This chapter examines the experience with market debt operations undertaken in IMF-supported programs over the evaluation period.60 After providing an overview of these debt operations, it examines how programs with debt operations have fared in terms of growth and adjustment during and after the program. The assessment focuses primarily on 12 market debt operations, most of which were in the context of GRA programs.
Market Debt Operations in IMF-Supported Programs
The basic principle underlying the IMF’s role in supporting debt operations is that it is for the member country to decide whether and how to restructure its debt and to manage the whole restructuring process. At the same time, the IMF must ensure that any IMF-supported program can successfully achieve its objectives while safeguarding the revolving character of IMF resources, which requires a satisfactory judgment on debt sustainability as a basic prerequisite for access. The financing assurances policy requires that a program should have adequate external financing and, in cases where the financing gap cannot be filled by other means, explicitly encourages debt restructuring operations on terms compatible with balance of payments viability (IMF, 2013b).
The lending into arrears policies (LIA/LIOA) require that a country under an IMF-supported program be making good faith efforts to negotiate the restructuring of the debt in default with its private or official creditors. These policies are supported by two carefully developed frameworks for debt sustainability analysis, for market access countries (MAC DSA) and for LICs (LIC-DSF), the latter prepared jointly with the World Bank. The LIC-DSF was last modified in 2017 and the MAC DSA in early 2021.61 Under these policies, where debt is assessed as unsustainable or even sustainable but not with high probability (in exceptional access cases), debt operations have been required as a condition for access to Fund resources. Beyond this requirement, debt operations can contribute to support long-term growth by relieving the burden of future debt service, making more fiscal resources available for productive public investment, and improving incentives for private investors.
Following this framework, the IMF has played an active role in supporting market debt operations while adhering to the neutrality principle. The IMF’s financing assurance and lending into arrears policies as well as the Fund’s catalytic role have provided balanced incentives for debt restructuring on both debtor and creditor sides. The IMF has also provided technical advice in identifying resources available for debt servicing under alternative scenarios and assessing the restructuring envelope, primarily based on the DSA framework. In addition, Fund staff has often played a role in encouraging creditor participation by communicating with stakeholders about a country’s economic policies and prospects and debt servicing potential.
The Fund’s DSA framework has been used extensively in the program context to make difficult judgments on whether to insist on debt operations as a condition for access to IMF financing and whether the extent of debt relief in packages under negotiation would be sufficient to achieve debt sustainability. In some cases (e.g., Belize), the Fund’s DSA has also played a crucial role in the surveillance context by serving as an independent baseline for negotiations between authorities and creditors.
While the MAC DSA is a highly sophisticated framework that has evolved over time, it does not include a systematic analysis of the impact of debt restructuring on growth and the timing of market access. While DSA was used to evaluate the effect of different debt operations on the gross financing need and debt path, the endogenous linkage between debt and growth was not always clearly grounded. In practice, Fund staff generally took an ad hoc approach to assess market access and the debt-growth nexus used in program design, typically embedding the effects of debt operations in the form of lower primary balance and reduced interest payments.62 In some cases, growth projections were adjusted to reflect the design of the debt operation, particularly the extent to which the burden of debt restructuring falls on the domestic creditors. This lack of firm foundations raises concerns since unless a program provides a path for a country to regain market access, the IMF does not have a strong basis to conclude that the program is addressing the underlying problems (Hagan and others, 2017; Guscina and others, 2017).
Experience with Debt Operations
Although limited in numbers, the 12 market debt operations in the evaluation period have been diverse in terms of modality and coverage of debt, reflecting a variety of country-specific factors related to creditor participation, cross-border spillovers, domestic financial stability, and social consequences. Preserving the health of the financial sector received significant attention when debt operations involved debt owed to domestic financial institutions. For financial stability concerns and other strategic reasons (e.g., political concern on distributional consequences for domestic creditors), differential treatment of creditors and debt instruments was common. As arrears existed in all cases and were very large in some cases, program conditionality generally targeted the clearance of arrears as part of overall debt restructuring.
Experience shows that while completed debt operations ultimately brought significant debt relief, negotiations of debt restructuring packages can adversely affect credit availability and confidence, with an adverse short-term impact on growth. Negotiations can be particularly disruptive if extended by technical difficulties in reaching agreement with creditors, litigation, or political concerns (IMF, 2020a).63 The increasingly heterogeneous creditor base has also complicated efforts to proceed promptly with debt negotiations. Prolonged litigation made the output costs of default felt for a longer period (e.g., in Antigua and Barbuda), while a confrontational stance with holdout creditors could compromise a country’s ability to raise new financing in international capital markets. In some cases (e.g., Grenada), capacity constraints limited authorities’ ability to develop and implement a debt restructuring strategy, delaying debt operations longer than expected. Most debt operations in the evaluation period involved default and accumulation of arrears with external creditors, adding to risks of delayed debt operations.
In this context, various legal approaches were used to facilitate creditor participation, limit holdouts, and avoid litigation. To increase creditor participation, for example, innovative contractual design and instruments were used including value recovery instruments and countercyclical and state contingent payouts (IMF, 2020b). In some cases, an aggregate collective action clause was retrofitted for restructuring of local-law debt (e.g., Barbados and Greece). In other cases, authorities modified financial regulations to make debt restructuring more palatable (e.g., Jamaica and St. Kitts and Nevis).
Experience with regaining market access was mixed. Greece, Grenada, and Jamaica all took longer than expected to regain market access, but for Côte d’Ivoire, Cyprus, and Ukraine regaining access was faster than expected. Despite successful implementation of debt restructuring, Grenada was judged to remain in debt distress due to the non-completion of all debt restructurings and the existence of arrears to official creditors, which delayed market access. In Jamaica (2010), the limited debt relief created a repayment bunching, resulting in delayed market access and leading to the need for a subsequent operation.
Adjustment and Growth Outcomes in Programs with Debt Operations
The time profile of adjustment and growth was quite different between GRA programs with and without debt operations during the evaluation period.64 Debt outcomes of the programs with debt operations were on average better than projected and those of other GRA programs (Figure 43, Panel C). While initial debt was far higher in programs with debt operations, debt ratios were on average put on a broadly declining trend while the opposite was the case for other GRA programs. Success in putting debt on a declining path in part reflected that in programs with debt operations, fiscal adjustment (measured as the change in the primary balance) was stronger and more front-loaded than in other GRA programs (Figure 43, Panel B).
Adjustment and Growth Trajectories: Programs with and without Debt Operations
Source: Erce (2021).Note: Based on the data for 10 programs with market debt operations, excluding Barbados (2018) and The Gambia (2017), and 42 GRA programs without debt operations. Outcomes and projections represent cross-country medians. All projections are initial projections at program approval (T). Data availability is not uniform across periods. Due to the presence of successor programs for some countries in the sample, there is overlap in the data presented over the period and, therefore, the results are not always fully consistent with those based on program periods only.Adjustment and Growth Trajectories: Programs with and without Debt Operations
Source: Erce (2021).Note: Based on the data for 10 programs with market debt operations, excluding Barbados (2018) and The Gambia (2017), and 42 GRA programs without debt operations. Outcomes and projections represent cross-country medians. All projections are initial projections at program approval (T). Data availability is not uniform across periods. Due to the presence of successor programs for some countries in the sample, there is overlap in the data presented over the period and, therefore, the results are not always fully consistent with those based on program periods only.Adjustment and Growth Trajectories: Programs with and without Debt Operations
Source: Erce (2021).Note: Based on the data for 10 programs with market debt operations, excluding Barbados (2018) and The Gambia (2017), and 42 GRA programs without debt operations. Outcomes and projections represent cross-country medians. All projections are initial projections at program approval (T). Data availability is not uniform across periods. Due to the presence of successor programs for some countries in the sample, there is overlap in the data presented over the period and, therefore, the results are not always fully consistent with those based on program periods only.As to the growth trajectory, growth on average rebounded sharply from a deep trough at T–1 in programs with debt operations, with the U-shaped pattern being much sharper than that of other GRA programs in both growth projections and outcomes (Figure 43, Panel A). Growth outcomes of the programs with debt operations on average slightly underperformed initial projections in early years of the program (T and T+1) but exceeded projections in later years. As with debt and fiscal outcomes, growth outcomes ranged widely across programs with debt operations as indicated by the interquartile range in shade, which is significantly wider than that for other GRA programs.
Overall, IMF-supported programs with market debt operations had only mixed success in terms of debt sustainability and the BOP position after the program. Half were followed by successor programs. And in half the cases, either follow-on debt operations were needed or debt ended up in distress or at risk according to the DSA. Where the debt operations fell short of restoring debt sustainability, the underlying reasons varied, including insufficient debt relief, lack of technical expertise, shortfalls in fiscal adjustment, political pressures and, in some cases, the discovery of previously undisclosed debts.
In broad terms, debt operations with principal haircuts and upfront fiscal adjustment were more successful in reducing debt than those with just debt reprofiling or lowered coupons. Specifically, programs with debt operations were able to reduce debt on average by 14 percent of GDP over the 3-year horizon following program approval, with debt reduction of 21 percent of GDP in programs with principal-based operations compared to an average increase in debt of 3 percent of GDP in programs with debt reprofiling, although this is an unfair comparison because it does not take into account the impact of coupon reduction on the net present value of the debt (Figure 44).
Debt Outcomes of Programs with Debt Operations
(Three-year cumulative change in debt from program approval; in percent of GDP)
Source: Erce (2021).Note: Based on the data for 10 programs with market debt operations, excluding Barbados (2018) and The Gambia (2017). “Without successor” stands for programs not followed by a successor program in less than 3 years; “Debt < projected” denotes programs where actual debt is less than projected; Public debt data represent the face value and do not take account of additional reduction in NPV terms.Debt Outcomes of Programs with Debt Operations
(Three-year cumulative change in debt from program approval; in percent of GDP)
Source: Erce (2021).Note: Based on the data for 10 programs with market debt operations, excluding Barbados (2018) and The Gambia (2017). “Without successor” stands for programs not followed by a successor program in less than 3 years; “Debt < projected” denotes programs where actual debt is less than projected; Public debt data represent the face value and do not take account of additional reduction in NPV terms.Debt Outcomes of Programs with Debt Operations
(Three-year cumulative change in debt from program approval; in percent of GDP)
Source: Erce (2021).Note: Based on the data for 10 programs with market debt operations, excluding Barbados (2018) and The Gambia (2017). “Without successor” stands for programs not followed by a successor program in less than 3 years; “Debt < projected” denotes programs where actual debt is less than projected; Public debt data represent the face value and do not take account of additional reduction in NPV terms.Turning to growth outcomes, the evidence suggests that more effective debt operations have on average been associated with better growth outturns in both program and post-program periods (Figure 45). Specifically, both within-program and post-program growth outcomes (during the five years after the program ended) relative to the growth benchmark discussed in Chapter 3 have been superior for: (i) operations based on principal reduction rather than reprofiling; and (ii) programs where debt trajectory has been better than projected during the program.65
Growth Outcomes of Programs with Debt Perations
(Average annual deviation from growth benchmark)
Source: Erce (2021).Note: Based on the data for 10 programs with market debt operations, excluding Barbados (2018) and The Gambia (2017). Post-program period spans fve years after the program ended. See Figure 44 for the definition of the categories on the horizontal axis.Growth Outcomes of Programs with Debt Perations
(Average annual deviation from growth benchmark)
Source: Erce (2021).Note: Based on the data for 10 programs with market debt operations, excluding Barbados (2018) and The Gambia (2017). Post-program period spans fve years after the program ended. See Figure 44 for the definition of the categories on the horizontal axis.Growth Outcomes of Programs with Debt Perations
(Average annual deviation from growth benchmark)
Source: Erce (2021).Note: Based on the data for 10 programs with market debt operations, excluding Barbados (2018) and The Gambia (2017). Post-program period spans fve years after the program ended. See Figure 44 for the definition of the categories on the horizontal axis.Finally, it is worth recalling the results reported in Chapter 4 from growth regressions that while a reduction in the public debt-to-GDP ratio supports post-program growth, the debt operation itself can have a lingering adverse effect, presumably through the impact on borrowing costs and market access. This provides a reminder that the process of debt restructuring can affect growth outcomes. Some supporting evidence is provided in the literature, which suggests that pre-emptive negotiations, which avoid accumulating arrears, result in lower overall output costs from debt strains (Asonuma and Trebesch, 2016), and that hard defaults are more damaging for growth (Trebesch and Zabel, 2017).
Lessons from Country Experience
The country case studies provide more support for the view that debt operations that involved upfront principal reduction have led to more decisive and credible impact on debt sustainability than otherwise, with favorable growth implications. In Grenada, debt restructuring involved haircuts on a wide array of government debts as well as maturity extension which, taken together, resulted in the total NPV haircuts on the order of 50–60 percent, and allowed for a large reduction in the overall debt-to-GDP ratio. The successful operation helped to sustain domestic support for the program as well as reducing the debt burden on the economy. In Ukraine, debt restructuring involved principal reduction aimed at reducing the debt-to-GDP ratio by 20 percentage points, which helped the country to regain its access to international capital markets within two years at reasonable costs.
By contrast, in Jamaica where most of the debt was owed to domestic financial institutions, major debt restructuring with principal haircut was constrained because of concerns about potential financial stability risks. Debt operations were eventually undertaken in two rounds and focused on lowering interest rates and maturity extension in order to preserve the health of the domestic financial system, while excluding external debt from restructuring for concerns on insufficient creditor participation as well as future market access. With more limited reduction of the debt burden, sustained large primary surpluses have been necessary to restore debt sustainability.
Country experience illustrates how delays in debt negotiation or agreement on terms insufficient to restore debt sustainability can be detrimental in terms of regaining market access and restoring confidence and investment. Grenada (2010) and Ukraine (2014) did not include debt restructuring operations, even though debt vulnerabilities were acknowledged as rising, while Jamaica needed two rounds of restructuring as the first round was sufficiently aggressive and fiscal slippage occurred after the 2010 restructuring. Drawing on an earlier IEO evaluation (IEO, 2016), in Greece, pressures from European partners related to contagion concerns and the adverse impact on creditor balance sheets delayed debt operations and then limited the ability of the Fund and country authorities to set a realistic restructuring envelope based on the DSA framework. With Greece’s commitment to membership of the euro area, delayed debt operations and related uncertainty imposed a high toll on growth and social inclusion.
Country case studies suggest that tailored design features can be helpful to ensure adequate creditor participation and to provide greater growth resilience in the face of adverse shocks. In Grenada, for instance, principal-based debt restructuring entailed two novel features: (i) a hurricane clause which provides for immediate debt moratorium in the event of another natural disaster, and (ii) a revenue-sharing clause in new bond contracts. These features helped to achieve wide creditor participation by providing creditors with upside potential while increasing resilience by making debt services contingent on shocks to growth.
Assessment
The IMF played a positive role in supporting debt operations in the program context during the evaluation period. Financing assurance and lending into arrears policies as well as the catalytic role of the Fund have provided balanced incentives on both debtors and creditors to engage constructively in debt operations to restore debt sustain-ability. Beyond this, the IMF’s main contribution has been technical advice based on its DSA frameworks, which have been used to determine the extent of debt relief needed to restore debt sustainability. In addition, Fund staff has often played a role in encouraging creditor participation in debt negotiation by communicating with stakeholders about a country’s economic policies and prospects and debt servicing potential.
Reviews of experience suggest that debt operations with principal haircuts supported by upfront fiscal adjustment tended to be more successful in restoring debt sustain-ability and supporting growth than those with just debt reprofiling or lowered coupons, which delivered too limited NPV reduction. Although reprofiling operations can be effective if accompanied by sufficiently committed fiscal adjustment, this was clearly a harder route and took longer to demonstrate success.
Overall, IMF-supported programs with market debt operations included in the evaluation period have had only mixed success in strengthening debt sustainability and improving the BOP position over the medium-term. This experience in programs with debt operations confirms that successful debt operations can contribute to progress in lowering debt trajectory and restoring growth—but that debt operations that are “too little and too late” can fail to achieve these goals.66
This conclusion raises the question of whether the IMF should be more demanding in ensuring that debt operations in the program context achieve their objectives in terms of debt sustainability and providing a stronger basis for growth. In some cases, with hindsight, it seems that the IMF should have insisted that more ambitious debt operations were needed upfront to address debt sustainability concerns in order to qualify for financing. Steps to make the DSA frameworks more rigorous in the recent revisions to the MAC DSA and LIC-DSF may help to provide a more effective basis for the IMF to insist on more timely and adequate debt operations as a condition for access to Fund financing. However, further attention could be paid to reflecting more systematically how debt operations may affect market access and growth prospects, particularly if debt operations involve default on external debt and a restructuring of domestic debt owed to financial institutions and social security systems.
The potential growth consequences of specific design features of debt operations could also receive more attention. For example, while it may be helpful to secure high creditor participation by sharing some upside with creditors, if such features are too generous, they could backfire and make it more difficult to grow out of debt. Also, it can be helpful to introduce counter-cyclical features in restructured debt, including to automatically adjust debt service obligations in the event of natural disasters which can enhance growth resilience in the face of shocks.
