1]) dt−1/ρt − pbt + sfa t - Conversion to real terms uses definitions: - Real exchange rate z where 1+z = (et/et−1)(1+πtf)/(1+πtd) - Real effective interest rate r where (1+rt) = (1+it)/(1+πt) - The (it − [ρt −1]) term factors to (rt − gt)/(1+gt) after substitution, linking interest-growth differentials to real interest and real growth. *Source: SRDSF GUIDANCE NOTE (excerpts).* ### Box 3. Debt Dynamics and Drivers in the SRDSF (concluded) ### Box 3. Debt Dynamics and Drivers in the SRDSF (concluded) ### Debt dynamics equation and debt-stabilizing primary balance - Final SRDSF equation (as presented) for change in debt Δdt: - Δdt = zt dt−1f (1+gt)(1+πtf) + rt − gt 1+gt dt−1 + πtd − πtf (1+πtf) ρt dt−1f − pbt + sfa t - Purple term: real exchange rate component. - Blue term: real growth–interest differential. - Red term: relative inflation component. - Debt-stabilizing primary balance (setting Δd = 0, assuming no change in real exchange rate and no stock-flow adjustments): - pbt = πtd − πtf (1+πtf) ρt dt−1f + rt − gt 1+gt dt−1 - Alternative expression: pbt = dt−1 1+gt (πtd − πtf 1+πtd αtf + rt − gt) - Where αtf corresponds to the foreign currency share in public debt. - Use: - This formula is used to calculate the probability of debt non-stabilization, with the time period corresponding to the average of these variables along the trajectory. ### Realism tools: purpose and general guidance - Purpose: - Suite of tools to assess realism of the baseline macroeconomic scenario and guard against excessively optimistic projections. - Nine realism tools scrutinize key drivers of public debt using cross-country and historical performance. - Tools are automatically produced after projections are entered into the template and included as standard outputs of any SRDSA. - Users encouraged to interact early to allow revision if significant issues arise. - When realism flags appear: - SRDSA should explain why they do not signal overoptimism using commentary fields. - Strong justification needed when warnings span multiple tools or show large deviations from benchmarks. - Examples when flags can be discounted: - Aggressive fiscal adjustment recently completed or legislated with high probability of execution and yield. - Documented track record of divergence from cross-country norms, with explanation. - Severe exogenous events (e.g., COVID-19) that distort historical and cross-country norms; justify why flags are less relevant. ### Realism Tool 1 — Forecast track record of debt drivers - Examines forecast errors for: primary deficit, real interest−growth differential, exchange rate depreciation, stock-flow adjustments, and public debt at 1-, 3-, and 5-year horizons. - Comparator groups: six options based on country’s relationship with the Fund, income level, and (for emerging markets) export earnings. - Color scale: dark green = below the 25th percentile (pessimism) to bright red = above the 75th percentile (optimism). - Interpretation: - Many red cells indicate persistent forecast optimism and signal risk, especially if projected debt trajectory improves dramatically. - Notes: - For real interest rate differential errors, consider contributions of nominal interest and inflation. ### Realism Tool 2 — Output gap revisions - Assesses historical revisions and bias in real-time output gap and 3- and 5-year ahead projections. - Based on Kangur et al. (2019) and staff analyses. - Uses October and April WEO vintages; for most countries October WEO proxies real-time projections. - Color scale: green (<25th percentile) to red (>75th percentile). Red indicates negative bias in output gap projections and raises a realism flag. ### Realism Tool 3 — Comparison of projected variables with recent realizations (debt drivers decomposition) - Graphically reports cumulative contributions of key debt drivers (primary balance, contributions from growth, interest, exchange rate, and other factors) over past 5 years and next 5 years. - Users should flag large shifts in the pattern of contributions (e.g., large drop in contribution from r − g) unless adequately justified. ### Realism Tool 4 — Distribution of debt-to-GDP ratio reductions - Compares projected 3-year change in debt-to-GDP with distribution for all market access countries (October 2020 WEO data, 1990–2019). - Large projected debt reduction defined as > 75th percentile of sample flags potential over-optimism. - Also compares projection to country’s maximum observed 3-year reduction during 1990–2019; exceeding domestic max is a realism concern. - Figure references: - 3-year debt reduction above 75th percentile: 5.9 ppts of GDP (threshold indicated in figure). ### Realism Tool 5 — Fiscal adjustments (cyclically adjusted primary balance) - Shows distribution of 3-year changes in cyclically adjusted primary balance (October 2020 WEO, 1990–2019) and country’s maximum historical adjustment. - If country projections unavailable, tool computes cyclically adjusted primary balance using output gap (Escolano 2010) or HP filter where teams do not calculate output gaps. - Flags: - Projected adjustment > 75th percentile of cross-country database. - Projected adjustment > country’s historical maximum. - Figure references: - 3-year adjustment above 75th percentile: 2 ppts of GDP (threshold indicated in figure). ### Realism Tool 6 — REER gap realism - Based on user-provided estimate of initial REER misalignment; extrapolates REER over/undervaluation gap using baseline REER projections, assuming no change in equilibrium REER. - Flag: - Initial misalignment that is not unwound (gap exceeds ±5 percent) by end of 5-year horizon signals potential realism concern. - Confidentiality: - Included in policy notes but not in staff reports. - Note: - If teams do not project REER, proxy with change in nominal exchange rate against the dollar. ### Realism Tool 7 — Real GDP growth realism - Compares baseline real GDP growth with potential growth, output gap, and historical average (10-year average). - Signs of optimism: - Output gap without fiscal stimulus is positive at end of projection period. - Significant increase in real growth relative to historical average. ### Realism Tool 8 — Consistency between fiscal adjustment and growth (fiscal multipliers) - Compares impact of planned fiscal adjustment on growth using multipliers 0.5, 1 and 1.5 and annual AR(1) persistence of 0.6 with the baseline growth path. - Flag: - Large discrepancies (e.g., growth pickup during consolidation) that are inconsistent with plausible multiplier/persistence assumptions. ### Realism Tool 9 — Financing terms and issuance composition - Scrutinizes assumptions on new private borrowing and financing terms. - Shows composition of issuances by maturity (long term, medium term, short term) versus historical average. - Displays projected spread against 10-year US Treasury yields and spread implied by Laubach rule. - Flags: - Rapid shift toward long-term financing suggesting understated gross financing needs. - Projected spread compression that greatly exceeds the reduction implied by Laubach rule. - Laubach rule: - Bond spreads increase linearly by about 4 bps in response to a 1 ppt increase in projected debt-to-GDP ratio (Laubach (2009)). *Source: Box 3. Debt Dynamics and Drivers in the SRDSF (concluded), SRDSF GUIDANCE NOTE, INTERNATIONAL MONETARY FUND* ### SECTION V. NEAR-TERM RISK ASSESSMENT ### SECTION V. NEAR-TERM RISK ASSESSMENT ### Logit Model to Predict Sovereign Stress Events — Standard Application - The SRDSF’s multivariate logistic (logit) regression model is the standardized near-term risk analysis tool. - Key output: Logit Stress Probability (LSP) — the fitted probability that a stress event will materialize within 1-2 years. - The model uses explanatory variables organized in four categories based on channels affecting sovereign stress: structural characteristics, cyclical position, debt burden and buffers, and global conditions. - The LSP is calculated from realized values of the explanatory variables (no projections). All required variables are available from a centralized SRDSF database and the SRDSF template computes the LSP automatically. Users may manually update variables with newer statistical releases. For the institutional quality indicator, users should extrapolate the last available observation when current-year data are unavailable. For VIX, users should input the year-to-date change even when other variables correspond to the previous year. - The near-term assessment is not meaningful when the country is already in stress; SRDSAs should not report the LSP, the mechanical signal, or standardized near-term reporting if any Table 2 stress criteria are satisfied. ### Explanatory Variables (summary of Table 1) - Structural factors - Institutional quality: Average of government effectiveness and regulatory quality components of the World Governance Indicators. Intuition: stronger institutions → lower probability of stress. - Stress history: If a country is in stress, previous observation + 1. If not in stress, 0.9 × previous year's observation. Intuition: recent stress episodes raise probability of renewed stress. - Cyclical position - Current account/GDP: Current account/GDP × 100. Intuition: weaker current accounts may signal overheating prone to reversal. - Three-year change in REER: [REER(t)/REER(t-3)-1] × 100. Intuition: strong appreciation can raise risk of abrupt depreciation that spikes FX debt. - Credit-to-GDP gap, lagged (if positive): Cyclical component from a one-sided HP filter on credit-to-GDP ratios with smoothing parameter of 400,000 if positive (zero otherwise). Credit-to-GDP = private credit/GDP × 100. Intuition: positive gaps suggest excesses that may yield contingent fiscal liabilities. - Debt burden and buffers - Change in debt-to-GDP ratio: [Total Public Debt(t)/GDP(t) - Total Public Debt(t-1)/GDP(t-1)] × 100. Intuition: sudden spikes are difficult to manage and result in stress. - Public debt/revenues: [Total Public Debt(t)/Total Revenues(t)] × 100. Intuition: more revenues to service debt → less stress likelihood. - FX public debt/GDP: [Forex Debt(t)/GDP(t)] × 100. Intuition: higher FX debt → higher vulnerability. - International reserves/GDP: [Gross International Reserves(t)/GDP(t)] × 100. Intuition: higher buffers reduce stress risk. - Global conditions - Change in VIX: Year-to-year level change in VIX, with VIX indexed to 2010 = 100. Intuition: weaker global market sentiment raises stress probability. - Alternate specification (currency union members): Number of countries in stress divided by number of countries in currency union. Intuition: contagion risk within currency unions. ### Criteria for Determining the Existence of Sovereign Stress (summary of Table 2) - Large IMF-supported programs and exceptional financing from other IFIs and donors - IMF-supported programs: nonprecautionary programs with access > 100 percent of quota and positive disbursement in first year of program; if positive disbursements occur in later years, the country is still in stress. - Other IFIs: arrangements > 5 percent of GDP and positive disbursements in year. - Donors: exceptional disbursement > 5 percent of external debt. - Default - External arrears ≥ 5 percent of public external debt and increasing at least 10 percent in nominal terms; or defaults on domestic debt instruments. - Debt restructurings - Renegotiations of repayment terms on outstanding debt instruments (not liability management ops). - Chronic excessive inflation (including hyperinflation) - (i) Doubling of inflation rate compared to the previous year and inflation above 25 percent; or (ii) any event with inflation above 100 percent. - Market indicators / Loss of market access - Advanced economies: (i) spreads ≥ 1.5 standard deviations above 10-year mean and above 150 bps; or (ii) spreads above 500 bps. - Emerging markets: (i) Doubling of EMBIG spreads relative to year before and spreads ≥ 500 bps; or (ii) if EMBIG spreads unavailable, doubling of real domestic interest rate relative to year before and real interest rate ≥ 10 pct. - Loss of market access defined as an inability to issue debt in markets when there is a financing need. - Financial repression indicators (any one): (i) Central bank claims on government: ≥ 4 pct of GDP and growth of 100 pct (y/y); (ii) Commercial bank claims on govt: ≥ 9.1 pct of GDP and growth of 100 pct (y/y); (iii) T-bill rate: y/y change > 4.5 percentage points (if rate < 11 percent) or y/y change > 50 percentage points (if rate ≥ 11 percent); or (iv) other reports derived through MCM TA reports or FSAPs. ### LSP Interpretation, Outputs, and Mechanical Signal - Mechanical signal thresholds (Annex IV.A): - Low risk if LSP is below 6.3 percent. - High risk if LSP is above 19.5 percent. - Moderate risk otherwise. - Additional outputs generated by the template: - Level of the LSP and its evolution in recent years compared to a chosen comparator group (advanced economies, commodity-exporting emerging markets, or non-commodity-exporting emerging markets). - Contributions to change in LSP by major variable category. - Probabilities of erroneous predictions associated with the LSP level. - Commentary: users should summarize points of interest in the standard report and disclose any judgment-based final signals with rationale. - The LSP is an early warning tool for the likelihood of near-term sovereign stress (horizon: "1-2 years ahead") and does not indicate whether stress could be resolved by policy/new financing or whether debt is sustainable. ### Considerations for Special Cases and Adjustments - International reserves adjustment for large financial assets: - When a country has large financial assets (e.g., sovereign wealth funds) that are reserve-like, readily available, liquid at fair value, and under government control, it may be appropriate to augment international reserves with these assets for the near-term assessment. - Users must ensure these assets can be used without: (i) creating conflicts with fiscal rules; (ii) violating regulations, laws, or conventions (for SWFs, e.g., usage should not entail major balance sheet effects for the fund); and (iii) clashing with other encumbrances (e.g., assets used as collateral). - Adjustments should be applied consistently for all years in standardized reporting to prevent measurement-induced LSP dynamics. Recalculate LSP using the augmented reserve metric. - Note: During global stress, some asset classes may be difficult to liquidate. - Additional note: reflect on the maturity of swap-related liabilities before adding corresponding assets to international reserves; when adding swap-related liabilities to the DSA, unused swap-related assets should be added to international reserves for the near-term tool. - Credit gap discontinuities: - If credit-to-GDP ratios exhibit unusual jumps not reflective of financial sector risk (due to financial inclusion policies, entry/exit of banks, M&A, regulatory changes, or data anomalies), users may adjust credit-to-GDP ratios (e.g., based on past trends). - Any adjustment must be clearly explained in the commentary box of the standardized reporting, including methodology and valid reasons. - Currency union members: - An alternate logit specification includes the share of currency-union members currently in stress. Use this if contagion risks from the region are assessed as high. Do not use if the country is not in a currency union or no union member is in stress. ### Using the Near-Term Assessment to Inform Policy - Users should examine contributions of variable groups to the LSP change and assess whether results are intuitive or driven by a single bucket; unintuitive, single-variable-driven results may justify judgment. - Policy implications vary by contributing variable group: - Structural characteristics: pursue structural and governance reforms to build debt carrying capacity, recognizing these measures typically take time to affect near-term risks. - Cyclical position: consider leaning against the wind via macroprudential measures, fiscal adjustment, and policies to avoid exchange rate misalignment. - Debt burden and buffers: mitigate risks through fiscal adjustment (including revenue mobilization) and debt management (currency, maturity, and terms). - Global risk appetite: recalibrate the balance between local and international debt issuance, weighing tradeoffs including cost of debt and potential crowding out of domestic investment. ### Use of Judgment in Final Assessment - Judgment may be applied when the model does not fully capture near-term risk drivers. Potential reasons for applying judgment include: - Proximity to a stress event: high-frequency indicators may indicate imminent breach of stress event triggers; in such cases, final assessment should be high risk regardless of the mechanical signal. - Temporary distortions in explanatory variables from identifiable one-off events. - Permanent, country-specific characteristics (including WGI-based institutions index) that make regressors structurally different; evidence of systematic model misspecification (missed crises/false alarms) could justify judgmental adjustments. - Timing issues: credible corrective policies whose effects are not yet reflected in model variables may justify a judgmental upgrade if market reaction is favorable. - Factors outside the model: additional relevant information not captured by variables should be linked to specific factors and properly disclosed. *Source: SECTION V. NEAR-TERM RISK ASSESSMENT, ppea2022039* ### SECTION VI. MEDIUM-TERM RISK ASSESSMENT ### SECTION VI. MEDIUM-TERM RISK ASSESSMENT ### Debt Fanchart Module — purpose and methodology - Analyzes risks from the evolution of indebtedness over the medium term by simulating many debt trajectories using a debt dynamics equation and randomly drawn shocks to key variables. - Stochastic trajectories imply distributions of debt outcomes for each year of the projection horizon and are summarized by key percentiles and presented as a debt fanchart. - Standard simulation setting: 10,000 trajectories. ### Required data inputs and sampling method - Users must enter historical observations beginning in 2000 until the most recent year available for: - debt-to-GDP; - real effective interest rates; - real GDP growth; - the primary (noninterest) deficit; - real exchange rates; - domestic inflation; - foreign inflation (measured as inflation in the United States). - Users must also enter the share of debt in foreign currency—both history and projections. - The standard deviation of debt revisions (available centrally in the SRDSF common database) is required to calibrate uncertainty around the initial debt level. - Sampling approach: block bootstrapping by selecting blocks of historical realizations. For the standard time horizon of 6 observations (current year plus five subsequent years), three two-year blocks are drawn sequentially to generate one debt path between t and t+5 using the debt stock-flow equation. ### Preliminary historical fanchart and realism diagnostic - The preliminary historical fanchart is constructed entirely from realized historical data and serves as a realism diagnostic comparing the user’s baseline debt-to-GDP projections to projections implied by past behavior. - Possible diagnostic outcomes: - Baseline rises toward or exceeds the upper edge of the fan: indicates substantial weakening of debt ratios relative to past; users should confirm intuitive justification (e.g., looser policies) or revisit the baseline. - Baseline converges to or falls below the lower edge of the fan: suggests potential optimism bias. For SRDSF, a realism concern is triggered when the projected debt-to-GDP ratio falls below the 20th percentile of the fanchart in two or more years; users should consider revising the baseline or a realism adjustment will be activated. - Baseline resides within the middle section of the fanchart: no realism concerns. ### Final fanchart construction: centered and adjusted variants - Final baseline-centered fanchart: - Constructed when no realism issue is detected. - Calculated by adding de-meaned shocks (from the historical fanchart) to the user’s baseline; reflects past behavior via width and skew while being centered on the baseline. - Final adjusted fanchart (realism adjustment activated): - Triggered when realism diagnostic flags optimism (see above). - Procedure: - Compute the deviation of the baseline projection from historical trends (median of historical fanchart). - Compare the terminal-year deviation to a histogram of deviations for a relevant comparator group (advanced economies; commodity exporting emerging markets; non-commodity emerging markets) to find its percentile. - Shift the entire fanchart upward so the baseline corresponds to that percentile. The baseline projection itself is unchanged; the fan is shifted upward, implying higher simulated debt trajectories. - Realism adjustment generally raises simulated debt levels and increases probability of non-stabilization. ### Debt Fanchart Index (DFI) — components and interpretation - The final fanchart is summarized by three metrics combined in the Debt Fanchart Index (DFI): - Fanchart width: - Calculated as (terminal debt level at the 95th percentile) minus (terminal debt level at the 5th percentile). - Measures uncertainty around the baseline; invariant to the realism adjustment. - Probability of debt non-stabilization: - For each trajectory, a trajectory-specific debt stabilizing primary balance is computed using the Box 3 equation with the baseline drivers in the final year plus the trajectory’s average shock. - A trajectory is stabilizing if the projected primary balance exceeds the debt stabilizing level. - Probability of stabilization = number of stabilizing trajectories / total number of trajectories. - Probability of non-stabilization = 1 − probability of stabilization. Higher values indicate weaker prospects for stabilizing debt and higher risk. - Probability of non-stabilization is usually higher when the realism correction is activated. - Terminal debt level adjusted using an institutions index: - The median terminal-year debt level is multiplied by an institutional quality index (transformed so higher values indicate stronger governance and higher debt carrying capacity, thus lower risk). - The metric generally signals higher risk when the realism adjustment is activated because the adjusted fanchart has a higher median debt level. ### Considerations and special cases - Excluding years from the historical sample is permissible for: - Short or spliced time series: users may set the start of the historical sample to the first year with full observed values if extrapolation is impossible; splicing should be documented in commentary. - Clear structural breaks: years may be dropped after careful iterative analysis; avoid shortening the sample by more than 3–4 historical observations where possible. - Debt restructurings: years with deep debt operations (nominal haircuts) should be excluded to avoid implying future restructurings in simulated trajectories; exclusions should be documented. - COVID-19 recovery adjustment for SRDSAs prepared in 2022: - Preliminary historical fancharts should be centered around the baseline for the first two years of the forecast horizon in 2022 to avoid unwarranted realism corrections. The template implements this automatically; the option should be de-activated after 2022. - Exit clause to deactivate realism adjustment in rare circumstances: - May be used when the preliminary historical fanchart ceases to be a relevant diagnostic and strong arguments exist that the adjustment was erroneously activated, plus substantial confidence in baseline features producing benign debt dynamics. - Situations for consideration include recently concluded debt restructurings, major sudden and unplanned structural or policy framework changes, or abrupt political transitions producing major social and economic changes. - For SRDSAs prepared by Fund staff, de-activation must be agreed through the interdepartmental review process (or Management adjudication if needed) and rationale must be documented in commentary. - Special adjustments near or after debt restructurings: - Realism adjustment is usually inappropriate and should be deactivated; final fancharts should be centered on the baseline. - Volatility of the real interest rate post-restructuring may be scaled down by a factor corresponding to the ratio of new and past debt issuances if lower volatility is likely. - Liquid asset buffers: - Although the module uses gross debt only, large financial asset buffers materially affect solvency and should be considered. - If assets exceed 75 percent of GDP and 100 percent of public debt, the mechanical signal from the debt fanchart module should automatically be low regardless of the DFI’s level. Any such adjustment must be indicated in the standardized reporting commentary. *Source: ppea2022039 - SECTION VI. MEDIUM-TERM RISK ASSESSMENT (SRDSF GUIDANCE NOTE).* ### 69. The tool’s results should be presented based on several automatically produced ### 69. The tool’s results should be presented based on several automatically produced outputs that are included in the standardized SRDSA reporting: ### Automatically produced outputs - The module automatically produces the final fanchart for inclusion in the output, whether it is baseline-centered or adjusted. - The DFI consistent with the final fanchart produces a mechanical signal that is low risk for DFIs below 1.13 and high risk for DFIs above 2.08; otherwise the signal is moderate risk (Annex IV.B). - A graphical comparison of the components of the DFI to values observed in a relevant group of peers provides an additional point of reference for analysis. Users should select the group whose members illustrate characteristics most consistent with the country being analyzed and it should be consistent with the same group as the near-term assessment (described above) and GFN Module (described below). This information can complement the absolute levels of the metrics. ### Investigating the DFI’s components - Users should investigate the DFI’s components to see how the tool is detecting sovereign risks. The three underlying metrics are generally related, but each illustrates a distinct source of risk: - Fanchart width: A wide fanchart indicates substantial uncertainty around the baseline and the possibility of large projection errors. If they were to materialize, public debt could turn out to be much higher than envisaged, and therefore a source of potential vulnerability. - Probability of debt non-stabilization: A high probability that debt does not stabilize in the medium-term is a key warning that policies are not correctly configured to deliver fiscal and macroeconomic stability. - Terminal debt level, adjusted for institutions: If the debt level remains elevated at the end of the projection horizon, then it is likely to represent a significant burden and limit options to cushion shocks if they were to materialize. - In addition to scrutinizing these metrics individually, users should examine whether any DFI component points to a different finding than the others. Explaining such a divergence (including in the commentary box of the standardized reporting), would be important information toward understanding the robustness of the mechanical results. ### Policy advice informed by the DFI - Policy advice to contain sovereign stress risks can be informed by the DFI. When the tool signals an elevated level or risk, the next step is to understand what can be done to prevent stress from materializing. - Examples of how components point to policy actions: - Wide fanchart contributing to the risk signal: typically appropriate to build buffers which could limit the fallout from shocks that tend to be large in that country. - High probability of debt non-stabilization or elevated terminal debt level: points to a need to identify additional fiscal adjustment measures beyond those envisaged in the baseline. - High terminal debt level but favorable repayment structure (as indicated by the GFN Module) may mitigate risks. ### Using judgment in the medium-term assessment - There is no final signal from the fanchart module, but users should consider whether the output warrants use of judgment in the medium-term assessment. - The fanchart index feeds directly into the medium-term mechanical signal (section VI.D), together with the results of the GFN Module (see Section VI.B), which also produces a risk index. Any judgment is applied to the resulting medium-term signal that combines both the Debt Fanchart and the GFN Indexes. - Key considerations for when judgment may be warranted: - Large liquid asset buffers, which may imply that the fanchart causes the medium-term index to signal excessive risk, requiring a judgmentally determined final medium-term assessment. - One component of the DFI acting as an outlier and pushing the overall signal to another result that is fundamentally inconsistent with the risk profile signaled by the other two indicators. - Situations where the fanchart is wide, where policy adjustments are expected (with a high degree of confidence) to result in substantially less future volatility. - Expectations of structural shifts in the future that will fundamentally change the underlying behavior of the macroeconomic and fiscal debt drivers (examples: discovery/depletion of natural resources phased in over several years). - A history of repeated false alarms or missed crises as revealed through SRDSAs conducted in the recent past and developments consistent with lower risks. ### Gross Financing Needs (GFN) Module — standard application - The GFN Module is the SRDSF’s main tool for assessing liquidity risks at a medium-term horizon. The underlying analysis examines: - The size of a country’s financing needs (as defined in Box 4) over the medium term. - Debt holders and new financing instruments across various creditor groups to check whether creditor composition and debt structure are risky. - The domestic banking system’s capacity to act as a residual creditor in adverse conditions. - Box 4. Measurement of Gross Financing Needs (GFNs) - GFNs are calculated as the sum of the primary deficit, debt service (interest and amortization), and realization of explicit and implicit contingent liabilities, less any interest revenue. - Debt issuance refers to the amount of public debt placed with creditors to either meet the GFN or to finance other transactions, including financial asset accumulation/decumulation. Any purchase (sale) of liquid assets should be included as a transaction that raises (lowers) debt issuance. - Special accounting considerations: - Costs related to transactions supporting/taking over firms no longer going concerns should be recorded in the primary deficit as capital transfers/grants instead of contingent liability materializations. - Debt relief can give rise to entries under capital grants and should be recorded there instead of other transactions that decrease debt issuance. - Cash vs accrual basis: In cash-based systems, users should add arrears (both external and domestic) to GFNs to improve risk analysis; accrual basis reflects arrears when they occur. - Footnotes/definitions retained from source: - Explicit contingent liabilities: legal or contractual arrangements giving rise to conditional payment requirements. - Implicit contingent liabilities: not legal/contractual but recognized when a condition/event is realized (examples provided in source). - The SRDSF’s GFN definition consistently includes contingent liabilities realizations in both historical and projected observations in contrast to the MAC DSA. ### Assumptions on financing (critical inputs) - Assumptions on financing by instrument and debt holder in the baseline are critical inputs and warrant careful consideration because different financing structures or creditor bases can lead to drastically diverging liquidity risk profiles. - By instrument: All new debt issuance must be entered in the SRDSF template by instrument type. For each instrument, define: - (i) local or foreign currency and whether linked to an index; (ii) fixed/floating/zero-coupon interest; (iii) marketable or not; (iv) local or external market of issuance; (v) frequency of debt payments; (vi) grace period; (vii) duration until maturity; and (viii) borrowing entity. - Allocate all new debt issuance among these debt types, internalizing committed financing and communicated plans. When financing is less certain, enter assumptions reflecting the most likely outturn guided by past performance and market conditions. - By holder: For each instrument, attribute financing across five creditor classes: - (i) domestic central bank; (ii) domestic commercial banks; (iii) other domestic creditors, including non-bank financial institutions; (iv) external official creditors; (v) external private creditors. - Share of each creditor ranges from zero to 100 percent and can vary year to year. Use past debt holder shares in the template as guidance when holders are not obvious, but reflect relevant policy information and ensure consistency with creditor balance sheet sizes. ### Examples for debtholder share settings when holders are clear (table content summarized) - Suggested holder share parameterization (percent) by type of debt: - IMF credit: Official 100, Private 0, Central bank 0, Commercial banks 0, Other 0. - Other multilateral and bilateral loans: Official 100, Private 0, Central bank 0, Commercial banks 0, Other 0. - Central bank advances, overdrafts, recapitalization bonds: Official 0, Private 0, Central bank 100, Commercial banks 0, Other 0. - Syndicated international bank loans: Official 0, Private 0, Central bank 0, Commercial banks 100, Other 0. - Local bank loan financing: Official 0, Private 0, Central bank 0, Commercial banks 0, Other 100. - Bonds issued to non-consolidated social security fund: Official 0, Private 0, Central bank 0, Commercial banks 0, Other 100. ### Generalized stress scenario (centerpiece of module) - The module’s generalized stress scenario features macro-fiscal and debt holder shocks automatically implemented by the template. - Under default settings (Table 4), shocks are based on those recently observed in a country’s history. The scenario generally produces larger GFNs and incorporates a debt holder shock in which foreign private investors roll over only part of existing holdings for several years and do not acquire new debt issued by the government to meet any other financing needs. - Key elements and default settings: - Growth: Growth is reduced for two years by 1 standard deviation based on the last 10 years' outturns. - Interest rate: Nominal marginal effective interest rates rise by 300 bps upfront; shock phased out over 5 years. - Exchange rate: One-off depreciation equal to the maximum depreciation observed in last 10 years. For currency unions and inflexible exchange rate regimes, inflation rate decreased by one half of the largest decrease in inflation observed in the last 10 years. For flexible exchange rate regimes, combination of a slack effect and exchange rate passthrough effect with specified calibration. - Primary balance: Baseline noninterest revenue/GDP ratio and baseline nominal noninterest expenditures held constant. - Maturity shortening shock: In year of shock, debt issuance split 50 percent between short-term and long-term (unless baseline's maturity assumptions are shorter, in which case the baseline is carried over). Share of long-term financing rises gradually over 5-year interval. - Debt holder shock: Foreign private creditors rollover rate is 67% for two years, thereafter 100 percent. Banks absorb the shortfall and fully roll over their holdings. Other creditors' rollover rates do not change. Foreign private creditors provide no new financing other than for existing debt rollover. Creditor groups other than domestic commercial banks provide financing proportionate to their shares in the baseline. Commercial banks are the residual source of financing. - Note: 1/In cases where the baseline maturity assumption indicates more than 50 percent of issuance in the short-term, an adjustment is made to keep the stress scenario’s maturity assumption at least as unfavorable as in the baseline. ### GFN Financeability Index (GFI) components - The module automatically constructs the GFN Financeability Index (GFI), combining risks measured by three components available after financing assumptions and stress scenario convergence: - Average GFN-to-GDP ratio in the baseline: Larger financing needs indicate higher vulnerability to funding shocks. - Initial (current) bank exposures to the government: Elevated value implies higher risk because significant existing exposures may constrain further financing from the banks. - Change in bank claims on the government in a generalized stress scenario: Summarizes the potential demand on the banking system if stress materializes. ### Considerations for special cases (adjustments to GFN Module) - Adjustments are appropriate when sovereigns can count on key creditor groups to exhibit certain behaviors, provided such transactions have very high prospects of clearing requisite approvals and are corroborated by public announcements or firm commitments. - Examples of valid adjustments: - Quantitative easing by the central bank: Set the central bank’s holder shares to match announced government bond purchases/rollovers. For currency unions, adjust for the foreign official sector as regional central banks are treated as external creditors. Central bank and foreign official sector are not subjected to holder shocks; higher shares lower the amount of “risky” financing subjected to shocks. - Live precautionary Fund arrangements: Add the amount of resources the country could draw to the pool of available liquid assets. This reduces demands on the banking system and improves the related risk metric. - Nonmarket financing: Firm commitments on non-market debt disbursements (e.g., loans) sufficiently substantial to blunt funding shocks may be used to finance the GFN in the stress scenario according to the same terms/conditions as in the baseline (muting maturity shortening and interest rate shocks on this instrument). *SRDSF GUIDANCE NOTE, INTERNATIONAL MONETARY FUND* ### 78. Other mitigating adjustments are appropriate only in a limited set of countries that ### 78. Other mitigating adjustments are appropriate only in a limited set of countries that ### Mitigating adjustments (limited set of countries with special characteristics) - Revising the pool of available government liquid asset buffers: - If sovereign asset holdings are sufficiently large that there are remaining buffers even after accounting for the holder shock in the stress scenario, users may subtract the residual assets (expressed in percent of bank assets) from the change in bank claims on the government in the stress scenario. - After making this adjustment, the value of this risk metric as well as the GFN Financeability Index will decrease, pointing to lower risk. - Conditions for this adjustment: assets can be liquidated in a manner reflective of their fair value, are not encumbered, and are legally available for the government’s use. - Domestic non-banks as a mitigating factor: - Where the domestic non-bank financial system is significantly larger than the banking system, is a major holder of sovereign debt, and/or can be called upon to provide reliable financing (e.g., large retirement/pension funds, life insurance companies, or other mutual/bond funds), it may be reasonable to assume these institutions would step in as residual creditors along with domestic banks. - The tool can aggregate both sectors’ claims on the government and express them as a percentage of their combined assets; users can calculate the change in the stress scenario and use this indicator to replace the banking-system-only measure. - Reporting requirement when adjustments are used: - Accompanying commentary in the standardized output should note both the existence of the adjustment and the factors that led to determining that the relevant criteria were satisfied. ### Exceptional adjustments to reflect special country circumstances - Domestic non-banks as an aggravating factor: - If domestic non-bank public debt holdings are more volatile (for example, due to short-term investor strategies or funding structures), the standard stress scenario may be adjusted to treat them like foreign private investors. - When activated: reduces the rollover rates of private other investors temporarily and in parallel to private external creditors; this sector is assumed to provide no new financing in the shock scenario; financing demands are assigned to commercial banks, increasing demand on banks and signaling higher risk. - Changing the size of banking system assets in the stress scenario: - If there is clear information about illiquid banks, plans to resolve troubled assets, or reorganizations/liquidations, users may lower projected banking assets to the level expected to continue as going concerns. - For consistency, a parallel adjustment should be made to bank assets when measuring initial claims. - Modifying the magnitude of the stress scenario’s shocks: - Historically-drawn shocks may be distorted by extreme crises or recoveries; if users are highly confident such events will not repeat, they can scale down these macro-fiscal shocks. - Users should analyze implications (for example, a stress scenario debt ratio that rises well above the upper bound of the debt fanchart) before changing shocks. - In rare circumstances, when robust capital flow measures effectively lock in foreign private investors, users may consider adjusting rollover rates upwards in shocked years. - Revising the timing of shocks: - If users have knowledge that shocks will materialize in a year other than the standard setting (the year after the first projection year), they can shift the onset of shocks across the medium-term horizon to reflect political cycles, lumpy maturities, or generalized global conditions. - Governance and prudence: - These exceptional changes imply significant deviations from the tool’s underlying design or assumptions and need to be exceptionally well justified. - For SRDSAs prepared by Fund staff, such changes are subject to interdepartmental review and Management approval. - When assumptions are less certain, it is generally better not to alter the mechanics of the tool but to apply judgment when reflecting tool results in the overall medium-term assessment. - If changes are implemented, users should clearly explain the changes and how they impact interpretation of the tool in the accompanying commentary. ### Interpreting the Tool (standard reporting and indicators) - Standard reporting automatically produced when the GFN Module is completed: - A mechanical signal based on the GFI is reported in the overall summary table. - Signal thresholds: low risk for GFIs below 7.6 and high risk for GFIs above 17.9; otherwise, the signal is moderate risk. - Final signals (that incorporate judgment, if any) are only reported for each of the three horizons in the SRDSA; users should not report a final signal for this module. - Figures showing evolution of GFNs in the baseline and stress scenarios, financing provided by domestic banks, and comparisons of the three GFI components relative to a relevant comparator group are produced. - Factors to consider when interpreting GFN-related risks: - Trajectory of financing needs: - Persistently high GFN-to-GDP ratios are a sign of potential risk, especially if upward-trending. - Consider whether large GFNs are driven by fiscal balances and/or debt service to identify corrective measures. - Riskiness of the debtholder profile: - Foreign private investors are more likely to exhibit sudden loss of appetite, particularly if holdings were accumulated during capital inflow surges. - Official and domestic creditors are usually more stable and less likely to trigger financing shocks. - Capacity of banks to absorb government debt: - Heavily exposed banks may have limited space to increase exposure in face of shocks. - Frequency and magnitude of shocks: - Countries with high macroeconomic volatility face higher liquidity risks and require larger financing buffers. - Level of buffers: - Ample liquid asset buffers can mitigate risk; drawn-down assets increase risk unless buffers are restored. ### Policy advice linked to interpretation - Fiscal adjustment: - When the primary deficit is the primary cause of elevated GFNs, fiscal adjustment may be the appropriate lever. - Debt management: - Borrowing plans and local market development initiatives need to internalize tradeoffs between interest rates, currency composition, and duration. - Debt management can be effective for lengthening maturities, reducing rollover risks, and containing reliance on market segments that may saturate certain creditor groups. ### Using judgment (applying judgment to the medium-term final assessment) - Judgment should be applied at the medium-term final assessment level; the GFN Module does not combine mechanical signal and judgment into a final assessment. - Examples where metrics may be distorted and merit judgment: - One-year outliers can skew average GFN-to-GDP in baseline; lumpy GFNs may result from pre-financed debt repayments, large capital expenditures financed by committed loans, liability management operations, or non-market operations (e.g., arrears clearance, central bank recapitalization). - In Fund-supported programs, firm commitments obtained during financing assurances are a mitigating factor. - Conversely, a large upfront one-year GFN with uncertain financing indicates high liquidity risk. - Relatively high GFNs or bank exposure might be less concerning if a recent track record shows these levels did not result in stress, provided the trend is toward less risk and country-specific explanations justify manageability. - Banking system capacity caveat: - The module implicitly assumes banks are healthy, liquid, and willing to act as residual financier; this may not hold if banks are unwilling or strained. - Detailed information from an FSAP can inform capacity to absorb additional government debt. - Applicable regulations and exposure limits should be considered; changes in bank claims in the stress scenario should be assessed against exposure limits for feasibility. - Other factors affecting analysis: - Trade-offs between liquidity and solvency risks must be reflected when applying judgment (for example, central bank financing that lowers liquidity risk could increase future solvency risk if central bank capital turns negative). - External sector characteristics: capital flow measures may prevent holder shocks; membership in global government bond indexes may limit volatility of foreign private investor flows. - Additional financing from key development partners may be available to meet GFNs if shocks arise. - Nonfinancial assets are usually not an important mitigating factor as they are rarely quickly deployable, except in exceptional cases such as ongoing privatization programs conducted according to best practices. ### Triggered Stress Tests — common features and interpretation - Purpose: - Triggered stress-tests capture specific risks not fully covered by the fanchart and GFN tools for countries subject to extraordinary event risks that have not materialized recently. - Five scenario analyses: (i) banking sector instability; (ii) commodity price shocks; (iii) contingent liabilities due to narrow public debt coverage; (iv) corrections of misaligned exchange rates; and (v) natural disasters. - Activation: - Stress tests are “triggered” when a country meets certain relevance criteria; countries can trigger multiple tests or none. - Users may manually activate a test even if a trigger is not formally met. - Activation is not an early warning signal and does not by itself indicate the likelihood of a shock. - Mechanics and outputs: - All tests require a user-entered baseline with the shock layered on top. - There is a standard calibration for each test, adjustable by users for country-relevance. - The mechanical SRDSF tool simulates debt-to-GDP and GFN-to-GDP paths automatically; results are superimposed on standardized reporting for the two medium-term tools. - Example (illustrative): contingent liabilities can make a large difference to the debt fanchart but only create a temporary spike in GFN. - Use in judgment: - Triggered stress tests inform judgment but do not have their own mechanical signal. - Users should consider both the magnitude of shocks and their qualitative probability of materializing. - If shocks are perceived as more likely and would produce major impacts, judgment could be warranted for the final medium-term signal. - Specific directions on incorporating activated stress tests into assessments are provided elsewhere in the guidance. *SRDSF GUIDANCE NOTE — Excerpt as provided* ### 89. The banking crisis stress test aims to capture contingent liabilities stemming from the ### The banking crisis stress test aims to capture contingent liabilities stemming from the ### Banking crisis stress test — triggers - Triggering indicators (test activated when country realization of either indicator equals or exceeds threshold): - Credit-to-private-sector-to-GDP gap from one-sided HP filter (percent): 10 - Mispricing risk index (within country percentile rank): 67 ### Banking crisis stress test — calibration - First-round fiscal cost of bank crisis resolution (Percent of GDP): - Advanced economies: 6.8 - Emerging markets: 10.0 - Second-round effects: - Modeled as negative shock to annual real GDP growth and inflation (proxied by the GDP deflator) in two consecutive years, and a positive shock on interest rates in the first year. - Primary balance assumed to remain unchanged from its baseline value in percent of the lower nominal GDP, except for the direct first-round effect on primary expenditures. - Shock magnitudes and propagation details: - Magnitude of the shock to growth: one standard-deviation, calculated using its last 10 annual realizations. - Size of the shock on inflation: set at ¼ percentage point for every percentage point decrease in real GDP growth in the stress scenario. - Increase in interest rates: equals ¼ percentage point for every one percent of GDP worsening of the primary balance in the stress scenario. - Modeling assumption: - Shock modeled as an increase of primary expenditures on a cash basis (simplifying assumption). ### Banking crisis stress test — customization - Users should tailor standard shocks to country-specific circumstances. - If the banking system is very large, increase fiscal costs to reflect higher expense for stabilizing the system. - Do not generally revise shock parameters downward for small banking systems; instead, consider whether the scenario is realistic and minimize the role of the test in applying judgment per Section VI.D. - Additional factors to consider when interpreting or customizing: - Whether potential banks at risk are domestically- or foreign-owned. - Available first-line-of-defense alternatives to direct government intervention. - Existence of a specific bail-in mechanism. - Resolution strategies adopted in the past (e.g., unlimited deposit guarantees, open-ended liquidity support, repeated recapitalizations, regulatory forbearance). ### Banking crisis stress test — special considerations - Activate test even if not triggered automatically when Fund staff analysis suggests systemic financial risk is high. - Inclusion of a well-articulated view on systemic financial risk is required for Article IV staff reports (IMF 2022c). - Interpretation of results should consider: - Broader assessment of systemic risk. - Quality of financial supervision and adequacy of financial policies. - Existence and severity of real-financial feedback loops. - Spillovers to or from the non-bank financial sector. - Current overall policy mix and position in the financial cycle. - Disregard results if a banking crisis has already recently occurred and its costs are reflected in public debt-to-GDP. ### Commodity price stress test — scope and triggers - Applied to emerging market commodity exporters and commodity importers with sizable fuel subsidies. - Trigger: - Applies to emerging market and developing economies classified in the latest WEO as having fuel or nonfuel primary products as their main source of export earnings. - For commodity importers, users should indicate whether subsidies have been present within the past 5 years; if so and unless a reform rules out reemergence, an alternative test examining fiscal risks from a commodity price boom should be applied. ### Commodity price stress test — calibration and propagation - Shock: sudden one standard deviation change in the prices of fuel and non-fuel commodities; commodity price gap applied in second year and closes over 5 years. - For commodity exporters: - Real GDP growth reduced by 1.1 percentage points for each 10-percent contraction of commodity prices. - Fiscal revenues-to-GDP reduced by 1.4 percentage points for each 10-percent contraction of commodity prices. - Inflation (GDP deflator) reduced by the impact of price gap and commensurate to the share of commodity exports. - For commodity importers with sizable fuel subsidies: - Expenditures-to-GDP ratio increased by 0.9 percentage points for each 10-percentage point increase in fuel prices starting the second year. - If strong reasons indicate subsidies would instead reduce growth-enhancing investment, a negative growth shock may be imposed and the expenditure shock reduced (changes should be disclosed). - Interest rate response: - Interest rate premium increased by 25 basis points for each percentage point deterioration in the primary deficit. - Average deterioration from baseline in the second and third years of projection is taken to calculate the interest rate premium. - Timing: - Shocks to real GDP growth; revenues and GDP deflator (for exporters); and expenditures (for importers) start in the second year, with full impact in second and third year, gap converging to baseline in 5 years. - Interest rate shock applied starting in the second year and does not converge to baseline during the projection period. - Net exports: - Where exporters also import a commodity, a net export figure is applied to capture mitigating effects of decline in commodity imports. ### Commodity price stress test — customization and special considerations - Users can customize for stabilization funds, exchange rate regime, commodities, size of price shock, and responses of real GDP growth and fiscal revenues. - For SRDSAs prepared by Fund staff, adjustments should follow close discussion with country authorities and be validated in review. - If country produces significant multiple commodities, examine differing price cycles and expected medium-term production scaling. ### Contingent liability stress test — trigger and purpose - Illustrates risks of debt surprises for countries using a narrower debt perimeter than the general government. - Trigger: - Activated when user indicates (i) debt coverage is narrower than general government and (ii) there are units of the general government with financial activities separate from the central administration. ### Contingent liability stress test — calibration and parameters - Captures impact of liabilities of public entities excluded from the SRDSA debt perimeter. - Many variables pre-populated in template using key statistical publications; users may adjust defaults with better information. - Designed to capture implications of below-standard debt coverage via a one-time contingent liability materialization in the second year of the forecast horizon. - Standard and optional parameters (examples of data linkages and items to consider): - State and local governments: Central bank survey (IFS)*, Other depository corporations survey (IFS)*, Zero or user estimate - Central bank, Commercial banks, Others, Official, Private: various links to IFS and external debt databases where available - Unconsolidated social security funds: Zero or user estimate - Optional customized parameters typically excluded from default calibration: Public sector enterprises (SOEs), PPPs — may use financial statements where available and private databases on Eurobond issuances - Domestic creditors, External creditors: Zero or user estimate - (*Denotes item that can be linked to existing database for IMF staff.) - Customization: - Shock can be augmented to include risks from SOEs and PPPs; commentary should note scope and sources of customizations and that the test captures a broader set of risks than below-standard debt coverage alone. ### Contingent liability stress test — interpretation - Users should describe the likelihood that modeled contingent liabilities will materialize for the sovereign. - Users should describe efforts by authorities to boost comprehensiveness of public debt coverage if not included elsewhere. ### Exchange rate shock — trigger and calibration - Trigger: - Applied when user-entered estimate of high initial overvaluation of REER is defined as above 5 percent and changes in REER over the medium-term are insufficient to eliminate overvaluation. - For SRDSAs prepared by Fund staff, estimated overvaluation should be consistent with the External Sector Assessment. - Scenario not applied if no large initial overvaluation or if corrected over the medium-term. - Calibration: - Scenario centered on a depreciation shock sufficient to close the country’s over-valuation gap during the projection horizon. - Floating regimes: shock fully applied to nominal exchange rate against the dollar in the second year; additional impact on deflator to mimic passthrough. - Pass-through magnitudes: emerging market economies: 25 basis points per 1 percentage point of depreciation; advanced economies: 3 basis points per 1 percentage point of depreciation. - Inflexible regimes (fixed rates, currency union members, no legal tender): internal devaluation shock equal to the depreciation shock is applied by lowering the GDP deflator in equal steps for the six years of the projection horizon. ### Exchange rate shock — customization and special considerations - For internal devaluations with expected rapid reserve loss, users may adjust shock distribution across years. - Test not designed to analyze risks of exchange rate over-shooting beyond unwinding of fundamental overvaluation. ### Natural disaster stress test — trigger and calibration - Trigger: - Applied to MAC countries meeting both: (i) two natural disaster events in a three-year window; and (ii) cumulative economic loss of at least 5 percent of GDP caused by those events. - Natural disaster events considered include climate-related, geophysical, hydrological, meteorological, biological, and extra-terrestrial events. - Information taken from EM-DAT database between 1980 and 2021. - Also applied to MAC Small States identified in IMF (2016). - Calibration: - Direct impact: one-off shock of 4.5 percentage points of GDP to public debt-to-GDP ratio in the second year of projection. - Interaction effect: real GDP growth lowered by 1.3 percentage points, with no subsequent rebound shock (implying some permanent output loss), introduced in the second year. - Customization: - Users can adjust parameters to capture mitigation policies including effects of catastrophe insurance. - Special considerations: - Not designed to capture structural and gradual impacts associated with physical and transition risks from climate change (these are analyzed using a long-term risk assessment tool). ### Medium-term index and aggregation - For the medium-term mechanical signal, results of the Debt Fanchart and GFN Modules are aggregated automatically in the SRDSF template. - Aggregation results in a medium-term index based on the DFI and GFI (Annex IV.D). *Source: IMF SRDSF GUIDANCE NOTE (excerpt).* ### 95. While there are no exceptions to the standard aggregation rule and thresholds, users ### ppea2022039 - Sections 95–110: Medium-term and Long-term Risk Assessment ### Interpreting the Standard Tool - The medium-term index mechanical signal in the SRDSA cover table: - Low risk for MTIs below 0.257. - High risk for MTIs above 0.395. - Otherwise, the signal is moderate risk. - Standardized reporting includes a figure showing the evolution of the medium-term index for the current observation and the past several years; commentary should typically indicate whether overall medium-term risks are rising or falling. - When the two medium-term tools diverge widely (one indicating high risk and the other low), the averaged medium-term index will typically fall in moderate risk. Users should: - Explore whether the result is intuitive and, if reasonable, document a careful explanation in the commentary box of the standardized reporting. - Otherwise, follow a judgment-based final signal as described under "Using Judgment." ### Using Judgment — Situations Warranting a Judgment-Based Final Signal - Strong presumption for a judgment-based final signal exists if: - A stress test is triggered, delivers a debt path above the 75th percentile of the debt fanchart, and users assess a high risk of that outcome materializing; typically a one-notch downgrade to the final signal is appropriate. - The medium-term mechanical signal is assessed as invalid due to considerations for judgment listed for either the Debt Fanchart or GFN Module (or both). - There is a wide divergence between Fanchart and GFN Module results; triggered stress tests may arbitrate, but users must add considerations about which tool is more reliable. - For SRDSAs prepared by Fund staff, expectations of a high risk of materializing should be consistent with a high likelihood of materialization in the Risk Assessment Matrix, when it includes a related risk. ### Medium-term reporting example (Figure 23) — Key numeric values and metrics - Sample country ("Ruritania") medium-term assessment outputs (values shown as in Figure 23): - Debt fanchart module: - Fanchart width (percent of GDP): 24.80.2 - Probability of debt non-stabilization (percent): 5.90.1 - Terminal debt-to-GDP x institutions index: 25.10.5 - Debt fanchart index (DFI): 0.8 - Risk signal: Low - GFN module: - Average baseline GFN (percent of GDP): 16.25.5 - Banks' claims on the gen. govt. (pct banks' assets): 7.42.4 - Chg. in banks' claims in stress (pct banks' assets): 0.70.2 - GFN financeability index (GFI): 8.2 - Risk signal: Moderate - Medium-term index components and outputs: - Weight on Debt fanchart index (normalized): 0.5 - Weight on GFN financeability index (normalized): 0.5 - Medium-term index value: 0.19 - Risk signal: Final assessment: Moderate - Probabilities reported: - Prob. of missed crisis, 2022-2027 (if stress not predicted): 9.1 pct - Prob. of false alarm, 2022-2027 (if stress predicted): 61.1 pct - Module thresholds (as stated in the figure notes): - DFI signal thresholds: low risk if the DFI is below 1.13; high risk if the DFI is above 2.08; otherwise moderate risk. - GFI signal thresholds: low risk if the GFI is below 7.6; high risk if the DFI is above 17.9; otherwise moderate risk. - Medium-term index signal thresholds: low risk if the GFI is below 0.26; high risk if the DFI is above 0.40; otherwise moderate risk. - Commentary example: the GFN Financeability Module pointed to a higher, but still moderate, level of risk; reinforced by potential contingent liabilities from government entities outside the central government and signs of strains in subnational governments' finances. ### Long-term Risk Assessment — Common Considerations (Section VII) - Purpose and nature: - Covers risks of debt-related stress that could materialize after the next five years (longer-run, more qualitative, higher uncertainty). - Required element in all Fund SRDSAs. - If long-term analysis reveals fiscal issues likely to have an impact at the 5-10 year horizon not reflected in SRDSF 10-year baseline, users should revise the 10-year baseline to reflect associated fiscal costs and macroeconomic impacts. - First-step guidance: - Use SRDSF extended (5-10 year) baseline projections for public debt and GFNs and describe level and direction of debt, GFNs, and debt-stabilizing primary balances. - In certain cases (e.g., debt restructuring), report a 10-year debt fanchart and emphasize probability of debt stabilization over the next decade. - Template can extrapolate debt-to-GDP and GFN-to-GDP paths based on terminal levels of the medium-term horizon or user-customized assumptions. - Optional long-term standardized modules (voluntary except where mandatory): - Four modules capture key risks relevant to many market-access countries: - Demographic change impacts on social security/pension funds and public health programs. - Effects of discovery or depletion of natural resource wealth. - Rollover risks from large future debt amortizations. - Consequences of adaptation and mitigation investments to combat climate change. - Use of modules is voluntary for Fund users unless: - (i) needed to comply with requirements for Resilience and Sustainability Facility (RSF) requests and augmentations; or - (ii) the country is highly exposed to natural disasters or undergoing debt restructuring, in which case the climate-change adaptation submodule is compulsory. - Assessments for the long-term horizon are more judgment based; none of the four long-term modules produces a mechanical signal. - Indicators of risk from modules: - Explosive simulated debt trajectory indicates risk. - Average or maximum GFN-to-GDP ratios exceeding medium-term levels signify potential vulnerability. - Whether risks are high or moderate depends on breadth of indications, confidence the risks will materialize, timeframe and availability of corrective actions, and country track record. ### Demographics Module — Overview - Purpose: capture long-term fiscal costs from demographic trends that impact pensions, social security, and public health programs. - Use when demographic trends likely impose significant funding pressures and heighten sovereign stress and debt sustainability risks. ### Pension/Retirement Benefit Sub-Module — Scope, inputs, steps, and outputs - Scope: - Generates long-term projections of financing needs from national pension schemes funded by the general government, including social security programs. - Simulates net pension expenditures and analyzes implications for long-term public debt and GFNs. - Criteria for mandatory use in SRDSF write-up: - Countries with significant current or future pension liabilities not already reflected in extended baseline projections should use the module. - Two conditions indicating materiality (either suffices): - Growth of pension expenditure over the period 2022-50 falls above the 75th percentile in the entire sample of countries. - Growth of old-age dependence over the period 2022-50 falls above the 75th percentile of the entire sample of countries. - Modeling steps (as described): - Generate annual gross pension expenditures as a share of GDP that increase with: - the old-age dependency ratio, - the coverage of the pension scheme(s), - the pension benefits per beneficiary; - and decrease with: - the number of people who are employed, - GDP per worker. - Mathematical expression (textual form preserved): - Benefits/GDP = Dependency Ratio ⋅ Benefits/Beneficiaries / GDPWorkers ⋅ Coverage Ratio ⋅ 1/(LFPR⋅(1−UR)) - Equivalent formulation provided in the source using Pop65/Pop15-64 and related terms. - Obtain annual net pension expenditure by netting pension contributions from gross expenditures; SRDSF assumes pension contributions remain constant as a share of GDP throughout the projection period (present until year 2100). - Annual financing needs equal shortfalls between pension system net asset value and net expenditures; if assets are sufficient, no financing need that year. - Default template assumes net assets grow annually at a risk-free rate. - Data sources and user inputs: - Pre-populated standardized data: dependency ratio, population by age group, pension coverage ratio, real GDP, labor force participation rate, unemployment rate. - User-entered inputs required for the first year of projection: - Total benefits paid by the pension system (in percent of GDP); - Total contributions to the system (in percent of GDP); - Assumption on the growth rate of contributions (either constant as a share of GDP, or grow at the same rate of GDP per worker); - Value of the pension system’s net assets (in percent of GDP). - Initial year of analysis choice ranges from 2020 to 2099; module computes projections for future years without preexisting data. - If user chooses a starting year (e.g., 2021), real GDP data are drawn from WEO up to 2026; for years after 2026, the submodule estimates GDP projections using the steady state rate of GDP growth. - Default discount and asset growth assumptions: - Real discount rate of 5 percent for NPV computation. - Real risk-free rate of 3 percent for pension assets growth. - Users may modify these assumptions. - Outputs: - Long-term projections for net pension expenditures, net pension asset balances, financing needs for the general government from present until year 2100. - Financing needs included with long-term projections to illustrate impact on debt and total GFNs. - Net present value (NPV) of entire stream of future financing needs from the pension system. - Permanent increase in primary balances required to offset them. - Additional guidance: - When available, present and discuss country authorities' projections alongside module outputs; explain significant differences and judgment on the more likely outcome. - If demographic measures are distorted (e.g., large migrant worker or refugee populations), evaluate tool reliability and consider not running the module. ### Healthcare Sub-module — Scope and use criteria - Scope: - Projects future financing needs arising from the healthcare system as a share of GDP; driven by demographic changes and other cost-increasing factors. - Use criteria: - Countries with sizable future public healthcare expenditures should use the sub-module and discuss risks in the SRDSF write-up. - Two criteria indicating potential long-term pressures (either suffices): - Growth of healthcare expenditure over the period 2022-2050 falls above the 75th percentile in the entire sample of countries. - Growth of old-age dependence over the period 2022-2050 falls above the 75th percentile in the entire sample of countries. *Source: IMF staff estimates and projections.* ### 111. The healthcare sub-module uses a model to generate projections for annual healthcare ### ppea2022039 - 111. The healthcare sub-module uses a model to generate projections for annual healthcare ### Healthcare sub-module: model and formula - Annual healthcare expenditure, as a share of GDP, is estimated from: - average health spending per population aged 20–64; - the relative share of healthcare expenditure for a particular age cohort (αi); - the distributions of the age cohorts in the population. - Mathematical expression (as presented in the source) links Health Expenditures20−64 / Population20−64 × Population20−64 / GDP with cohort-adjustment factors: - cohort-adjustment includes [1 + α0−19 · Population0−19 / Population20−64 + α65+ · Population65+ / Population20−64]. - Definition of αi: - αi = (Health Expendituresi / Populationi) / (Health Expenditures20−64 / Population20−64) - i denotes each age cohort. - Since α65+ is larger than 1 in most cases, healthcare spending-to-GDP also grows with the old-age dependency ratio. - An additional amount for excess growth of healthcare expenditure due to non-demographic factors is applied to the standard expenditure obtained from the formula. ### Inputs, user options, and default assumptions - Users need only choose the first year of projection by default; optional manual adjustments for the initial year include: - Health expenditure as a share of GDP; - Health expenditure per capita for the age cohort 0-19 as a share of health expenditure per capita for the age cohort 20-64 (i.e., α0-19). - Default module assumptions (users may modify): - real discount rate of 5 percent; - constant excess growth rate of healthcare costs of 0.6 percent for EMs and LIDCs; - constant excess growth rate of healthcare costs of 1.4 (percent implied for AEs). - Users with better country estimates can replace module outputs but should still discuss long-term fiscal risks in the SRDSF write-up. ### Outputs and presentation - Key outputs: - projection for the future path of healthcare expenditures from present until year 2100; - NPV of the future healthcare expenditures; - required fiscal adjustment for every year. - If country authorities produce their own estimates, those should be shown alongside the module’s standardized output with discussion of relative realism. ### Natural Resources Module — purpose and scope - Captures long-run mitigating factors or debt risks from scaling up/down of natural resource extraction and its impact on resource revenues, economic growth, and debt profile. - Relevant when future extraction volumes deviate from past volumes due to exhaustion, new discoveries, or political choices. - Does not analyze sovereign risks arising from future commodity price movements (module uses WEO price projections up to t+5 and a constant nominal inflation rate of 2 percent thereafter). - Valid only for countries producing naturally exhaustible commodities; not for secondary commodities or exhaustible but abundant commodities (e.g., agricultural commodities and fertilizers). - Tool calculates if extraction volumes over years t+6 to t+15 deviate by more than one standard deviation from the historical average (calculated over the past ten years); if criterion met, inclusion in SRDSA is strongly recommended. ### Natural Resources Module — data needs and modeling - Commodity-side inputs (up to three macro-significant commodities) may include: - size of proven commodity reserves; - commercial viability of reserves; - expected discoveries; - expected changes in extraction rates and domestic consumption; - expected capital investment expenditures; - share of commodity revenues to budget; - share of commodity-dependent sectors to GDP. - Macroeconomic non-commodity variables required: - government expenditures; - non-commodity revenue; - nominal non-commodity long-run GDP growth. - Module behavior: - produces cash flow projection associated with long-term increase/decline in commodity production and traces net effect on debt and GFNs; - GDP growth endogenously determined: direct and indirect commodity-dependent sectors grow at pace of commodity production; independent sectors follow exogenous stable growth; - projects primary balances-to-GDP over periods t+5 to t+10 and feeds these into long-term SRDSA baseline to calculate changes in GFN-to-GDP and debt-to-GDP paths. - Considerations and caveats: - accounts for leakages due to discounts for domestic consumption and private shareholders; - if SOE investment expenditures are not covered in general government, users should adjust shares of government revenues to resource exports to account for expected investment costs; - option to channel commodity revenues to a sovereign fund and simulate fiscal rules on transfers to government. - Users advised to run sensitivity analysis and provide caveats in commentary where uncertainty is significant. ### Large Debt Amortizations Module — purpose and applicability - Illustrates risks from abnormally large debt amortizations over the longer-term horizon, focusing on rollover risks and implications of refinancing/repaying obligations for long-term debt levels. - Particularly relevant when: - recent restructurings or large concessional financing will require future refinancing on less favorable terms; - sizable pickup in amortization beyond medium horizon, especially foreign-currency external amortizations; - countries qualifying for Resilience and Sustainability Trust analyses; - any country meeting an indicative criterion of debt amortizations 6 to 25 years ahead that significantly exceeds the 10-year historical average. - Module projects GFNs over a 25-year horizon. Users must input debt service on existing debt for this period (principal and interest, by currency composition). Payments on debt projected to be issued in medium-term and maturing in long-term are auto-calculated. ### Large Debt Amortizations Module — projection options and diagnostics - Three standardized assumptions for projecting remaining GFN-to-GDP components beyond medium term: - Option 1: constant t+5 values extrapolated for t+6 to t+25; - Option 2: uses t+5 values to compute a primary deficit that stabilizes the debt-to-GDP ratio from t+6 onward, with constant t+5 values for remaining variables (implying constant debt-to-GDP); - Option 3: historical 10-year averages used to project GFNs into the long term. - Users should examine individually: - (i) GFN-to-GDP ratio; - (ii) amortization-to-GDP ratio; - (iii) level of amortization. - Risk indication rule: - a variable gives a risk indication whenever it exceeds the 10-year historical average by more than one standard deviation in any year of the long-term projection; - two or more risk indications would strongly suggest including module findings in long-term SRDSA reporting. - Further scrutiny tools if risks appear: - recalculation of Medium-term GFN Financeability Index using average GFN-to-GDP over 25-year horizon; - realism indicator plotting maximum annual changes in GFN-to-GDP and Amortization-to-GDP against observed distribution between 2009 and 2019 for market access countries; - changes greater than the 75th percentile of sample changes suggest liquidity risk beyond medium term; - module reports average deposit accumulation levels required to meet projected principal repayments, as levels and as percentage of nominal GDP — users should assess feasibility of such buildup. ### Reporting guidance for Large Debt Amortizations Module - Reporting considerations: - module strength depends on realistic amortization and interest schedules for existing debt entered by users; - assess and compare the three projections against one another for context; - module outputs should serve as a starting point for further analysis and incorporation of any additional relevant considerations. ### Climate Change Module — structure and use requirements - Two sub-modules: - adaptation investments (building resistance to effects of climate change); - mitigation (efforts to reduce greenhouse gas emissions). - Both sub-modules project debt-to-GDP and GFN-to-GDP over a 30-year horizon under: - a standard scenario (default assumptions); - a customized scenario (user-adjusted, country-specific assumptions). - Use requirements: - Use of both submodules is required in context of RSF requests or augmentations and for pre-defined groups of countries where fiscal costs of adaptation or mitigation are expected to be significant; - adaptation module is compulsory in debt restructuring cases; - for all other MACs, submodules are optional; in some cases only one sub-module may be relevant. - Scenario mechanics: - baseline debt drivers extended from t+5 to t+30; fiscal costs of adaptation and mitigation added as deterministic shock from t+6 onwards (removing any adaptation or green investment already assumed from t+6 onwards); - customized scenario allows adjusting financing terms of climate-related investments, underlying primary balance assumptions, and long-term GDP growth path; - if t+5 debt drivers are too optimistic, a customized scenario with more conservative debt-driver projections should be considered for t+6 to t+30; - assumptions for customized scenario must be justified in the long-term risk analysis write-up. *Source: SRDSF Guidance Note (excerpts from the provided content unit).* ### 126. The key outputs from both sub-modules are extended projections for debt-to-GDP ### 126. The key outputs from both sub-modules are extended projections for debt-to-GDP and GFN-to-GDP ### Overview and purpose - Key outputs: extended projections for debt-to-GDP and GFN-to-GDP to inform long-term risk assessment. - Assessment character: qualitative given substantial uncertainty about future climate evolution and sovereign risk impacts. - User responsibility: after entering required data and assumptions, users should scrutinize resulting debt and GFN paths and come to an overall judgment, qualifying that judgment by the uncertainty around the level and range of potential climate investment. ### Interpreting module outputs and risk signals - Indicators that climate change may constitute an important risk: - Debt and GFN trajectories that are substantially higher and/or on an upward trend after factoring in investments in adaptation and mitigation. - Debt and GFN trajectories still upward trending in the customized scenario—or substantially higher over the extended 25-year period than in the first 5 years of the projections. - Indicators that climate change may be manageable: - Both the expected investment level and its uncertainty are small. - Where investment needs are both large and uncertain: - Interpretation is more difficult; users should carefully describe the basis for their assumptions (particularly investment needs). ### Scenarios and incorporation into SRDSF baselines - Both standard (default) and customized scenarios should inform user judgement: - Standard scenario: typically yields substantially higher debt and GFN trajectories. - Customized scenario: used to show how expected financing terms and policy reactions could mitigate debt-related risks. - Use in SRDSF baseline projections: - The two submodules should inform the SRDSF’s extended (5-10 year) baseline projections for public debt and gross financing needs. - Minimum requirement: the customized scenario of the adaptation sub-module must be incorporated in the 5–10-year baseline projection. - Incorporation of the customized scenario of the mitigation sub-module is optional because mitigation fiscal cost estimates are even more uncertain than those for adaptation. ### Limitations, evolution, and guidance - Techniques and data are expected to evolve; important data and modeling gaps exist for many countries. - Guidance is a starting point offering practical workarounds while better information and tools are developed. - Users should monitor trends in data availability and analytical methodologies and reflect these in assessments as appropriate. ### Adaptation sub-module — scope and mechanics - Purpose: capture fiscal cost of adaptation investment and main debt drivers over a 30-year horizon in a standard and a customized scenario. - Required inputs: - Estimates of expected capital expenditures (adaptation investment costs) for the years they are expected to be implemented. - Long-run assumptions on key debt drivers, including a separate underlying primary balance-to-GDP assumption beyond the user’s 5-year projection (by default held constant over the horizon and assuming no public adaptation investment beyond the 5-year projection period). - Standard scenario: - Prepopulated with IMF adaptation cost estimates from Aligishiev, Bellon and Massetti, 2022. - Extrapolates t+5 values of debt drivers over the remainder of the 30-year projection period. - Assumes growth remains constant at the t+5 level projected by the user over the following 25 years (implicit assumption: adaptation investment exactly cancels negative long-term impact of climate change on growth, except for impacts already in t+5). - Adaptation cost coverage: floods, storms, and sea level rise; does not capture droughts and heatwaves; may be less accurate for very small countries. - Customized scenario: - Users can modify adaptation investment costs, long-term growth projections, private sector participation, financing sources (public debt, grants, international cooperation), and other factors based on country-level information. - Encouraged use of DIGNAD-based analysis to supplement submodule estimates as DIGNAD is further developed. ### Adaptation sub-module — mandatory use and country coverage - Compulsory use: - For countries highly exposed to natural disasters, including: - (1) countries for which the natural disasters stress test is triggered; and - (2) countries at high risk from climate change, defined as the top quartile of an Adaptation Ranking Index (combines EM-DAT propensity to natural disasters; Aligishiev, Bellon and Massetti 2022 adaptation cost estimates; Composite Index using ND-GAIN, IMF-INFORM, and WRI). - Both sets of countries are generated and updated automatically. ### Adaptation investment proxies and data sources - Available regional and global estimates (Table 8: Regional Estimates of Public Adaptation Investment Needs, in percent of GDP per year): - World: 0.25 (Aligishiev, Bellon, Massetti, 2022) - Middle East and central Africa: 0.1 - 3.3 (Duenwald, et al, 2022) - Sub-Saharan Africa: 2-3 (IMF 2020a (regional outlook)) - Latin America and the Caribbean: 0.5 (ECLAC, 2014) - Asia and Pacific: 0.85 (ESCAP, 2019) - East and Northeast Asia: 3.8 (ESCAP, 2019) - Pacific SIDS: 1.5 (ESCAP, 2019) - Sources for country-level estimates (ordered from most to least reliable): - Individual country studies (e.g., joint IMF-World Bank CCPAs, Disaster Resilience Strategies, World Bank CCDRs, IMF CMAPs, regional development bank studies). - National Adaptation Plans and updated Nationally Determined Contributions (NDCs); caution: NDCs may mix adaptation and development costs and may over-estimate investment needs. - Proxies based on NDCs of similar MACs using IMF climate INFORM risk index comparators: - Median NDC adaptation investment cost proxies: 1.3 percent of GDP for relatively low climate risk (INFORM score between 2-4); 1.6 for medium climate risk (INFORM score between 4-6). - For very low climate risk MACs, the income group estimate for AE of 0.3% GDP per year could be implemented when some adaptation is assessed as needed. - Practical note: finding good estimates beyond Aligishiev, Bellon and Massetti, 2022 is currently a challenge. ### Mitigation sub-module — scope and applicability - Purpose: capture impact on debt sustainability of upfront investment needed to ensure a transition to a low carbon economy over a 30-year horizon. - Rationale: market shift to clean energy may require non-negligible upfront investment in green technology and infrastructure to complement carbon pricing and other measures. - Required use: - Required for all countries with an ambitious zero net carbon emission target (targeting zero net carbon emission before 2050). - Required for the 25 largest CO2 emitters per unit of output (highest mitigation needs per unit of output), even if they have not set emissions targets. - Standard and customized scenarios mirror adaptation submodule structure: - Standard scenario: prepopulated with public investment estimates for mitigation based on EU-reported data and t+5 debt driver values extended through t+30. - Estimates cover upfront investment needs for a 10 to 15-year period to meet zero net carbon by 2050, addressing carbon footprint of main sectors: Industry / Energy sector / Electricity generation; Transport / Mobility; Agriculture and land use; Buildings / Build-up environment. - Methodology to scale EU data to country-level: - Sectoral investment needs per 100 Mt CO2 in EU data are combined with country’s average CO2 emissions per unit of output (IMF Climate Change Indicators Dashboard using OECD IOT industries data). - Sectoral investment needs calculated by multiplying EU investment cost estimates by Dashboard’s average emissions in the country; aggregated into economy-wide proxy by averaging sectoral needs weighted by each sector’s share in total output. - illustrative benchmarks from EU data: - A EU7 country with average CO2 emissions of 150 Mt per unit of output would need upfront investment of about 1 percent of GDP per year (excluding buildings sector). - A EU9 country with average CO2 emissions of 360 Mt per unit of output would need close to 4 percent of GDP per year (excluding buildings sector). - Data caveats: - Investment estimates assume complementary measures similar to the EU (carbon market, regulatory measures like LULUCF). - Investment in buildings/build-up environment in EU data does not distinguish normal maintenance from additional green investment and was excluded in some aggregate proxy estimates. *Source: SRDSF Guidance Note (selected excerpts).* ### 140. Other considerations may also be relevant when judging potential investment needs, ### 140. Other considerations may also be relevant when judging potential investment needs ### Caveats in extrapolating European data - Energy industries have the highest CO2 emissions per unit of output in all countries, however, their share in total output in the EU is generally small—between 3 and 5 percent in EU countries—and therefore, in the EU they are tackled through market mechanisms such as the EU Emissions Trading System to reduce greenhouse gas emissions more cost-efficiently. - Advanced economies, such as the EU7, may already benefit from high infrastructure and network quality, and therefore, might have lower costs in sectors such as transport / mobility, or greening of buildings and the environment (as in Figure 28). - For most countries the green investments will take the form of imported capital goods, with low labor intensity. - Note: the total cost per 100 Mt of CO2 emissions in total economy is a weighted average, where the weights are the shares of each sector in total output. ### Default scenario, customized scenario, and Figure 29 estimates - The default settings use: - the extended baseline described in paragraph 2; - the estimated proxy on public investment needs per year from Figure 29; - the off-setting factors on the GDP growth and fiscal revenues described in paragraph 14. - Figure 29 shows upfront investment needs per year estimates for countries in the IMF Climate Change Indicators Dashboard with average emissions above 200 Mt by applying the same investment costs as in the EU9 per 100 Mt of CO2 emissions from fuel combustion, given the carbon intensity of their economic structure, and assuming a climate objective of reaching zero carbon net emissions by 2050 as in the EU. - Footnote: The Figure 29 estimates are obtained under the assumption that all upfront investment needs are publicly funded. Thus, in general, the proxy would give an upper bound for public investment needs per year. A larger private sector participation and /or a stronger assumption on the ability of carbon taxation to timely shift existing investment from high-carbon intensive sectors to low-carbon intensive sectors would lead to much lower estimates of public investment needs per year. - Example referenced: The G20 note on “Reaching net zero emissions“ prepared by IMF staff finds additional global public investment needs are 0.27 percent of annual global GDP. ### Calibration parameters for the customized scenario (users should consider) - a. Investment needs per year: - Consider different carbon intensity from the Climate Change Dashboard indicator and/or a different climate objective for reaching zero net carbon emissions (for example, a longer time horizon, which would lead to a smoother investment profile). - In a few countries, authorities have comprehensive studies on climate mitigation investment needs over the next 10 to 15 years that users could alternatively use. - b. The share of investment financed through government debt: - Consider financing mixes other than government debt (private sector participation, higher share of concessional lending / foreign aid) which would mitigate impact on debt ratio and GFNs. - The default scenario assumes fully government debt financing, hence gives an upper bound of potential impact on debt ratio and GFNs. - c. GDP effects and fiscal revenues: - The default scenario assumes short-term adverse output effects are offset by benefits of green investment, and these complement a policy mix in terms of carbon taxation/subsidies which is budgetary neutral. - Users can depart from this assumption using other tools (for example, the CPAT model developed by FAD) to adjust the GDP and primary balance path. ### Data guidance and limited-data procedures - When investment data is available, users should use reported investment data to customize defaults, including the share of privately financed investment. - For Fund SRDSAs where module is relevant (countries exposed to transition risks as in Figure 25, paragraph 138) but emissions data are insufficient, staff should discuss missing data in the long-term assessment and note what is needed to strengthen information collection so the module could be run in the future. - Note: For countries where the share of energy industries is higher than 5 percent of GDP (e.g., in oil-producing countries such Saudi Arabia, Russian Federation) the investment needs are underestimated (e.g., due to lack of data on needed complementary measures such as emission trading schemes, Land Use, Land-Use Change and Forestry (LULUCF) sector). ### Figure 29 context (investment cost per year for a 10-year horizon; climate objective reached by 2050) - Figure 29 covers: - Top 6 countries with CO2 Emissions from Fuel Combustion > 560 Mt - Countries with CO2 Emissions from Fuel Combustion > 200 Mt and < 750 Mt - Sample country labels shown include Saudi Arabia, China, P.R.: Mainland, India, Kazakhstan, Rep. of, Russian Federation, South Africa, and many others; aggregates such as "Total (excl. buildings)" are used in the figure. - Source: Fund staff calculations. ### Transition to assessments of sovereign stress and debt sustainability - The guidance continues into Section VIII on bottom-line assessments on risks of sovereign stress and debt sustainability (see paragraphs 142 onward). ### Key principles for overall risk of sovereign stress (paragraphs 142–143) - Overall signal should lie within the range of final signals at various horizons. - If a signal (e.g., moderate risk) appears most often across horizons, strong reasons are needed for a different overall assessment. - When weighing horizons, consider: - degree of confidence in results in each horizon (more uncertainty → less emphasis); - existence of time to take feasible corrective actions; - whether risks are lessening or growing over time, including if medium-term debt trajectory is expected to stabilize. - Communication should make clear sovereign stress is a broad concept and does not constitute a debt sustainability assessment. ### Debt sustainability definition and interpretation (paragraph 144) - Definition (IMF Executive Board): public debt sustainable when the primary balance needed to at least stabilize debt under both baseline and realistic shock scenarios is economically and politically feasible, such that the level of debt is consistent with an acceptably low rollover risk and with preserving potential growth at a satisfactory level. - Notes on interpretation: - Debt sustainability depends on solvency (debt stabilization) and liquidity (rollover risk). - Debt may be sustainable even if baseline primary balance does not stabilize debt, provided feasible adjustment measures can deliver debt stabilization with acceptable rollover risks. - Determining economic and political feasibility involves judgment, considering track record, cross-country comparisons, legal requirements, and prospects for ratifying sensitive adjustments. ### Probabilistic thresholds for sustainability (paragraph 145) - Any sustainable assessment should be associated with at least 50 percent probability. - "High probability" of sustainable debt: at least 80 percent probability. - Complementary cutoff: unsustainable debt event would occur with less than 20 percent probability. - These cutoffs balance avoiding unnecessary restructurings and avoiding mislabeling unsustainable situations as sustainable. ### Mechanical tools and combination for probabilistic sustainability assessments (paragraph 146) - Components used: - Sustainability logit model: calibrated on past episodes of unsustainable debt (defaults, restructurings, hyperinflations, large transfers) to estimate probability of unsustainable debt over a 1-4 year horizon. (Internal model available only to Fund staff.) - Debt fanchart index (DFI): quantifies prospects for medium-term debt stabilization. - GFN financeability index (GFI): metric of rollover risk; baseline must reflect all components of program financing (including prospective Fund disbursements) when assessing sustainability in program cases. - Signal on debt sustainability: - The fitted probability of unsustainable debt, the DFI, and the GFI are combined into a numerical sustainability index compared against calibrated thresholds to derive the mechanical sustainability assessment. - Thresholds calibrated on past instances of unsustainable debt are set such that: - sustainable assessments associated with at least 50 percent probability; - "sustainable with high probability" corresponds to probability of an unsustainable event of less than 20 percent. ### Role of staff judgment and review (paragraph 147) - Staff judgment, validated in review, is critical and may override mechanical results when appropriate. - Considerations when applying judgment: - Borderline results: evaluate sensibility and prospects for maintaining assessment if small shocks occur. - Conflicting results across tools: explore reasons and weigh credibility. - Distorted variables: temporary or manageable statistical/transactional distortions should not unduly affect assessment. - Omitted factors: include mitigating or aggravating conditions from stress tests and long-term modules. - If debt mechanically assessed as sustainable with high probability but a stress test/long-term module reveals vulnerabilities and the shock has high likelihood of materialization (e.g., 30-50 percent), revise to "sustainable but not with high probability." - Downgrade from "sustainable but not with high probability" to "unsustainable" would be rare and require demonstration that downside outcomes have more than a 50 percent probability of materializing. - A country’s track record: patterns of false alarms or missed crises may indicate factors not captured; verify no changes in underlying circumstances. - Clear signals of unmanageable public debt warrant an "unsustainable" assessment regardless of mechanical tools. Key indications include: - looming defaults in the near term due to lack of financing to make repayments, creditors’ unwillingness to roll over claims, and resistance/inability to implement fiscal adjustment or achieve urgent official financing; - announcements of an intention to restructure debt; - defaults or emergence of arrears constituting a default per SRDSF event definitions; - sudden, severe explosions of debt caused by economic crisis and not planned fiscal policies (>25 percent of GDP in one year). ### Requirements for DSAs under Fund arrangements (paragraph 148) - For DSAs prepared under Fund arrangements, a clear bottom-line debt sustainability assessment that incorporates mechanical analysis and any judgment is required to verify that prerequisite of sustainable debt has been met. - Disclosure rules vary by arrangement type: - For all Fund arrangements and emergency financing requests: precise aggregation method and index cutoffs are strictly confidential; staff should provide a concise descriptive statement synthesizing intuition from mechanical tools and judgment. - For arrangements and emergency financing requests with normal or no access: SRDSA statement should note whether debt is (i) sustainable with high probability; (ii) sustainable but not with high probability; or (iii) unsustainable. For publication, distinction between (i) and (ii) is removed—both published as "debt is assessed as sustainable." - For arrangements and emergency financing requests with exceptional access: the "high probability" component must be disclosed to the Board and retained in publication to justify application of the Exceptional Access Policy. *Source: Excerpt from SRDSF GUIDANCE NOTE (IMF). Note: Figures and staff calculations referenced in the text are from the original document.* ### 149. In certain circumstances, including arrangements treated as precautionary, ### 149. In certain circumstances, including arrangements treated as precautionary, ### When sustainability assessments must include a drawing (adverse) scenario - Additional analysis based on a scenario that assumes Fund resources are drawn (the drawing scenario) is required in the following three settings: - in exceptional access cases, where the sustainability assessment needs to be based on the drawing scenario; - if shocks that may trigger a drawing are not adequately captured by the medium-term (fanchart and GFN) modules. This situation would arise in cases where the shocks in the drawing scenario have not arisen at a comparable magnitude in the country’s historical sample used to run the tools; or - when review departments have doubts about the realism of the DSA baseline that cannot be resolved through discussions with the country team. This would typically provide additional information to Management when they are called upon to weigh in on the disagreement between departments. - If any of these criteria are applicable (including for precautionary SBAs as well as PLL arrangements) a debt sustainability assessment will need to be undertaken based on the adverse scenario that justifies the level of access. - The drawing scenario specifications and implications: - The scenario would entail a full draw of the Fund credit (treating the IMF’s claims as part of public debt and which may correspond to an overall increase in debt, where the effect of higher debt on the risk signals could be mitigated by any effect of higher buffers arising from the drawing). - The scenario would need to be sufficiently specified to serve as an input to the DSA. - This additional scrutiny will indicate whether debt is sustainable in the event downside shocks materialize and Fund credit is drawn. - There would generally be a high presumption of meaningful fiscal/debt impacts if adverse shocks arose that led to a drawing of Fund resources, including from a substantial change in the financing mix. ### SRDSAs in surveillance-only contexts (paragraph 150) - Sustainability assessments in SRDSAs prepared for surveillance-only countries are optional and are expected to be undertaken only in special circumstances. - Baseline projections in surveillance-only contexts reflect the policies that are most likely to be implemented; the baseline may not incorporate adjustment policies if none are planned. - Where the projected debt trajectory stabilizes and rollover risks are low with minimal or no adjustment, debt is sustainable. - In surveillance cases where an upward debt path and/or high rollover risks may occur and debt sustainability is ambiguous, staff may opt to perform additional analysis: - Developing an alternative scenario: - Should include reform measures representing the maximum possible effort. - Realism of successful implementation is central; users should employ a wide range of inputs to evaluate plausibility, including diagnostics in the realism tools, the experience of the country or relevant peers, the authorities’ commitments, and any relevant constraints. - Running the debt sustainability tools for the alternative scenario: - The same procedure for programs, outlined in paragraph 146, should be run for this scenario, as though it was a program baseline. This provides the three-way mechanical signal on debt sustainability. - Apply any judgment: - There is no presumption that the final assessment must align with the mechanical signal. The considerations outlined in paragraph 147 remain relevant and should be applied to the final assessment if appropriate and in a manner parallel to a program engagement with the Fund. - Prepare outputs: - Standard SRDSA reporting should solely focus on the (no adjustment) baseline. - Add a self-contained final element that includes: (i) a description of the scenario; (ii) a table showing the paths for debt, GFNs, and major economic variables; and (iii) the standard outputs for the Debt Fanchart and GFN Modules. - Fund staff should be aware of the requirements for deletions of certain SRDSA elements under the Transparency Policy (paragraph 154). - Practical cautions: - It is often difficult to identify what policies are feasible outside of a Fund-supported program; weighing benefits of a clear statement on sustainability against costs of errors is critical. - In high-vulnerability countries that remain current on obligations, have not definitively lost market access, and may soon request a program, a detailed unsustainable debt assessment could trigger loss of market access or premature disclosures of authorities’ adjustment policies. - Detailed optional assessments are expected to be performed only rarely and for strong reasons. - Where detailed disclosure would be counterproductive but a bottom-line assessment is appropriate, staff may add a more qualitative bottom-line assessment on debt sustainability in the accompanying SRDSA commentary. ### Use of SRDSF tools to inform debt restructuring (paragraph 151) - When debt is unsustainable and authorities have decided on a debt restructuring, the SRDSF tools should inform the determination of targets for debt relief. - Key practices: - Medium-term tools link naturally with the sustainability definition and do not require major modification to provide information about the magnitude of required debt relief. - It is generally appropriate to use a longer (10-year) horizon, which is common in restructurings. - This 10-year baseline should include the costs associated with essential adaptation investments to respond to climate change, as recorded in the customized version of the adaptation sub-module discussed in Section VII.E. - GFN targets, derived from the GFN module, serve as a convenient starting point to verify that the resulting financing needs after the restructuring are manageable, including under adverse circumstances. - A post-restructuring debt trajectory can be derived from the new debt structure that would attain the GFN targets; the debt fanchart module should verify an appropriately high probability of debt stabilization. - If the fanchart module does not indicate sustainability, the debt relief envelope should be adjusted until both the GFN module and the fanchart module signal sustainability. - Judgment and complementary targets to address specific country vulnerabilities should inform the targets when warranted. - Consideration should be given to any issues arising from the transition back to the standard SRDSF methodology once the restructuring concludes. - Role clarification: - Targets derived according to the SRDSF are in line with the Fund’s usual role in restructurings to define the needed envelope of debt relief to restore sustainability. - Specific restructuring decisions remain the responsibility of country authorities, in consultation with their legal and financial advisors. ### DSA requirements and publication (paragraphs 152–154) - SRDSAs are required under the following circumstances: - For surveillance cases, the SRDSA is a critical input to the Article IV consultation and should be included in the corresponding policy notes and staff reports; SRDSA would be updated at roughly an annual frequency. - All requests for IMF financing must be accompanied by an SRDSA. - Thereafter, for normal-access Fund arrangements, an SRDSA should be included annually, unless major changes to a country’s circumstances/outlook warrant an updated assessment. - When a program review is combined with an Article IV consultation, an updated SRDSA should be included regardless of the last SRDSA’s date. - In programs that involve exceptional access, the SRDSA needs to be updated at every program review to verify compliance with the exceptional access criteria. - Standardized SRDSA reporting elements (paragraph 153): - An overall summary consisting of a standard summary table and a chapeau summarizing key points of interest. - The summary table should report: (i) the mechanical and final risk signals at each horizon; (ii) the mechanical signals from the debt fanchart and GFN modules; (iii) any triggered stress tests; (iv) the sustainability assessment; and (v) a statement on whether debt stabilizes in the baseline. - For Fund SRDSAs, staff can add relevant commentary to the table, especially noting any use of judgment where the mechanical signal differs from the final assessment. - A summary assessment chapeau paragraph beneath the table that reports the overall assessment of risks and considerations supporting that determination; when included, the intuition behind an included sustainability assessment (but not the mechanical sustainability assessment) should be mentioned. - Debt disclosures reporting key metadata on public debt statistics and details of any consolidations performed to net out cross-holdings within the SRDSA’s debt perimeter. - Debt profile figures by currency, holders, governing law, marketability, and maturity. - A summary of the baseline scenario reporting: (i) public debt; (ii) contributions to the change in public debt by component (tabular and graphical); (iii) GFNs by component; and (iv) other key macroeconomic variables. - The output of the realism tools (panel figure of tools). - Mechanical output of the near-term assessment (if applicable): summary table reporting the logit stress probability and contributions to its change and a figure showing evolution vis-à-vis thresholds. - Mechanical outputs of medium-term assessment tools: (i) final debt fanchart; (ii) figures illustrating GFNs analyzed by the GFN Module; (ii) charts illustrating the components of the DFI and GFI, with a comparison to relevant peers; and (iv) a summary figure illustrating the evolution of medium-term index against thresholds. - Outputs of long-term modules if used. - Publication and Transparency Policy constraints (paragraph 154): - For Fund documents subject to the IMF’s Transparency Policy, certain SRDSA elements may require deletion prior to publication. - When the country document is sent to SPR for clearance after review department comments, staff should include both the SRDSA presented to the Board and a version showing required deletions for publication. - After circulation to the Board, deletions must be submitted through the Transparency Portal like any other deletion to Board documents. - Specific deletion requirements: - Near-term assessment (when included with the SRDSA): In line with the IMF Executive Board’s decision in January 2021, no elements of the near-term assessment are to be published prior to a review that will be undertaken 12 months after the SRDSF roll-out. Prior to this review, staff should delete the near-term signal, final assessment, and commentary from the summary table. The standardized reporting for the near-term assessment should also be deleted prior to publication. - A special case: when the near-term assessment is high risk, the medium and long-term mechanical signals are low or moderate risk, and staff assesses overall risk as high, staff should strongly consider how near-term risks could extend past the short-term and judgmentally assess medium-term risks as high when preparing the DSA to avoid unintended signals. - The mechanical signal on debt sustainability provided to the Board should be deleted regardless of whether the assessment was mandatory or optional (surveillance-only). - The probability of sustainable debt should be deleted from the final sustainability assessment except when required by the Fund’s lending policies. In practice, this deletion involves removing the words “with a high probability” or “but not with high probability”, leaving only an indication that debt is “sustainable”. - Certain cases require the final three-way assessment to remain because of applicable lending policies: In arrangements subject to the Exceptional Access policy, it is essential for evaluating Criterion 2. Similarly, in certain arrangements, sustainability with a high probability is a qualification criterion which must be noted (e.g., in FCLs, PLLs, and SLLs). - Any commentary that refers to elements subject to deletion will need to be removed prior to publication; staff should craft commentary on sensitive issues so it can be easily removed with minimal rephrasing. *Source: SRDSF GUIDANCE NOTE (excerpts).* ### References ### References ### Cited works - Aligishiev, Zamid, Matthieu Bellon, and Emanuele Massetti, 2022, “Macro-Financial Implications of Adaptation to Climate Change,” IMF Staff Climate Note 2022/002 (Washington: International Monetary Fund). - Arslanalp, Serkan and Takahiro Tsuda, 2012, “Tracking Global Demand for Advanced Economy Sovereign Debt,” IMF Working Paper No. 12/284, (Washington: International Monetary Fund). - Arslanalp, Serkan and Takahiro Tsuda, 2014, “Tracking Global Demand for Advanced Economy Sovereign Debt,” IMF Working Paper No. 14/39, (Washington: International Monetary Fund). - Borio, Claudio and Philip Lowe, 2002, “Assessing the risk of banking crises,” BIS Quarterly Review, December. - Buchheit, Lee, Guillaume Chabert, Chanda DeLong, and Jeromin Zettelmeyer, 2019, “The Restructuring Process,” in S. Ali Abbas, Alex Pienkowski, and Kenneth Rogoff, “Sovereign Debt: A Guide for Economists and Practitioners,” pp. 328-64, Oxford University Press. - Cantelmo, Alessandro, Leo Bonato, Giovanni Melina, and Gonzalo Salinas, 2019, “Policy Trade-Offs in Building Resilience to Natural Disasters: The Case of St. Lucia,” IMF Working Paper No.19/54, (Washington: International Monetary Fund). - Céspedes, Luis Felipe and Andrés Velasco, 2013, “Was This Time Different? Fiscal Policy in Commodity Republics,” NBER Working Paper No. 19748. - Climate Policy Initiative (CPI), 2021. Global Landscape of Climate Finance 2021. - Dabla-Norris, Era, James Daniel, and Masahiro Nozaki, 2021, “Fiscal Policies to Address Climate Change in Asia and the Pacific,” IMF Asia and Pacific Department -Fiscal Affairs Department No 21/07, (Washington: International Monetary Fund). - Drehmann, M., S. Pradhan, P. Wooldridge, and R. Szemere, 2016, “Recent enhancement to the BIS statistics,” BIS Quarterly Review, September. - Duenwald, Christoph, Yasser Abdih, Kerstin Gerling, and Vahram Stepanyan, 2022, “Feeling the Heat: Adapting to Climate Change in the Middle East and Central Asia,” IMF Departmental Paper DP/2022/008. March. (Washington: International Monetary Fund). - Economic Commission for Latin America and the Caribbean, 2014, The economics of climate change in Latin America and the Caribbean: Paradoxes and challenges, United Nations. November. - Economic and Social Commission for Asia and the Pacific, 2019, “The Disaster Riskscape across Asia-Pacific: Pathways for resilience, inclusion and empowerment, “ Asia-Pacific Disaster Report. United Nations, August. - Escolano, Julio, 2010, “A Practical Guide to Public Debt Dynamics, Fiscal Sustainability, and Cyclical Adjustment of Budgetary Aggregates,” Fiscal Affairs Department, January 2010, (Washington: International Monetary Fund). - Gaspar, Vitor, David Amaglobeli, Mercedes Garcia-Escribano, Delphine Prady, and Mauricio Soto, 2019, “Fiscal Policy and Development: Human, Social, and Physical Investment for the SDGs,” IMF Staff Discussion Note 19/03. (Washington: International Monetary Fund). - Global Commission on Adaptation, 2019, “Adapt Now: A Global Call for Leadership on Climate Resilience.” - International Monetary Fund, 2011, “Modernizing the Framework for Fiscal Policy and Public Debt Sustainability Analysis.” IMF Policy Papers Series, August, (Washington). - International Monetary Fund, 2012, “Commodity Price Swings and Commodity Exporters,” Chapter 4, World Economic Outlook, April, (Washington). - International Monetary Fund, 2013a, “Public Sector Debt Statistics: Guide for Compilers and Users,” (Washington). - International Monetary Fund, 2013b, “Staff Guidance Note for Public Debt Sustainability Analysis in Market Access Countries,” IMF Policy Papers Series, May, (Washington). - International Monetary Fund, 2014, “Government Finance Statistics Manual,” (Washington). - International Monetary Fund, 2015b, “Where Are Commodity Exporters Headed? Output Growth in the Aftermath of the Commodity Boom,” Chapter 2, World Economic Outlook, October, (Washington). - International Monetary Fund, 2016, “Small States’ Resilience to Natural Disasters and Climate Change – Role for the IMF,” IMF Policy Paper. December. Page 66. (Washington). - International Monetary Fund, 2018a, “Guidance Note on the Bank-Fund Debt Sustainability Framework for Low Income Countries,” IMF Policy Paper, (Washington: International Monetary Fund). - International Monetary Fund, 2018b, “Guidance Note for the Use of Third-Party Indicators in Fund Reports,” IMF Policy Paper, (Washington). - International Monetary Fund, 2020a, “Adapting to Climate Change in Sub-Saharan Africa,” Regional Economic Outlook: Sub-Saharan Africa, Chapter 2, (Washington). - International Monetary Fund, 2021a, “Review of the Debt Sustainability Framework for Market Access Countries,” IMF Policy Paper No. 2021/003, (Washington). - International Monetary Fund, 2021b, “Guidance Note on Implementing the Debt Limits Policy in Fund Programs,” IMF Policy Paper No. 2021/037, (Washington). - International Monetary Fund, 2021c, “Guidance Note for Fund Staff on the Treatment and Use of SDR Allocations,” IMF Policy Paper No. 2021/059, (Washington). - International Monetary Fund, 2021d, “Issues in Restructuring of Sovereign Domestic Debt,” IMF Policy Paper No. 2021/071, (Washington). - International Monetary Fund, 2021e, “Fiscal Policies for Climate Change Adaptation,” (forthcoming), (Washington). - International Monetary Fund, 2021f, World Economic Outlook: Recovery During a Pandemic, October, (Washington). - International Monetary Fund, 2022a, “Proposal to Establish a Resilience and Sustainability Trust,” IMF Policy Paper No. 2022/013, (Washington). - International Monetary Fund, 2022b, “Modification to the Transparency Policy,” IMF Policy Paper No. 2022/035, available at: https://www.imf.org/-/media/Files/Publications/PP/2022/English/PPEA2022035.ashx (Washington). - International Monetary Fund, 2022c, Guidance Note for Surveillance under Article IV Consultations, forthcoming (Washington). - Iossifov, Plamen and Thomas Dutra Schmidt, 2021, “Cyclical Patterns of Systemic Risk Metrics: Cross-Country Analysis,”. IMF Working Paper No. 2021/028, (Washington: International Monetary Fund). - IPCC, 2021, “Summary for Policymakers, in: Climate Change 2021: The Physical Science Basis,” Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press. In Press. - Kahn, Matthew, Kamiar Mohaddes, Ryan N.C. Ng, M. Hashem Pesaran, Mehdi Raissi, Jui-Chung Yang, 2019, “Long-Term Macroeconomic Effects of Climate Change: A Cross-country Analysis, “IMF Working Paper No. 2019/215, (Washington: International Monetary Fund). - Kangur, Alvar, Koralai Kirabaeva, Jean-Marc Natal, and Simon Voigts, 2019, “How Informative Are Real Time Output Gap Estimates in Europe?” IMF Working Paper No. WP/19/200, (Washington: International Monetary Fund). - Kaufmann, Daniel, Aart Kraay, and Massimo Mastruzzi, 2010, “The Worldwide Governance Indicators: Methodology and Analytical Issues,”. World Bank Policy Research Working Paper 5430. (Washington: World Bank). - Laeven, Luc and Fabian Valencia, 2020, “Systemic Banking Crises Database II,” IMF Economic Review 68(2), pp 307-61. - Laubach, Thomas, 2009, “New Evidence on the Interest Rate Effects of Budget Deficits and Debt.” Journal of the European Economic Association 7 (4), June, pp 858-85. - Phillis, Yannis, Nektarios Chairetis, Evangelos Grigoroudis, Fotis Kanellos, and Vassilis Kouikoglou, 2018, “Climate security assessment of countries,” Climate Change 148, pp 25-43. - OECD, 2019, “Pensions at a Glance,” Country Profiles-Brazil. - Spatafora, Nikola, and Issouf Samake, 2012, “Commodity Price Shocks and Fiscal Outturns,” IMF Working Paper No. WP/12/112, (Washington: International Monetary Fund). - Tiedemann, Johanna, Veronica Piatkov, Dinar Prihardini, Juan Carlos Benitez, and Aleksandra Zdzienicka, 2021, “Meeting the Sustainable Development Goals in Small Developing States with Climate Vulnerabilities: Cost and Financing,” IMF Working Paper No. 21/62. (Washington: International Monetary Fund). - UNEP, 2021, Adaptation Gap Report 2021: The Gathering Storm - Adapting to climate change in a post-pandemic world, (Nairobi, Kenya: United Nations Environment Programme). - United Nations, 2019, World Population Prospects 2019, Department of Economic and Social Affairs, Population Division. - Willems, Tim, and Jeromin Zettelmeyer, 2021, “Sovereign Debt Sustainability and Central Bank Credibility,” CEPR Discussion Papers No. 16817, (London: Centre for Economic Policy Research). ### Annex I — Comparison of the SRDSF and MAC DSA - Coverage - SRDSF: GG as default; justification required for narrower coverage; broader coverage (including central bank) in some cases - SRDSF: Disclosure requirements on coverage definitions, debtholder profile, and guidance on certain instruments (like swaps) - MAC DSA: Narrower than GG in some cases; no disclosure requirement on coverage - Horizon - SRDSF: 10-year debt and GFN projections for all cases - SRDSF: Risk assessments for near-, medium-, and long-term horizons - MAC DSA: 5-years projections - MAC DSA: No distinction in horizons - Realism tools - SRDSF: Cover additional drivers (exchange rate, financing terms on external debt, stock-flow adjustments), and public debt - SRDSF: In-depth tools for potential growth and fiscal multipliers. - MAC DSA: Cover growth, inflation and primary balance. - Near-term risks / Stress indicators - SRDSF: 10 indicators, in five categories: quality of institutions, stress history, cyclical, debt burden, and global - Heatmap - SRDSF: Debt and GFN levels, five debt profile and market indicators each producing a risk signal for heatmap - MAC DSA: No aggregation/overall signal - Composite index - SRDSF: Multivariate logistic regression combines indicators in a continuous metric (fitted probability of stress) - Signal derivation - SRDSF: Stress probability split in low, moderate, and high-risk zones, (calibrated to 10% missed crisis and false alarm rates) - Medium-term risks - Debt fanchart - SRDSF: Fanchart tool - MAC DSA: Visual tool based on symmetric shocks (asymmetric shocks used at team’s discretion) - Stress indicators - SRDSF: Three indicators: i) probability debt does not stabilize in medium term, ii) fanchart width, and iii) debt level at t+5 controlling for debt-carrying capacity (fanchart accounts for deviation of baseline projections from historical trends via skewed shocks) - MAC DSA: No signal/indicators; - Composite index - SRDSF: Index based on 3 indicators weighted by predictive power - MAC DSA: No signal/indicators; - Signal derivation - SRDSF: Index split in low, moderate, and high-risk zones, (calibrated to 10% missed crisis and false alarm rates) - MAC DSA: Interpretation of fancharts at team’s discretion. - GFN Tool / Macro-fiscal shocks - Stress indicators (SRDSF): - Three indicators: (i) initial bank claims on government, (ii) maximum cumulative change in bank claims over projection period under a generalized stress scenario; (iii) average projected GFN/GDP in baseline. - MAC DSA: Effect of shocks to primary balance, real GDP growth, real interest rate, and exchange rate on debt and GFN levels reflected in heat map signals - Composite index - SRDSF: Index based on 3 indicators weighted by predictive power - Signal derivation - SRDSF: Index split in low, moderate, and high-risk zones, (calibrated to 10% missed crisis and false alarm rates). - Triggered stress-tests - SRDSF: Simulate debt and GFN paths under: (i) contingent liabilities related to narrow coverage, (ii) banking crisis, (iii) natural disasters, (iv) commodity price shocks, and (v) REER shock. - SRDSF: Allows for customized stress-tests for idiosyncratic risks. - MAC DSA: Allows for customized stress-tests for idiosyncratic risks. - Long-term risk analysis - SRDSF: Optional tools for risks from: population aging, natural resource discovery/depletion, debt amortizations; and climate change. - MAC DSA: Option to extend debt and GFN projections - Judgment and communication - SRDSF: Judgment-based risk assessments at each horizon, with deviation from mechanical signals explained. - SRDSF: Overall risk assessment based on user judgment. - MAC DSA: No aggregate mechanical signals; lack of standardized bottom-line assessments; unclear application of judgment. *International Monetary Fund — References and Annex I content as provided.* ### Annex II. Model SRDSA Prepared by Fund Staff ### Annex II. Model SRDSA Prepared by Fund Staff ### Purpose and scope - Presents model SRDSA items along the lines of those to be issued to the IMF’s Executive Board. - Explains variations in SRDSA output for countries under different circumstances. - Summarizes items that may be subject to deletions prior to publication under the IMF’s Transparency Policy. ### Variations by engagement type - Surveillance-only context (optional DSA performed): includes near-term assessment and sustainability assessment. (Figure A.II.1a) - Note: If this assessment were not performed, the mechanical signal and final assessments would have “n.a.” and the commentary would indicate “optional for surveillance-only countries”.1 - Program treated as precautionary (no drawings expected): includes near-term assessment and required sustainability assessment. (Figure A.II.1b) - Non-precautionary program: near-term assessment not applicable; sustainability assessment present. (Figure A.II.1c) - Publication implications: when included, the near-term and sustainability assessments have implications for automatic deletions under the Transparency Policy (summarized in section IX). ### Item 1. Summary of the Sovereign Risk and Debt Sustainability Assessment - Purpose: summarizes results of all SRDSA elements including the mechanical and final risk signals, sustainability assessments, debt stabilization in the baseline, and user commentary. - Mechanical signal deletion rule: "The mechanical signal of the debt sustainability assessment is deleted before publication."2 - Qualifier deletion rule: "In surveillance-only cases or cases with IMF arrangements with normal access, the qualifier indicating probability of sustainable debt ('with high probability' or 'but not with high probability') is deleted before publication."2 - Near-term assessment applicability: - "The near term assessment is not applicable in cases where there is a disbursing IMF arrangement. In surveillance-only cases or in cases with precautionary IMF arrangements, the near-term assessment is performed but not published."1 ### Illustrative SRDSA findings (Ruritania examples) - Overall risk: Moderate - Near term: - Surveillance-only and precautionary program examples: Near-term risk = Moderate - Non-precautionary program example: Near-term assessment = n.a. - Medium term: - Mechanical signal: Low - Final assessment: Moderate - Comments: Medium-term risks elevated by contingent liabilities from a narrow debt coverage and sub-national governments with symptoms of weak finances. - Debt Fanchart: Low - GFN (GFN Financeability Module): Moderate - Long term: Moderate - DSA summary assessment: "Sustainable but not with high probability" (appears in surveillance-only, precautionary, and non-precautionary examples where DSA is performed) - Debt stabilization in the baseline: - "The projected debt path is expected to stabilize and GFNs will remain at manageable levels, conditional on the implementation of fiscal adjustment measures that are assessed as feasible in spite of not yet being legislated." - Key commentary excerpts (preserved language): - "This SRDSA finds that Ruritania's debt is sustainable but not with high probability." - "Following a recent shock, public debt increased significantly and international reserves fell to very low levels, raising the risk of stress in the near term and leading the Logit module to provide a risk signal of medium." - "Market conditions are beginning to show signs of normalization, however, and the strong policy measures set to be implemented should strengthen confidence further." - "Fiscal adjustment will deliver a debt trajectory that stabilizes at somewhat lower levels. Consequently, the Debt Fanchart module gives a low risk signal." - "Contingent liabilities from sub-national governments that are excluded from the SRDSA's debt perimeter are plausible and would have high impacts if realized. Thus, the medium-term risk assessment has been set at moderate risk despite the low risk mechanical signal." - "Over the longer run, Ruritania should continue with reforms to tackle risks arising from population aging on the social security fund. However, the long time horizon at which these risks would materialize and the authorities' planned measures will help contain risks." ### Item 2. Debt Coverage and Disclosures - Purpose: summarizes key elements of the public debt concept used in the DSA; has no variations depending on engagement type; no automatic deletions under the Transparency Policy. - Coverage statement (example): - "This SRDSA covers the central government, with no consolidation of other public entities" - Debt coverage entry (exact text preserved): - "1. Debt coverage in the DSA: 1/CGGGNFPSCPSOther" - "1a. If central government, are non-central government entities insignificant? No" - Subsectors included in the chosen coverage (as presented): - "Subsectors captured in the baseline — Inclusion" - 1 Budgetary central government — Yes - 2 Extra budgetary funds (EBFs) — No - 3 Social security funds (SSFs) — Yes - 4 State governments — No - 5 Local governments — No - 6 Public nonfinancial corporations — No - 7 Central bank — No - 8 Other public financial corporations — No - Remaining items in the figure (instrument coverage; accounting principles) are listed headers in the source but not expanded in the provided excerpt. *Annex II. Model SRDSA Prepared by Fund Staff — Presented as in the source document.* ### 5. Debt consolidation across sectors: ### 5. Debt consolidation across sectors ### Coverage and institutional context - Coverage in this SRDSA is for the central government. - Authorities are taking efforts to expand the perimeter of public debt statistics: - A new census of state and local government finances is being conducted, with preliminary results expected next year. - Notes and definitions referenced: - 1/ CG=Central government; GG=General government; NFPS=Nonfinancial public sector; PS=Public sector. - 2/ Stock of arrears could be used as a proxy in the absence of accrual data on other accounts payable. - 3/ Insurance, Pension, and Standardized Guarantee Schemes, typically including government employee pension liabilities. - 4/ Includes accrual recording, commitment basis, due for payment, etc. - 5/ Nominal value at any moment in time is the amount the debtor owes to the creditor. - 6/ The face value of a debt instrument is the undiscounted amount of principal to be paid at (or before) maturity. - 7/ Market value of debt instruments is the value as if they were acquired in market transactions on the balance sheet reporting date (reference date). Only traded debt securities have observed market values. ### Current public debt structure and projections - Perimeter shown in the charts: - Central government for "Debt by currency". - General government for "Public debt by legal basis, 2020". - Other perimeters noted for some instrument breakdowns. - Commentary: Public debt will remain about evenly split between foreign and local currency-denominated instruments, but maturities are expected to lengthen. Marketable instruments form the bulk of public debt and it is held by domestic creditors. - Residual maturity: 5.5 years (projection chart). - Baseline scenario (percent of GDP unless indicated otherwise): - Public debt: 2020: 79.8; 2021: 73.1; 2022: 75.6; 2023: 75.9; 2024: 75.1; 2025: 75.2; 2026: 74.5; 2027: 72.4; 2028: 72.9; 2029: 71.8; 2030: 70.3; 2031: 67.9. - Change in public debt: 2020: 4.7; 2021: -6.7; 2022: 2.5; 2023: 0.3; 2024: -0.8; 2025: 0.1; 2026: -0.7; 2027: -2.1; 2028: 0.5; 2029: -1.1; 2030: -1.5; 2031: -2.5. - Contribution of identified flows: 2020: 2.2; 2021: -10.6; 2022: 1.5; 2023: -0.2; 2024: -1.4; 2025: -0.4; 2026: -1.4; 2027: -4.0; 2028: 0.4; 2029: -2.6; 2030: -1.7; 2031: -4.1. - Primary deficit: 2020: 2.2; 2021: 2.6; 2022: 2.4; 2023: 0.6; 2024: -0.6; 2025: 0.3; 2026: -0.5; 2027: -1.5; 2028: 1.6; 2029: 1.1; 2030: 1.4; 2031: 0.7. - Noninterest revenues (percent of GDP): 2020: 20.0; 2021: 20.4; 2022: 21.5; 2023: 21.7; 2024: 22.6; 2025: 22.7; 2026: 23.2; 2027: 24.0; 2028: 20.8; 2029: 21.0; 2030: 21.8; 2031: 21.4. - Noninterest expenditures (percent of GDP): 2020: 22.2; 2021: 23.0; 2022: 23.9; 2023: 22.3; 2024: 22.0; 2025: 23.1; 2026: 22.8; 2027: 22.4; 2028: 22.5; 2029: 22.1; 2030: 23.2; 2031: 22.1. - Automatic debt dynamics: 2020: 0.0; 2021: -9.7; 2022: -1.0; 2023: -1.0; 2024: -0.8; 2025: -0.8; 2026: -0.8; 2027: -0.8; 2028: -1.1; 2029: -2.0; 2030: -2.7; 2031: -3.1. - Int. rate-growth differential: 2020: 6.5; 2021: -9.8; 2022: -0.9; 2023: -0.8; 2024: -0.8; 2025: -0.8; 2026: -0.9; 2027: -0.9; 2028: -1.2; 2029: -2.2; 2030: -3.1; 2031: -3.4. - Real interest rate: 2020: 1.3; 2021: 0.9; 2022: 1.4; 2023: 1.9; 2024: 1.7; 2025: 1.9; 2026: 1.6; 2027: 1.6; 2028: 1.3; 2029: 0.3; 2030: -0.7; 2031: -1.0. - Real growth rate: 2020: 5.2; 2021: -10.7; 2022: -2.3; 2023: -2.7; 2024: -2.5; 2025: -2.7; 2026: -2.6; 2027: -2.5; 2028: -2.4; 2029: -2.5; 2030: -2.4; 2031: -2.4. - Relative inflation: 2020: 0.3; 2021: 0.1; 2022: -0.1; 2023: -0.2; 2024: 0.0; 2025: -0.1; 2026: 0.2; 2027: 0.2; 2028: 0.1; 2029: 0.2; 2030: 0.4; 2031: 0.3. - Stock-flow adjustment: 2020: 2.6; 2021: 0.4; 2022: 1.1; 2023: 0.7; 2024: 0.6; 2025: 0.6; 2026: 0.5; 2027: 0.2; 2028: 0.0; 2029: -0.2; 2030: -0.2; 2031: -0.1. - Contingent liabilities and Other transactions: all listed as 0.0 in the baseline table. - Residual: same as stock-flow adjustment series. - Gross financing needs (GFN): 2020: 15.4; 2021: 14.6; 2022: 16.5; 2023: 14.8; 2024: 15.4; 2025: 16.4; 2026: 19.5; 2027: 14.2; 2028: 18.2; 2029: 19.0; 2030: 17.1; 2031: 15.2. - of which: debt service: 2020: 13.2; 2021: 12.0; 2022: 14.1; 2023: 14.2; 2024: 15.9; 2025: 16.0; 2026: 20.0; 2027: 15.7; 2028: 16.6; 2029: 17.9; 2030: 15.7; 2031: 14.5. - GFN by currency: Local currency: 2020: 11.7; 2021: 9.1; 2022: 11.0; 2023: 10.9; 2024: 12.6; 2025: 12.4; 2026: 16.1; 2027: 11.5; 2028: 9.1; 2029: 10.2; 2030: 10.4; 2031: 10.4. Foreign currency: 2020: 1.5; 2021: 3.0; 2022: 3.1; 2023: 3.3; 2024: 3.3; 2025: 3.6; 2026: 3.9; 2027: 4.2; 2028: 7.5; 2029: 7.6; 2030: 5.3; 2031: 4.1. - Memo macro assumptions: - Real GDP growth (percent): 2020: -6.5; 2021: 15.5; 2022: 3.3; 2023: 3.7; 2024: 3.4; 2025: 3.8; 2026: 3.6; 2027: 3.5; 2028: 3.5; 2029: 3.5; 2030: 3.5; 2031: 3.5. - Inflation (GDP deflator; percent): 2020: 2.2; 2021: 2.4; 2022: 1.9; 2023: 1.7; 2024: 2.0; 2025: 1.8; 2026: 2.3; 2027: 2.4; 2028: 2.2; 2029: 2.6; 2030: 3.0; 2031: 2.9. - Nominal GDP growth (percent): 2020: -4.4; 2021: 18.3; 2022: 5.2; 2023: 5.4; 2024: 5.5; 2025: 5.6; 2026: 6.0; 2027: 6.0; 2028: 5.7; 2029: 6.2; 2030: 6.6; 2031: 6.5. - Effective interest rate (percent): 2020: 3.9; 2021: 3.8; 2022: 4.0; 2023: 4.3; 2024: 4.4; 2025: 4.5; 2026: 4.7; 2027: 4.7; 2028: 4.0; 2029: 3.0; 2030: 2.1; 2031: 1.3. - Staff commentary: Public debt will rise a bit but then stabilize, reflecting expectations of a narrowing of primary deficits and stable economic conditions. ### Near-term and realism assessments - Near-term Risk Analysis (Logit Stress Probability, LSP): - Year of data: 2017 2018 2019 2020 (columns for prediction windows). - Logit stress probability (LSP): 2018-19: 0.095; 2019-20: 0.045; 2020-21: 0.103; 2021-22: 0.078. - Change in LSP: 0.050; -0.049; 0.058; -0.025 (by period). - Contributions to change in LSP (selected): - Debt burden & buffers: 0.058; -0.080; 0.077; -0.009. - Global conditions: -0.012; 0.031; -0.021; -0.022. - Cyclical position: 0.001; 0.001; 0.001; 0.006. - Institutional quality and Stress history: mostly 0.000 in the table. - Missed crisis probability in 2021-22 if stress not predicted: 15% - False alarm probability in 2021-22 if stress predicted: 72% - Staff commentary: Ruritania's risk of near-term stress has improved, but remains within the moderate risk range. The improvement largely reflects the normalization in global conditions from the COVID-19 shock as well as an improvement in debt indicators as the domestic recovery has gained steam. - Realism of baseline assumptions (qualitative summary): - Recovery from COVID-19 imparts complicated effects on growth path. - Realism analysis highlights: - Past forecast errors do not reveal systematic biases. - Projected fiscal adjustment are well within norms. - Debt reduction projections are in the top-quartile of cross-country database. - Spreads are projected to compress in a period of tightening global financial conditions. ### Medium-term risk assessment (selected metrics and signals) - Debt fanchart module: - Fanchart width (percent of GDP): 24.8 - Probability of debt non-stabilization (percent): 5.9 - Terminal debt-to-GDP x institutions index: 25.1 - Debt fanchart index (DFI): 0.8 - Risk signal: Low - Gross financing needs (GFN) module: - Average baseline GFN (percent of GDP): 16.2 - Banks' claims on the general government (pct banks' assets): 7.4 - Change in banks' claims in stress (pct banks' assets): 0.7 - GFN financeability index (GFI): 8.2 - Risk signal: Moderate - Medium-term index and final assessment: - Medium-term index combines normalized DFI and GFI contributions; Final assessment: Moderate. - Prob. of missed crisis, 2022-2027 (if stress not predicted): 9.1 pct - Prob. of false alarm, 2022-2027 (if stress predicted): 61.1 pct - Commentary: Of the two medium-term tools, the GFN Financeability Module is pointing to a higher, but still moderate level of risk. This is reinforced by the potential for contingent liabilities from government entities outside the central government, given signs of strains in subnational governments' finances. ### Long-term risks and sustainability under maximum adjustment effort - Long-term considerations: - Over the long-run, Ruritania has meaningful risks from population aging. - Under a no-reform scenario, public sector debt and GFNs would be noticeably higher than the long-term baseline. - Authorities have appointed a commission to explore alternatives; additional fiscal and structural reforms will help contain longer-term risks. - Maximum adjustment scenario (summary table, percent of GDP unless otherwise indicated): - Public sector debt: 2020: 79.8; 2021: 73.1; 2022: 75.6; 2023: 75.9; 2024: 75.1; 2025: 75.2; 2026: 74.5. - Gross financing needs: 2020: 15.4; 2021: 14.6; 2022: 16.5; 2023: 14.8; 2024: 15.4; 2025: 16.4; 2026: 19.5. - Primary balance: 2020: -2.2; 2021: 0.0; 2022: 0.5; 2023: 0.8; 2024: 1.0; 2025: 1.0; 2026: 1.0. - Real GDP growth (percent): 2020: -6.5; 2021: 15.5; 2022: 3.3; 2023: 3.7; 2024: 3.4; 2025: 3.8; 2026: 3.6. - Inflation (GDP deflator; percent): 2020: 2.2; 2021: 2.4; 2022: 1.9; 2023: 1.7; 2024: 2.0; 2025: 1.8; 2026: 2.3. - Nominal GDP growth (percent): 2020: -4.4; 2021: 18.3; 2022: 5.2; 2023: 5.4; 2024: 5.5; 2025: 5.6; 2026: 6.0. - Effective interest rate (percent): 2020: 3.9; 2021: 3.8; 2022: 4.0; 2023: 4.3; 2024: 4.4; 2025: 4.5; 2026: 4.7. - Medium-term tools under the maximum adjustment scenario: - Debt fanchart index: 0.3 - Fanchart width: 24.8 - Probability of debt stabilization: 5.9 - Debt level x institutions index: 25.1 - Avg. GFN, max. adjustment scenario (pct of GDP): 5.5 - GFN financeability index: 8.2 - Bank claims on the general government (pct bank assets): 2.4 - Change in bank claims in stress (pct bank assets): 0.2 - Sustainability assessment: "Sustainable but not with high probability" - Policy-relevant simulation result: - Revenue mobilization measures (identified through recent Technical Assistance) with a cumulative yield of 2 percent of GDP and realistic cuts to non-priority expenditures totaling about 1 percent of GDP could deliver primary surpluses of about 1 percent of GDP over the medium term. If delivered, these measures would be consistent with a debt sustainability assessment of "sustainable but not with high probability." *Source: IMF staff estimates and projections as presented in the SRDSA chapter.* ### Annex III. Summary Directions to Run the SRDSF ### Annex III. Summary Directions to Run the SRDSF ### Overview - Purpose: Provide concise operational directions to IMF staff running the SRDSF; not a substitute for the full guidance in the main text. - Structure: Stepwise process to run the framework, with section references to the guidance note for full details. ### Step-by-step operational directions (process) - Step 1: Answer the questionnaire on debt disclosures in the template (Section III.A). - Step 2: Upload the debt data and projections into the template to run the standard tools (Table AIII.1) and populate the debt structure charts (Section III.B). For SRDSAs prepared by Fund staff, link the input sheet to the underlying macro framework database or a bridge file that converts items from the MAC DSA template. - Step 3: Enter the financing assumptions on new debt issuance (e.g., amounts, maturities, interest rates, etc.). - Step 4: Scrutinize the baseline debt and GFN projections (Section III.C) and the realism tools (Section IV). For SRDSAs prepared by Fund staff, ensure paths are consistent with the macro framework. Check realism tools for any flags; revise the baseline or document explanations in commentary if issues arise. - Step 5: Perform preliminary analysis by running the debt fanchart module to see if the realism adjustment is activated (Section VI.A, paragraphs 58 to 60). If activated, typically revise the baseline so the downward debt trajectory is less favorable and re-run the module; if revision is impossible, explore criteria for disregarding this realism diagnostic. - Step 6: Prepare the near-term assessment (Section V). - For surveillance-only or precautionary program cases: template automatically produces mechanical results if data correctly entered. Note the mechanical signal, Logit Stress Probability, and contributions to the change in the probability. If results are counterintuitive, explore whether a judgment-based final assessment is warranted. - In non-precautionary program contexts: set the final near-term assessment to “not applicable”. - Step 7: Finalize the debt fanchart (remainder of section VI.A). Note the Debt Fanchart Index, its components, and the mechanical signal. If counterintuitive, judgment may be needed in step 10. - Step 8: Run the GFN module (section VI.B). Note the GFN Financeability Index, its components, the mechanical signal, and whether judgment might be needed later (step 10) to explain non-sensible medium-term results. - Step 9: Check activation criteria for any triggered stress tests (section VI.C). If a stress test is performed, check its position on the fanchart. If it is in the 75th percentile and the underlying shock has a sufficiently high probability of materialization, expect a judgmental downgrade for the final medium-term assessment in the next step. - Step 10: Complete the medium-term assessment (section VI.D). Note the Medium-Term Index and mechanical signal. Apply judgment if results are counterintuitive for the Debt Fanchart (Step 7) or GFN Module (Step 8). If any stress test (Step 9) met criteria for a judgmental downgrade, typically weaken the final medium-term assessment by one notch. - Step 11: Conduct the long-term risk assessment (section VII). Arrive at a final judgment-based assessment of high/moderate/low risk based on the extended baseline and other relevant factors. Examine debt trajectory for stabilization and levels/trends of gross financing needs. Long-term modules optional but encouraged; data requirements in Table A.III.2. - Step 12: Prepare the overall risk assessment (section VIII.A). Judgment-based but should reside within range of near-, medium-, and long-term assessments. Include arguments (including prospects for debt stabilization) in SRDSA commentary. - Step 13: Sustainability assessments (section VIII.B): - Part a: Determine if a sustainability assessment is warranted. Required in DSAs for Fund-supported programs; optional for surveillance-only countries. If unnecessary, skip to step 14. - Part b: Conduct the sustainability assessment (IMF staff only). Calculate probability of unsustainable debt from the internal sustainability logit model. Combine that with the Debt Fanchart Index and GFN Financeability Index (using the approved aggregation rule) to obtain the mechanical sustainability metric. Compare against thresholds to give the mechanical signal (e.g., sustainable with a high probability; sustainable but not with high probability; or unsustainable). If counterintuitive, consider scope for a judgment-based sustainability assessment. - Step 14: Finalize the output for the IMF’s Executive Board (section IX, paragraphs 152-153). Ensure all standard reporting figures and tables are correctly populated and provide additional commentary to interpret results or explain uses of judgment. - Step 15: Finalize the output for publication (section IX, paragraph 154). Apply deletions per Transparency Policy: items to delete relate to near-term assessment (surveillance-only or precautionary program countries); mechanical signal on debt sustainability (when not required by IMF lending policies); and the probability of sustainable debt (when not required by IMF lending policies). ### Data and module inputs (near-, medium-, and long-term) - Near- and medium-term modules require fiscal data/projections up to t+5, debt breakdowns (by residency, currency, maturity, holder, legal basis), interest bill (existing debt) and receipts, amortization of existing debt, assumptions on new debt issuance, effective real interest rate, gross financing need (calculated), stock-flow adjustment, government liquid assets, cyclically adjusted primary balance, forecast track record, average maturity, debt coverage disclosures, intra-governmental debt holdings, major macro variables (real and nominal GDP and deflator, current account balance, nominal bilateral ER, international reserves, potential GDP and output gap), and a range of financial sector and structural indicators. - Long-term modules require baseline projections after t+5 for primary revenues, expenditures, balance; interest bill and receipts; debt; amortization of existing debt; assumptions on new debt issuance; gross financing need (calculated); stock-flow adjustment; macro variables (real and nominal GDP and deflator; nominal bilateral ER). Optional specialized long-term analyses include demographics, natural resource scenarios, large amortizations, and climate change inputs. - See Table AIII.1 and Table AIII.2 for detailed variable-by-module requirements and centralization scope. ### Key calibration and threshold rules (design and interpretation) - General rule: low-to-moderate threshold associated with a 10 percent missed crisis probability; moderate-to-high threshold associated with a 10 percent false alarm probability. - Logit Stress Probability (LSP): - Horizon: estimates probability of a stress event at a 1-2 year ahead horizon. - Average posterior probabilities by fitted signal: - High: 40 percent - Moderate: 16 percent - Low: 2 percent - Model captures nonlinearities via logistic curve positions and calibrated thresholds. - Debt Fanchart Index (DFI): - Components are normalized using empirical standard deviations from the historical sample for MACs (each akin to a z-score without mean subtraction). - Weights determined by AUCs from ROC analysis. - Normalization factors and weights (as presented): - Fanchart width: 0.38 0.32 - Probability of debt non-stabilization: 0.22 0.33 - Terminal debt level x institutions index: 0.16 0.36 - Average posterior probabilities by signal: - High: 44 percent - Moderate: 23 percent - Low: 3 percent - Thresholds for mechanical signal: - Low risk: Below 6.3 percent - Moderate risk: Between 6.3 and 19.5 percent - High risk: Above 19.5 percent - GFN Financeability Index (GFI): - Components aggregated without normalization (similar magnitudes). - Weights determined by AUCs. - Average posterior probabilities by signal: - High: 42.1 percent - Moderate: 4.1 percent - Low: 3.5 percent - Thresholds for mechanical signal: - Low risk: Below 7.6 - Moderate risk: Between 7.6 and 17.9 - High risk: Above 17.9 - Medium-Term Index (MTI): - Aggregates normalized DFI and GFI; normalization divides DFI and GFI by the respective maximum level observed in the historical calibration sample. - Normalization factors: - Debt Fanchart Index: 4.5 - GFN Financeability Index: 52.0 - Thresholds for mechanical signals: - Low risk: Below 0.257 - Moderate risk: Between 0.257 and 0.395 - High risk: Above 0.395 - Average posterior probabilities by MTI signal: - High: 43 percent - Moderate: 9 percent - Low: 4 percent *Source: IMF staff estimates (content from Annex III and Annex IV of the SRDSF guidance note).* --- _Source: https://www.imf.org/-/media/files/publications/pp/2022/english/ppea2022039.pdf_