IMF reviews its debt sustainability framework
The Executive Board of the International Monetary Fund (IMF) reviewed on January 14, 2021 the IMF Debt sustainability Framework for Market Access Countries (MAC DSA). The review revealed scope to improve the MAC DSA framework’s ability to identify risk of sovereign stress and better align it with the IMF’s lending framework, to be achieved by replacing the current approach with a new methodology.
The MAC DSA plays a key role in the Fund’s core functions of surveillance and lending. In surveillance, the framework helps identify a member’s vulnerability to sovereign stress to steer the member away from such stress. In Fund-supported programs, which often take place after the stress has already developed, the DSA helps determine if sovereign stress can be resolved via a combination of IMF financing and economic reforms, or if measures such as debt restructuring are needed to deliver medium-term debt sustainability. The framework is also used in developing IMF conditionality and informing the need for debt relief in debt restructuring operations undertaken in the context of Fund-supported programs.
Since its introduction in 2002, this framework has been reviewed in 2003, 2005, and 2011-13. The 2011-13 review introduced key features, including a risk-based approach through distinction between high and low scrutiny countries, standardization of writeup and publication requirements, realism tools to guard against optimistic economic projections, a heatmap summarizing debt vulnerabilities, and debt fancharts to give a sense of the uncertainty around the projected path of the debt/GDP ratio.
A careful review over the past two and a half years has revealed scope for further improvements, so as to predict sovereign stress with greater accuracy. The new framework includes a broader and more consistent debt coverage, a longer projection horizon, new tools at multiple horizons based on superior analytical methods that account for countries’ structural characteristics, and enhanced transparency in the bottom-line assessments, including the exercise of judgment. Furthermore, the new tools support probabilistic debt sustainability assessments, as required by the Fund’s lending framework.
The framework is expected to be operationalized in the final quarter of 2021/first quarter of 2022. This will be preceded by the completion of the accompanying Guidance Note and template, and extensive engagement with country authorities and other external stakeholders. The transition between the old and the new framework will be carefully managed to ensure consistency.
Executive Board Assessment
Executive Directors welcomed the wide-ranging and comprehensive review of the Debt Sustainability Framework for Market Access Countries (MAC DSA), to be renamed “Sovereign Risk and Debt Sustainability Framework for Market Access Countries” (MAC SRDSF) to capture the full range of its analysis. Against the backdrop of rising vulnerabilities related to the pandemic, they broadly supported the proposed reforms aimed at improving the framework’s capacity to predict sovereign stress, enhancing transparency and communication of its results, and aligning it with the three-zone sustainability assessment required under the exceptional access framework. Directors recognized that the framework would require some further technical fine-tuning in the run up to the preparation of the Staff Guidance Note and implementation.
Directors supported the continued application of the existing definition of debt sustainability, and most concurred that General Government (GG) debt, defined per GFSM 2014 classification, should be the default institutional coverage. A few Directors suggested that the expansion of debt coverage to GG be implemented in a phased manner, as two-fifths of EMs currently report data for the central government only. Directors welcomed the incorporation of public sector liquid financial assets as a mitigating factor, and most Directors supported the riskâ€‘based approach under which central bank liabilities and/or SOE contingent liabilities would need to be included in the debt perimeter. However, a few Directors advised the incorporation of a broader range of public sector assets and wider adoption of net public debt concepts in the framework. Directors stressed that capacity-development support would be needed to bring country data coverage to adequate levels. A few Directors preferred the continuation of the existing 5-year time horizon in certain cases in view of large uncertainties regarding public debt projections.
Directors welcomed the expanded realism toolkit for baseline projections and tools to assess sovereign risks at three horizons: short, medium, and long term. They supported the use of the proposed new tools, with slight adjustments, to produce the probabilistic debt sustainability assessments required in Fund-supported programs and evaluate the consistency of restructuring targets with restoring sustainability in debt restructuring cases. A number of Directors emphasized the need to adequately account for the impact of climate change on sovereign risk and debt sustainability. A few Directors questioned the expansion of the existing realism toolkit to cover exchange rate analysis, especially for pegged regimes. A number of Directors expressed concern about the use of perceptions-based third-party indicators to build the institutional quality variable used in the short- and medium-term models. In addition, these Directors asked to leave adequate room for judgment and, as a cross-check, compare results using alternative indicators of institutional quality that are not perceptions-based.
Directors agreed that a sovereign risk analysis should generally be prepared in both program and surveillance contexts. In a program context, staff reports should contain the full range of risk-of-sovereign-stress outputs for the medium and long term (but not for the near term), as well as an overall risk assessment. In surveillance and precautionary arrangement cases, most Directors endorsed full disclosure of sovereign risk analysis to the Board but limited disclosure (omitting the near-term risk signal and assessment) to the public for a 12-month period, at which time full disclosure to the public would be reconsidered based on the experience gained with the new framework. A number of Directors expressed concern about the unintended consequences from potential market sensitivities of full disclosure of sovereign risk. A number of other Directors favored moving to full disclosure of sovereign risk analysis to the public immediately. Directors noted that implementing the limited disclosure options would require a targeted modification to the Transparency Policy, which would be proposed on a lapse of-time basis.
Directors agreed that sustainability assessments should be required for arrangements involving GRA resources (including precautionary arrangements) as well as for the PCI. While most Directors agreed that sustainability assessments should be optional in surveillance cases, a few Directors favored preparing a sustainability assessment in surveillance cases with high risk of sovereign stress, with the results disclosed to the Board but not to the public, although a few other Directors would favor public disclosure even for such cases. With respect to program cases, a range of view were expressed. Some Directors preferred maintaining the current practice by which a three-zone assessment is included in staff reports in exceptional access cases but not in normal access cases. A few Directors suggested full disclosure (to the Board and the public) of three-zone assessments in both normal and exceptional access cases. In the end, Directors could go along with disclosure to the Board of three-zone assessments in both normal and exceptional access cases, and to the public only in exceptional access cases, with experience assessed at the end of a 12-month period.
In the context of precautionary arrangements, Directors agreed that sovereign risk assessments would be informed by the baseline scenario, while sustainability assessments would be informed both by the baseline and, when appropriate, by an adverse (full drawing) scenario. They agreed that the latter would be appropriate in exceptional access cases (excluding FCL cases), if shocks triggering a drawing are not adequately captured by the medium-term tools, or when review departments have doubts about the realism of the baseline that cannot be resolved through discussions with the country team, although a few Directors stressed that the appropriate use of the new realism tools should resolve any such doubts.
While most Directors supported the proposed timeline, with a carefully planned roll-out expected for Q4 2021 or Q1 2022, some Directors favored a more accelerated schedule, and a few others considered the proposed timeline could be ambitious. In this context, the transition between the old and the new framework should be carefully managed to ensure consistency. Directors looked forward to the preparation of a guidance note and new templates underpinning the new framework, accompanied by early engagement with a subset of country teams to test the new tools in parallel with the current framework. They encouraged the provision of appropriate capacity development support and maintaining close engagement with the Board as the framework is implemented, as well as ensuring an effective communication strategy with member-country authorities and external stakeholders during this process.