The International Architecture for Resolving Sovereign Debt Involving Private-Sector Creditors-Recent Developments, Challenges, And Reform Options
Author:
International Monetary Fund
Search for other papers by International Monetary Fund in
Current site
Google Scholar Close
Publication Date:
01 Oct 2020
eISBN:
9781513557472
Language:
English
Keywords:
PP; exchange offer; sovereign bond; debt transparency; vulture fund debt speculation; restructuring negotiations; debt restructuring; domestic law; debt service; foreign currency; enhanced CACS; debt management; secondary market; Bonds; International bonds; Sovereign debt restructuring; Debt restructuring; Sovereign bonds; Asia and Pacific; private-sector debt resolution; debt crisis; debt relief
Download PDF (2 MB)

There have been significant developments in sovereign debt restructuring involving private-sector creditors since the IMF’s last stocktaking in 2014. While the current contractual approach has been largely effective in resolving sovereign debt cases since 2014, it has gaps that could pose challenges in future restructurings.

Abstract

There have been significant developments in sovereign debt restructuring involving private-sector creditors since the IMF’s last stocktaking in 2014. While the current contractual approach has been largely effective in resolving sovereign debt cases since 2014, it has gaps that could pose challenges in future restructurings.

Introduction

1. In response to a request from the G20 International Financial Architecture (IFA) Working Group, this note discusses the architecture for the resolution of sovereign debt problems involving private-sector creditors. Sovereign debt is the only debt class without a bankruptcy mechanism and the international community has relied on the contractual approach to prevent and resolve sovereign debt problems.1 The note highlights recent experience with sovereign debt restructurings, changes in the instruments and creditor base, and contractual and statutory techniques that have been used to address collective action problems. It then identifies challenges in the current system and presents reform options. This note does not cover official sector involvement that forms the other part of the sovereign debt restructuring and affects the resolution of sovereign debt problems involving private creditors, or the coordination between official-sector and private-sector sovereign debt restructurings.

2. In 2014, as part of the IMF’s work program on sovereign debt launched in 20132, the IMF Executive Board analyzed certain aspects of developments in the contractual framework for sovereign debt restructurings.3 The IMF endorsed key features of enhanced contractual provisions for international sovereign bonds to strengthen the contractual framework to address collective action problems. These included strengthened collective action clauses (CACs) that include a menu of voting procedures, including the “single-limb” voting mechanism, as well as modified pari passu clauses. IMF staff has been tasked with promoting the inclusion of these provisions in international sovereign bonds.

3. The architecture has improved and has generally performed well in recent years, but the system has gaps that could pose challenges, particularly if sovereign debt distress were to reach systemic levels. Recent restructurings have generally proceeded smoothly, were largely preemptive, and had high creditor participation. The uptake of enhanced collective action clauses has been strong. However, enhanced CACs have only just begun to be used in sovereign debt restructurings and there is still a large outstanding stock of international sovereign bonds without these clauses. Sub-sovereign debt, which may also need to be restructured alongside debt of the sovereign, may pose additional challenges. In many low-income countries (LICs), there is a significant stock of non-bonded privately-held debt, which does not have majority restructuring provisions for payment terms. Finally, there have been changes in the structure of claims—including growing use of collateralized and collateral-like debt instruments—and a diverse and more fragmented creditor base, that together make debt restructurings more challenging.

4. There are several options that could be used to strengthen the existing architecture. Beyond the continued inclusion of enhanced CACs in new international bonds, these include legal innovations such as the development of model clauses to facilitate modification of payment terms in loan agreements by a majority of creditors and the promotion of their use, greater use of trust structures, and increased issuance of state-contingent bonds to protect debtors from downside risks. There is also a question about the desirability of wider use of carefully tailored “anti-vulture fund” legislation of the kind adopted in a few countries that limits holdout creditor recovery to complement the contractual approach, taking into account the important legal and policy issues that these provisions raise. Targeted domestic law tools or international law options (such as a UN Security Council resolution as used previously in one case), both as have been proposed elsewhere, may also need to be considered only as a last resort and on a time-bound basis if a COVID-related systemic sovereign debt crisis were to materialize that cannot be effectively addressed by the existing resolution toolkit. Moreover, given the role of IFIs in facilitating speedy and orderly debt restructuring, the planned reviews of key IMF policies could lead to further reforms and consideration could also be given to a review of the effectiveness of relevant policies of other IFIs. Finally, the international community should further support enhancing debt transparency, and help countries strengthen their public debt management frameworks ex ante through, for instance, technical assistance.

5. The rest of the paper is organized as follows. Section I discusses the evolution of the sovereign debt landscape, including uptake of enhanced contractual provisions, developments in the instrument and creditor base, and statutory tools that have been used to address collective action problems. Section II identifies challenges and gaps for the framework for sovereign debt restructurings, and Section III lays out reform options to strengthen the resolution toolkit going forward. Section IV focuses on potential last resort responses should a systemic crisis arise in the interim. Finally, Section V concludes with ways to strengthen the current contractual framework and highlights areas in which further work is needed.

Section I. Evolution of the Sovereign Debt Landscape

6. Since the IMF’s last stocktaking in 2014, there have been significant changes in the contractual framework for sovereign debt restructuring, as well as over a dozen completed and forthcoming restructurings involving private claims. Enhanced CACs were introduced in 2014 and have by now been widely adopted in recently issued international bonds. Argentina (2020), Barbados (2019), Belize (2017), Chad (2018), Ecuador (2020), Grenada (2015), Mongolia (2017), Mozambique (2019), and Ukraine (2015) have restructured privately-held sovereign debt and others, such as Republic of Congo, Lebanon, Venezuela, and Zambia, are forthcoming or underway (Table 1).4 In addition, there have been completed (as well as ongoing and forthcoming) restructurings at the sub-sovereign level (e.g., City of Kiev, Puerto Rico, a number of Argentine provinces, and PDVSA, the Venezuelan state oil company). Several of these raised novel issues, as discussed below.

Table 1.

Sovereign Debt Exchanges in 2014–20

article image
Sources: Anthony, Impavido and van Selm (2020), Asonuma and Trebesch (2016) updated dataset, Cruces and Trebesch (2013) updated dataset, IMF country reports, Moody’s (2017), and relevant bond prospectuses. Notes:

Asonuma and Trebesch (2016) updated dataset. Start and end of restructurings correspond to either the announcement or default, and completion of exchange, respectively.

Anthony, Impavido and van Selm (2020) for Barbados domestic debt episode. Staff calculations for Argentina and Ecuador, Asonuma, Niepelt and Ranciere (2018) updated dataset for all remaining episodes. Weighted average (respect to debt outstanding) of instrument-specific NPV and market haircuts. NPV and market haircuts correspond to 1 – (PV of new bonds/PV of old bonds), and 1 – (PV of new bonds/Face value of old bonds), respectively.

Participation rate after collective action clauses (CACs) of bonds were triggered.

Barbados adopted legislation that retrofitted a collective action mechanism into domestic debt in September 2018.

Enhanced CACs are also included in all but one of the original bonds. CACs were triggered in all bonds in the exchange to achieve full participation.

While CACs are included in all original bonds, enhanced CACs are included in some of original bonds.

IMF staff estimate.

A. Recent Sovereign Debt Restructurings

Aided by CACs, recent sovereign debt restructurings have generally proceeded smoothly, were largely preemptive, and had shorter average duration and higher average participation than in previous periods.

7. Compared with earlier periods, sovereign debt restructurings more often have been preemptive, have been much shorter in duration, and have obtained higher participation on average due to the use of CACs. Since 2014, of the nine debt restructuring cases, a majority were preemptive (i.e., pre-default), while four cases (Argentina, Barbados, Grenada, and Mozambique) were post-default5 Prior to 2014, post-default restructurings had been more common than preemptive ones. Eight cases involved restructuring of mainly external debt, while two (Barbados and Grenada) restructured primarily domestic debt (i.e., debt issued under domestic law). Bonds continued to be the dominant type of debt restructured; Chad was the only case where only loans were restructured. The duration of restructurings—from either announcement or default until debt settlement—was 1.2 years on average. This is shorter than the average duration of 3.5 years for privately-held external debt restructurings over 1978–2010.6 Among the cases occurring since 2014, CACs were used to achieve full participation in over half of cases, and no ex-post litigation with private creditors has arisen in cases where CACs were used.

8. Some sovereign debt restructurings have incorporated state-contingent features in new bond contracts to incentivize creditor participation and to protect the sovereign from future downside risks (see Table 2).7

  • Value recovery instruments (VRIs) (such as GDP-linked warrants) that provide upside to creditors in good economic scenarios can help bridge creditor-debtor differences regarding economic growth and debt-serviceability. This in turn may allow for appropriately conservative base-case payouts that minimize the risk of future defaults. However, creditors have historically discounted these instruments due to their illiquidity and idiosyncratic risk profiles, potentially limiting their usefulness. After having been used in the earlier restructurings of Argentina (2005 and 2010) and Greece (2012), GDP warrants were used in the recent restructuring of Ukraine (2015). An upside instrument tied to revenues from a citizenship-by-investment program was used in Grenada (2015). In some of these cases, payouts may ultimately prove to be quite large in comparison to the instruments’ initial valuations.8

  • Debt restructurings in Grenada (2015) and Barbados (2018) have provided opportunities to include hurricane and natural disaster clauses that allow for automatic maturity extensions and interest forbearance following severe shocks. These clauses provide valuable insurance at low cost against exogenous risks and may be increasingly important given growing climate risks and other environmental concerns. A recent initiative promoted by the Eastern Caribbean Central Bank (ECCB), involving introduction of “hurricane or natural disaster clauses” into debt instruments, led to the production of a draft term sheet for these clauses by the International Capital Market Association (ICMA).9

Table 2.

State-Contingent Instruments Issued

article image

These haircut calculations do not account for the value of the state-contingent instruments.

Payments to be discounted back to May 2015 using average yield on the 2030 bond in the year in which they occur.

The Glencore loan to Chad that was restructured in 2018 includes specific contingencies to reduce the debt payments in low oil revenue scenarios and accelerate the payment of debt in case of high oil revenue scenarios.

Similar clauses were included in restructured debts with the Import-Export Bank of Taiwan, Province of China and the Paris Club.

Only for debt held by private (domestic and external) creditors.

9. The G20’s Debt Service Suspension Initiative (DSSI), launched in April 2020, has not yet resulted in any deferral of debt service to the private sector.10 Rescheduling debt service on DSSI-comparable terms would generally have resulted in losses to private creditors, as market rates at the time typically exceeded interest rates embedded in outstanding bonds and commercial loans. Although the Institute of International Finance (IIF) attempted to provide a framework to facilitate private creditor participation, voluntary private sector participation in the DSSI required the debtor country to make a request to its private creditors. It appears that in this context, the potential benefits of such a request (a limited debt service suspension) have been viewed by many sovereigns as smaller than the potential costs (see Box 1).11 This does not need to be the case in other contexts, such as a deep debt restructuring rather than reprofiling, or where private-sector involvement is a requirement for official sector support (as shown in the recent case of Ecuador, whose international bonds were reprofiled through a consent solicitation).12

Private-Sector Involvement under the G20’s DSSI

On April 15, G20 countries endorsed the Covid-19 Debt Service Suspension Initiative (DSSI). The initiative covers 73 International Development Association (IDA) and UN Least Developed Countries that are current on their debt service to the IMF and WB. Under the initiative, official bilateral creditors commit to suspend debt service payments due between May 1 and December 31, 2020. New repayments are due to begin in June 2022 and are phased over three years in semi-annual installments. The initiative is understood to be “net-present value (NPV)-neutral” in that there is no reduction in the nominal amount of principal or interest and the contractual rate of interest will be paid on deferred amounts. Under the DSSI, private creditors are called upon but not required to participate on comparable terms.

Despite appeals by the official sector and a coordination attempt led by the IIF, private creditors have not participated in the DSSI based on available information. The IIF has produced Terms of Reference for private-sector participation on a voluntary basis, but this has not been used. The reasons include lack of incentives on both the creditor and the debtor side.

Lack of incentives on the side of private creditors. At the time of the launch of the DSSI, rescheduling debt service on comparable terms (NPV neutrality, based on using the prevailing contractual interest rate as the discount rate) would for the most part have implied extending maturities at below-market interest rates. As a result, private creditors were not inclined to voluntarily offer debt service suspensions.

  • Lack of incentives on the side of debtor countries. While debtors stood to benefit from a private-sector debt service suspension, these benefits appear to have been generally outweighed by reputational concerns, fears of ratings downgrades, transactions costs, loss of market access, and concerns about adverse legal implications (in some combination).

  • Reputational concerns. Since private-sector participation on DSSI terms would have inflicted losses on most private creditors, some debtor governments may have feared that they could have led to a loss of market access or more expensive debt issuances.

  • Ratings downgrades. While no credit rating agency has downgraded any country merely for requesting the DSSI, one agency (Moody’s) placed several participating countries on a negative watch, citing the G20’s call for private-sector creditors to participate in the DSSI on comparable terms. Furthermore, all three major credit agencies have made it clear that requesting private-sector participation on G20-comparable terms could lead to a downgrade (although this might be temporary).

  • Legal risks. Depending on the terms of private debt agreements, requesting debt service suspension under the DSSI or comparable relief from private creditors could potentially trigger event of default clauses and possibly cross-default clauses in private debt contracts, which could lead to acceleration of those contracts, as well as litigation.1/ To help mitigate such concerns, the IIF has drafted a model waiver agreement that countries may use to obtain waivers from their private-sector creditors of any possible event of default that could arise from seeking debt relief from its official-sector creditors under the DSSI. Similar waivers could be sought for debtors seeking relief from private-sector creditors as well.

1 Bond contracts and loan agreements typically contain cross-default or cross-acceleration clauses. With respect to bond contracts, these clauses are typically only triggered upon an event of default or acceleration on other external indebtedness which may be broadly (any foreign currency lending) or more narrowly (any publicly traded foreign currency denominated instrument) defined. Loan agreements may be more broadly worded, such that even a voluntary rescheduling of other external indebtedness may give rise to an event of default under the loan agreement.

B. Instruments and Creditor Base

Since 2014, total sovereign debt has increased as a share of GDP, and the range of debt instruments has proliferated. Coupled with a continued diversification of the creditor base, these developments imply that future restructurings may face greater challenges achieving high participation.

10. Total external government debt of emerging market and developing economies (EMDE) has increased from 14 to 24 percent of GDP between 2008 and 2018, driven by increased debt owed to the private sector (Figure 1). Debt issued under domestic law has also increased.

  • The fall in the share of bilateral loans in total external debt by 7 percentage points is broadly matched by the rise in the share of commercial borrowing by 8 percentage points.

  • Bank loans have declined in overall importance for commercial general government borrowing and as of 2018 make up only 4 percent of the stock of emerging market commercial debt. In the 73 DSSI-eligible lower-income countries, non-bonded debt comprises 39 percent of aggregate public and publicly guaranteed (PPG) debt, and continues to provide the majority of commercial financing for many LICs, as only the largest and most developed have access to international bond markets (Figure 2).13

  • The share of external debt from state-owned enterprises (SOEs) has remained steady at 12 percent, although inadequate coverage of SOEs is one of the main gaps in public external debt data.14 Similarly there does not appear to be an upward trend in public guarantees of privately-issued debt according to the World Bank data; but coverage is likely to be an issue in this area, too. Sub-sovereign entities—both SOEs and provincial governments—increasingly issue foreign law-governed bonds in New York and England, the two primary jurisdictions of issuance (Figure 3).

  • Local currency debt of governments has continued to grow relative to 2017 across EMs and LICs and remains the majority share of total public debt in EMs.15 With respect to governing law, for advanced economies almost all debt is governed by domestic law; for emerging markets, there is more variance across countries (Figure 4).16 Countries with the largest share of local currency government debt are in the Asia Pacific region. On a regional basis, the most significant increases have occurred in the Asia Pacific and the Africa and Middle East regions. However, significant variation among countries exists regarding the importance of their local currency government bond markets. The countries with the lowest local currency debt shares are LICs outside the Asia Pacific region, with sovereign issuers in Africa and the Caribbean more heavily reliant on commercial foreign-currency debt (Figure 5). While debt governed by domestic law is often easier to restructure than international debt from a legal perspective, restructuring such debt can negatively impact financial stability (for example, due to the bank-sovereign nexus) and lead to further debt difficulties down the road.17

Figure 1.
Figure 1.

EMDE External Public and Publicly Guaranteed Debt by Creditors 1/

Citation: Policy Papers 2020, 043; 10.5089/9781513557472.007.A001

Sources: International Debt Statistics and IMF staff calculations.1/ Based on a sample of 56 EMDEs that report their government debt stocks to the World Bank as part of their borrowing agreements.
Figure 2.
Figure 2.

DSSI Countries: External Bonded and Other Non-official PPG Debt, 2018

(in billions of USD)

Citation: Policy Papers 2020, 043; 10.5089/9781513557472.007.A001

Note: Out of 73 DS5I countries, 23 are excluded for not having any private debt.
Figure 3.
Figure 3.

SOE and Sub-National International Bond Issuance

Citation: Policy Papers 2020, 043; 10.5089/9781513557472.007.A001

Figure 4.
Figure 4.

Government Debt Securities Issues by Jurisdiction, end 2017

Citation: Policy Papers 2020, 043; 10.5089/9781513557472.007.A001

Source: Arslanalp (2018)
Figure 5.
Figure 5.
Figure 5.

Developments in EM Local Currency Government Debt

Citation: Policy Papers 2020, 043; 10.5089/9781513557472.007.A001

11. Collateralized bond and syndicated loans issued by EMDE public sector make up about 15 percent of EMDE bond and syndicated loan issues since 2002, and issuance has been rising since the beginning of the decade (although it remains below its peak at the global financial crisis) (Figures 6 and 7).18 A debt instrument is collateralized when the creditor has a lien over an asset or revenue stream that would allow it to rely on the asset or revenue stream to secure repayment of the debt in case of default. For certain types of transactions, collateral directly related to the purpose of the financing is quite common (e.g., project or acquisition financing); but there are also transactions that involve unrelated collateral (e.g. budgetary borrowing secured by commodity exports). The enforceability of collateral depends on the type of collateral and the governing law of the jurisdiction where it is located.19 From the perspective of the borrower, collateral can have both benefits (lower borrowing costs on collateralized debt) and costs (debt is much harder to restructure in the face of a bad shock; borrowing on non-collateralized debt may become more expensive). Related collateral is generally preferable to unrelated collateral, because it preserves some risk sharing and makes it less likely that collateralized lending will be used for consumptive purposes (see Box 2).

Figure 6.
Figure 6.

Public Sector Collateralized Bonds and Syndicated Loans

(US$ billion)

Citation: Policy Papers 2020, 043; 10.5089/9781513557472.007.A001

Source: Dealogic.
Figure 7.
Figure 7.

Collateralized Bonds and Loans as a share of total Syndicated Bonds and Loans, average 2002–17

Citation: Policy Papers 2020, 043; 10.5089/9781513557472.007.A001

Source: Dealogic, Kaufmann, Kraay and Mastruzzi (2010). And Standard and Poor’s.

Welfare Implications of Collateralized Borrowing

Whether collateralized borrowing helps or hurts the borrowing country depends on the form of collateral and on the underlying problems (“distortions”) that complicate international borrowing and its domestic impact. To the extent that the main distortion is limited (or costly) enforcement of sovereign debt repayment, collateral eliminates the possibility of opportunistic default. This benefits the debtor through lower borrowing costs. At the same time, collateral also makes restructuring in response to an adverse shock more difficult (if not impossible). Perfectly enforceable collateral eliminates the possibility of sharing risk with the creditor by rescheduling the debt in the face of a bad shock. The welfare implications of these two effects run in opposite directions, and hence the overall welfare effect is often ambiguous.

One specific form of collateralization that is generally welfare improving is borrowing collateralized using the future flow receivable generated by the project that is being financed through collateralized debt. This insures the creditor against opportunistic default while still preserving risk sharing, as there will be no collateral if the project fails. For this reason, the IMF and World Bank have generally considered related collateral to be less problematic than collateral unrelated to the underlying transaction.1

The welfare implications of collateral can be complicated by the presence of additional distortions, such as weak governance (or government overconsumption) in debtor countries, lack of transparency, and debt dilution. In these settings: collateral (1) gives incumbents more leeway to borrow cheaply, saddling future generations with secured debt that is much harder to renegotiate than non-collateralized debt; (2) can significantly raise the borrowing cost on unsecured debt, both because unsecured borrowers may not know how much debt is collateralized, and because they may fear that future collateralized borrowing increases the loss-given-default of their claims. As discussed below, negative pledge clauses can help in this regard.

1/ International Monetary Fund and World Bank 2020a, Collateralized Transactions: Key Considerations for Public Lenders and Borrowers.

12. In recent years, some sovereigns have borrowed using sovereign repurchase (repo) agreements, which have features similar to collateralized agreements.20 In these contracts, bonds newly issued by the debtor or gold are used as collateral for cash lending, typically through an investment bank which the sovereign commits to repurchase at the end of the term (typically 3–4 years). These arrangements generally involve a high degree of overcollateralization (sometimes in excess of 100 percent) and require borrowers to post mark-to-market collateral against a decrease in the value of the underlying collateral, helping to protect the creditor against many default scenarios. Repos may also allow sovereigns to raise funds more quickly than through a formal issuance process, taking advantage of lower interest rates. Sovereigns have also noted that these arrangements help them avoid saturating their institutional investor base, especially if they face large issuance needs over short periods. However, repos can lead to significant problems during periods of stress, when the sovereign’s creditworthiness deteriorates, bond prices fall (as investors factor in their potential dilution in the default scenario), and borrowers must post variation margin while facing liquidity pressures. In recent episodes of sovereign stress, such sovereign repos were repaid (e.g., Argentina, Ecuador) rather than restructured, demonstrating the leverage of such de facto secured creditors at the expense of all other creditors.

13. Fragmentation of the creditor base contributed to challenges in creditor coordination in recent bond restructurings but were managed in these cases. In the Argentina debt restructuring, three separate creditor committees were formed, representing in total approximately 40–45 percent of the eligible bonds.21 Other bondholders did not join any of the committees, with one large bondholder preferring to negotiate directly with the government. The lack of a common creditor group appears to reflect the diversity of the creditor base with different interests, instrument holdings—with and without enhanced CACs—and engagement strategies with the authorities.22 Each committee had different views on the approach to the negotiations and the financial and legal terms of the counteroffers, which made coordination difficult. Nonetheless, an agreement in principle with all committees was eventually reached after a number of negotiation rounds that spanned more than six months. In the Ecuador bonded debt restructuring, the creditor base was also diverse and three creditor committees were formed.23 However, the Ecuadorian authorities facilitated creditor coordination and achieved high creditor participation in the exchange due to a transparent engagement strategy with creditors.24 During the restructuring renegotiations, the authorities made it publicly known that the larger creditor committee had already expressed support for the proposal, allowing them to reach bondholder approval of around 60 percent in aggregate prior to the launch of the exchange offer, which helped persuade the remaining bondholders to agree to the exchange.

C. Uptake and Use of Enhanced Contractual Provisions

The uptake of enhanced CACs in new bond issuances continues to be high, though they are still absent in about half of the outstanding stock of international bonds. The first uses of enhanced CACs in sovereign debt restructurings have relied on the two-limb voting mechanism; the single-limb voting mechanism remains unused.

14. Since the IMF’s endorsement of key features of enhanced CACs in October 2014, almost all new international sovereign bond issuances have included such clauses (see Box 3).25 Based on information available as of June 30, 2020, there have been around 690 international sovereign bond issuances since October 1, 2014, for a total nominal principal amount of approximately US$870 billion.26 Of these, approximately 91 percent of new issuances have included the enhanced CACs, as compared with 88 percent as of end-October 2018. While incorporation has been largely the same in the two primary jurisdictions of England and New York and amongst frontier and emerging markets, enhanced CACs have not been included in issuances outside these jurisdictions. In particular, issuances under Chinese and Japanese law do not include these clauses, with Japanese law bonds continuing to use series-by-series CACs, and bonds under Chinese law largely using either series-by-series or two-limb aggregated CACs.27 28

Collective Action Clauses

Collective action clauses are a provision in bond contracts that allow a majority of creditors either within—or across—series of bonds to bind the minority to the terms of a restructuring.1 These provisions limit the risk that a minority of creditors will disrupt an orderly restructuring process by “holding out”—or threatening to hold out—in order to receive payment in full at the expense of the restructured majority.

CACs have evolved over the years through different generations. First generation CACs—which became widespread in NY law-governed bonds beginning in 2003,2 allow a majority of bondholders within a single series (i.e., one bond issuance) to bind the minority to the terms of a restructuring. Second generation CACs include series-by-series voting procedures and a “two-limb” aggregated voting mechanism, which requires a minimum threshold of support be achieved both (a) in each series and (b) across all series being restructured. Third generation CACs are so-called “enhanced CACs,” with a menu of voting procedures: series-by-series voting, two limb aggregated voting, and a “single-limb” aggregated voting which allows a majority of creditors across all series to bind the minority to a restructuring.3 In 2014, the IMF endorsed the key features of the enhanced CACs (Table 3), and staff has been promoting the inclusion of these clauses in international sovereign bonds.4

1/ As in the past, the term “international sovereign bonds” is used in the following sense: (i) a bond is “sovereign” if it is issued or guaranteed by a government or a central bank; (ii) a bond is considered “international” if it is governed by a law other than the law of the issuer or gives a foreign court jurisdiction over any claims that may arise under the bond; and (iii) “bonds” are freely traded debt instruments with fixed maturities, normally in excess of one year. See the 2014 Paper at para. 6. The focus on bonds governed by a law other than that of the sovereign has been motivated by a recognition that, with respect to bonds governed by domestic law, the legal leverage possessed by holdout creditors is more limited given the capacity of the sovereign debtor to modify its domestic law. Id. 2/ English law-governed bonds have included CACs for the past century. 3/ Under the enhanced CACs, in order to use the single-limb voting procedure, all bondholders aggregated together for voting purposes must receive the same bond or the same menu of bonds (i.e., the offer must be “uniformly applicable”). A uniformly applicable offer across all series may be particularly useful when investors are somewhat homogenous across maturities and they think of their bond holdings as a portfolio rather than bond by bond. The single-limb voting procedure also allows for aggregation of groups of bonds together for voting purposes (so called “sub-pooling”), which adds to the flexibility of the procedure and allows a debtor to achieve a level of NPV equivalence amongst bondholder groups. 4/ See the 2014 Paper and the accompanying Press Release No. 14/459.
Table 3.

ICMA Enhanced CACs: Menu of Voting Procedures

article image

15. However, owing to the gradual maturity of pre-existing debt, a sizeable portion of the outstanding stock of international sovereign bonds does not yet include enhanced CACs. While the total outstanding stock of international sovereign bonds without enhanced CACs continues to decline slowly through amortization, about 50 percent of outstanding bonds as of end-June 2020 still do not include them. Moreover, it will take some years for all existing bonds without enhanced clauses to mature. About 30 percent of those bonds will mature in more than 10 years (about 40 percent of which are below investment grade). However, as noted above, almost all international sovereign bonds include some forms of CACs which, despite the absence of the single limb voting mechanism, could help facilitate debt restructurings.

16. In a significant step towards market harmonization, Euro-Area finance ministers recently agreed in principle to incorporate updated CACs with a single-limb voting mechanism into all Euro-Area sovereign bonds as of January 1, 2022.29 In line with enhanced CACs, the clauses will include a single-limb voting mechanism (as well as a series-by-series voting procedure) but will not include the two-limb variation.

17. Recent empirical analysis has found that bonds with CACs trade at a premium to those without them, indicating their value to investors. Earlier research demonstrated negligible market differentiation due to the inclusion of CACs.30 A recent paper by Chung and Papaioannou (2020) that examines the now substantial pool of bonds with enhanced CACs in the post-2014 period has found bonds with CACs to trade at significantly lower yields in secondary markets. This is particularly true for non-investment grade bonds, where a collective action clause is presumably more valuable given the greater probability of sovereign default.31

18. Investors appear to have differentiated bonds with enhanced CACs from those with traditional CACs at times of debt distress in recent debt restructuring cases. Market participants have noted that investors focus on CACs and value the bond differently at times of debt distress. For example, in the Ecuador (2020) restructuring, the sole bond (maturing in 2024) without enhanced CACs32 traded at a premium compared to bonds with enhanced CACs (Figure 8). A similar pattern was observed in the Argentine restructuring, where bonds issued during Argentina’s 2005 and 2010 bond exchanges, which had two-limb CACs with higher voting thresholds and provided greater legal protections to creditors generally, traded at higher prices than the bonds issued in and after 2016 (which included enhanced CACs) (Figure 9). While differences in legal provisions may have played a role in the pricing differences, other factors could also have influenced Argentine bond prices: i.e., the 2005/2010 bonds had been previously restructured and contained additional terms whose impact on the bonds’ prices cannot be separated from the price impact of the CACs.33

Figure 8.
Figure 8.

Ecuador: Price Developments of Eurobonds under Restructuring

(in percent)

Citation: Policy Papers 2020, 043; 10.5089/9781513557472.007.A001

Figure 9.
Figure 9.

Argentina: Price Developments of Eurobonds under Restructuring

(in percent)

Citation: Policy Papers 2020, 043; 10.5089/9781513557472.007.A001

19. Enhanced CACs were used for the first time in the recent Ecuador and Argentina bond exchanges. The market haircuts were 41 percent in the case of Ecuador and 50 percent in the case of Argentina.34 In both cases, the two-limb voting mechanism under the enhanced CACs was used. In Ecuador, both limbs of the voting thresholds were met with respect to all aggregated series; in Argentina the CACs thresholds were met with respect to all but two bond series.35 As described in Box 4, the use of certain legal techniques in conjunction with enhanced CACs in both exchanges raised concerns amongst some creditors and legal scholars. Creditors in these two cases have negotiated the inclusion of clauses in the new bonds issued in the exchange that would limit the use of these legal techniques in future restructuring of those bonds. The potentially more robust single-limb voting feature of enhanced CACs has not yet been used.

Ecuador and Argentina Experience with Enhanced CACs

The recent restructurings of bonded external debt in Argentina and Ecuador provided the first case studies for the use of enhanced CACs. Some operational issues that were not envisaged at the time of drafting arose, and the parties were able to agree to specific fixes.

Re-designation. Both restructurings signaled that the debtor reserved the right to “re-designate” at any time—even after the exchange offer closed—which series of bonds would be aggregated together for voting purposes.1/ Creditors raised a number of concerns that re-designation could allow the debtors to “gerrymander” ideal voting pools to maximize the cram down of holdout creditors, which creditors believed could undermine procedural fairness and integrity.

  • To allay these concerns, in both cases, the parties agreed to adopt language in the new exchange bonds to help address creditor concerns for future restructurings, providing that re-designation of voting pools will only be permitted if (i) bondholders are given five business days after the exchange offer closes to withdraw their votes, or (ii) the offer was approved by holders of more than 66⅔ percent of the aggregate principal amount of the originally designated pool.

“Pac Man” strategy. In its exchange documents, Argentina put creditors on notice that in future exchange offers, it could decide to use a strategy that commentators have referred to as “Pac Man” to sweep up creditors who hold out of the current exchange offer. The strategy would work as follows: after concluding its initial exchange, a debtor could launch one or more subsequent exchange offers, with the aim of binding those creditors who had refused to consent to the initial restructuring. It would do so by using enhanced single-limb CACs, which require only an aggregated voting threshold to be met, with no per-series requirement. The debtor could include in subsequent exchanges both (i) those holders who had consented to the initial exchange (and thus have new bonds) and (ii) those holders that held out of the initial exchange (and thus still have old bonds). The expectation is that the debtor would offer slightly better terms than in the initial exchange to entice the creditors with new bonds to consent to the subsequent exchange. If the aggregate voting threshold were met, the erstwhile holdout creditors would be bound regardless of the support in a particular series. Creditors raised concerns about this technique when used in conjunction with re-designation, as it would allow the debtor to effect a restructuring that originally only had garnered support of a minority (or a bare majority) of creditors.

  • To allay these concerns, both Argentina and Ecuador agreed to include language in the new exchange bonds, which provided that in restructurings involving the exchange bonds, the debtor can only “Pac Man” if in the first-round restructuring, holders of more than 75 percent of the aggregate principal amount of all bonds included in the original restructuring offer accept the deal. If the first-round restructuring does not meet this 75 percent threshold, the sovereign must wait at least 36 months before using a single-limb vote to cram down remaining holdouts.

1/ The Ecuador bonds contained the pre-2015 ICMA model clause with no explicit limitations on re-designation. While the Argentine bonds contained the NY ICMA model clause requiring vote pooling to be announced in advance and frozen, the offering documents sought bondholder consent to amend those terms to allow Argentina to re-designate voting pools at any time. While legal scholars have posited that Argentina would have been free to change voting pools during the offer period even without this consent, this would have constituted an amendment in the terms of the offer. As such, Argentina would have been required to offer bondholders that had already submitted their votes the opportunity to withdraw them. The re-designation amendment allowed Argentina to make changes to the voting pools after the closing of the offer, while maintaining the votes as submitted. See Buchheit, Lee and Gulati, Mitu, “The Argentine Collective Action Clause Controversy”, Capital Markets Law Journal (forthcoming 2020).

20. Other legal techniques have been used to complement and bolster the effectiveness of CACs. For instance, Ecuador conditioned its exchange offer on achieving a minimum participation threshold (MPT) of 80 percent after the activation of CACs.36, 37 MPTs are designed to encourage participation by providing that the restructuring will only go forward if a critical mass of creditors also consent, allaying concerns that a participating creditor may end up holding an instrument which (due to low participation) has low value. Another technique used is to couple an exchange with a creditor vote to modify the existing instruments through the use of exit consents. Exit consents can be used to make the existing instruments less attractive to hold and more difficult to enforce by limiting legal remedies, narrowing waivers of sovereign immunity, or eliminating certain creditor protections.38 Ecuador’s use of exit consents, for example, eliminated certain provisions that would have otherwise ensured that the existing bonds could not be left with terms less favorable than the exchange bonds.

21. Both Argentina and Ecuador also used CACs to give less favorable financial terms to holdout creditors, providing an additional incentive to participate in the exchange. Bondholders who did not tender their bonds but are bound by the restructuring through CACs were not given additional compensation in the form of deferred interest on their bonds.39 Moreover, the payment terms in the existing bonds held by non-consenting bondholders in Ecuador were amended to match the longest-dated exchange bond (which matures in 2040). In the case of Argentina, non-consenting holders were mandatorily exchanged for new bonds, which, in some cases, have the least favorable maturity structure and do not contain creditor protections in the event of a future, more favorable offering. This less favorable financial treatment imposed on non-consenting holders is a novel use of CACs and a larger incentive to participate than in the case of typical exit consents, where bondholders may have been left with less favorable non-financial terms, but at least were entitled to their original claim for principal and interest. Here, a non-tendering but participating bondholder has its claim impaired and is treated materially worse than tendering bondholders.40

D. Targeted Statutory Tools

Targeted statutory tools have been adopted in some cases as a complement to the contractual framework to facilitate restructurings by further limiting holdout risks.

22. In addition to contractual techniques, some countries have adopted targeted statutory tools to facilitate restructurings and limit risks posed by litigious holdout creditors. These techniques provide a backdrop against which restructurings occur and thus in principle can help set expectations amongst creditors and debtors in restructuring negotiations. They can also help limit legal risks and the associated costs to debtors and cooperating creditors where contractual restructuring tools fall short.

23. A few countries have adopted domestic legislation designed to discourage holdout creditors by limiting potential profits. Such laws—colloquially referred to as “anti-vulture fund legislation”—some of which have been adopted relatively recently and by only three countries,41 and have not yet been invoked, so there is little evidence to date about their effectiveness.42 The laws adopted vary in scope and intent. Following significant debt relief from official creditors to LICs under the HIPC Initiative, the UK law was motivated by the idea that its court system should not be leveraged to allow other creditors full recovery on debts where comparable treatment was expected. Therefore, the UK law is focused on limiting claims against countries that benefitted from debt relief under the HIPC Initiative; it prevents creditors from suing in UK courts to enforce payment on the pre-2004 sovereign debt of HIPC debtors on terms more favorable than agreed under the HIPC Initiative. By contrast, the Belgian law is broad, applying to the debt of any sovereign, and is focused instead on curtailing enforcement by secondary-market purchasers.43 This law limits in certain circumstances—including where a creditor refuses to participate in a debt restructuring process—a creditor’s ability to seek from a Belgian court enforcement of a claim that the judge determines is clearly disproportionate to the price the debt was purchased for in the secondary market. It is also noteworthy that in 2018, the European Parliament called on member states to adopt “a regulation based on the Belgian law on combating vulture fund debt speculation” and the European Commission has since commissioned a study on whether to recommend such legislation and how best to tailor it.44 However, at present, there appears to be little if any support for this approach. Both the UK and Belgium laws apply to debt contracts entered into before the passage of those laws, while the French law is focused only on debts purchased after its 2016 entry into force and forbids court authorization for seizure of assets of certain beneficiaries of official development assistance under certain conditions after a default or restructuring announcement.

24. Two countries, Barbados and Greece, have also used domestic law to retrofit collective action mechanisms into domestic-law debt. Where sovereign debt is governed by domestic law, debtor authorities have the ability to make amendments to domestic law (the so-called “local-law advantage”) to retrofit debt with collective action mechanisms. This has been used to allow debtors to enable the majority of creditors to agree on a restructuring by way of a single-limb vote, ensuring a smoother domestic restructuring by binding minority holdout creditors to the terms of the restructuring. Two recent restructurings used this approach: Greece in 2012 and Barbados in 2018, which both achieved 100 percent participation in their domestic exchanges (see Box 5).

Collective Action Mechanisms in Domestic Debt Exchanges—Greece and Barbados

Greece: In February 2012, the Greek Parliament enacted legislation1 enabling €177 billion of Greek law-governed bonds subject to the debt exchange and/or consent solicitation (the “affected domestic bonds”) to be restructured, if the consent of a requisite majority of bondholders across all bond series was obtained (i.e., via a single-limb voting mechanism).2 The requisite voting threshold required a quorum of at least 50 percent of the aggregated outstanding principal amount of all affected domestic bonds and a consent threshold of at least two-thirds of the outstanding principal amount of such bonds that are voting.3

On March 9, 2012, the Greek Finance Ministry announced that holders of approximately €152 billion in outstanding principal amount of affected domestic bonds–or approximately 85.8 percent–accepted and consented to the debt exchange and/or consent solicitation.4 As the requisite voting thresholds set forth in the Greek Bondholder Act were met, the terms of all affected domestic bonds were restructured.

Barbados: The 2018 restructuring of Barbados’ domestic debt also promoted robust participation with domestic legislation. The Debt Holder (Approval of Debt Restructuring) Act was introduced in parliament in late September and passed on October 31, 2018. Like the Greek Bondholder Act, it effectively applied aggregated collective action procedures to domestic law-governed debt. In the case of Barbados, the law made the restructuring binding where: (1) creditors holding at least 50 percent of the aggregate outstanding principal amount voted; and (2) among voting creditors, holders of 75 percent of the aggregate outstanding principal amount voted in favor.5 The Barbados mechanism differed from the Greek precedent in a few pertinent ways. First, the coverage was much broader than just bonds, covering virtually all public sector debt. Second, the collective action mechanism in Barbados did not provide identical options or menus to all creditors.6 Instead, the exchange terms were tailored by creditor type. Third, the Barbados legislation did not include a disenfranchisement of public sector entities; this was relevant given that the National Insurance Scheme held nearly half of the restructured domestic debt. Agreement of over 90 percent of domestic creditors, including all banks and insurers, was announced by the government on October 14, while the bill was under debate. The transaction was closed on November 19, and the Debt Holder Act ensured full participation.

1/ Law 4050/2012 “Rules on modification of titles, issued or guaranteed by the Greek Government with the Bondholders agreement”, otherwise commonly known as the “Greek Bondholder Act”. 2/ See also Zettelmeyer, Trebesch, Gulati, “The Greek Debt Restructuring: An Autopsy” (2013) Econ Pol 513. which referred to this as a “retrofit CAC”. See also Zandstra, “New Aggregated Collective Action Clauses and Evolution in the Restructuring of Sovereign Debt Securities” (2017) Capital Markets Law Journal, Volume 12, Issue 2. 3/ See the Press Release by the Ministry of Finance of the Hellenic Republic, dated February 24, 2012. For further detail, see Sovereign Debt Restructuring – Recent Developments and Implications for the Fund’s Legal and Policy Framework, April 2013, paragraphs 37–38. See also Zettelmeyer et al. 2013. 4/ See the Press Release by the Ministry of Finance of the Hellenic Republic, dated March 9, 2012. See also the Acts of the Ministerial Council dated March 9, 2012, which implements the Greek Bondholder Act pursuant to the achievement of the requisite voting thresholds. 5/ Barbados Parliament, Debt Holder (Approval of Debt Restructuring) Act, 2018, passed October 31, 2018, at paragraph 5. 6/ For details on the exchange terms by creditor type, see Anthony, et al., Barbados’ 2018–19 Sovereign Debt Restructuring—A Sea Change?, WP/20/34, February 2020.

Section II. Challenges to the Current Framework

25. Although the current architecture has been largely effective in resolving the relatively limited number of cases of sovereign debt problems in recent years, gaps remain that could pose challenges. With respect to bonded debt, the overarching challenge remains ensuring that creditors contribute to restructurings where needed, that creditors are bound by decisions of the majority, and that restructurings are not disrupted and delayed through actions not in the interest of creditors as a whole. As further discussed below, enhanced CACs are a significant step forward in mitigating collective action problems in international sovereign bonds but have some limitations. Regarding non-bonded debt, changes to payment terms require unanimous creditor consent, which is particularly challenging in cases where syndicated loans represent a significant portion of the country’s debt stock (as in many low-income countries) or where loans are bilateral. Restructurings are particularly difficult when countries have provided collateral to certain creditors, as secured creditors can enforce on their collateral to be compensated. The impact of these challenges on debt restructurings would depend on the composition of a country’s debt.45 Finally, the perennial issue of information asymmetry and lack of clarity on the perimeter of and treatment of claims in the debt restructuring continues to add friction to restructurings.

A. Bonded Debt

While enhanced CACs are designed to facilitate majority control of the restructuring process, they are not a panacea. Holdouts can still disrupt restructurings given the large outstanding stock of international sovereign bonds without enhanced CACs and the very limited operational experience with the clauses, as well as the possible ability of creditors to hold out.

26. While enhanced CACs have become the market standard, they have some limitations:

  • Although over 95 percent of international sovereign bonds include some form of CACs, 50 percent of the outstanding stock lacks enhanced CACs.46

  • Holdout behavior is still possible even under the single-limb voting mechanism under enhanced CACs, particularly in countries (such as frontier and low-income economies) where the total outstanding debt stock is relatively small and a holdout could assemble a blocking position at fairly low cost (see Figure 10).

  • CACs may also be absent in international bonds issued by sub-sovereigns (such as provinces) and state-owned enterprises, possibly complicating a restructuring and putting further financing pressure on the sovereign, which could prove challenging for countries with significant financial ties to large, commodity-based SOEs.

Figure 10.
Figure 10.

Low Income Country Eurobond Issuers: Minimum Bond Purchase for Holdouts 1/ 2/

(billion US$)

Citation: Policy Papers 2020, 043; 10.5089/9781513557472.007.A001

Sources: Asonuma, Niepelt and Ranciere (2018), Bloomberg, Dealogic, Eurobond Prospectus.1/ Low Income Country (LIC) Eurobond (with aggregate CACs) issuer sample is comprised of 20 countries: Armenia, Bolivia, Benin, Cameroon, Cote D’Ivoire, Ethiopia, Ghana, Honduras, Kenya, Mongolia, Mozambique, Nigeria, Pakistan, Papua New Guinea, Senegal, Sri Lanka, Tajikistan, Uzbekistan, Viet Nam, and Zambia.2/ We use average of lowest levels of bond prices in recent external private debt restructurings in 1999–2015 from Asonuma, Niepelt and Ranciere (2018) and assume that investors can purchase Eurobonds with the same level of price prior to future debt exchange.

27. There is also an ongoing debate about whether creditors with a judgment against a sovereign can be bound by CACs. In recent litigation, legal arguments have been asserted that bondholders who obtain a judgment against a defaulting sovereign prior to the operation of a CAC would no longer be affected by such clause under the doctrine of “merger” under U.S. and English law.47, 48 While this would only be a concern in protracted post-default restructurings, in such cases it would not be difficult for a creditor to show the sovereign is in breach of contract and obtain a judgment. In effect, this theory would provide that a bondholder who obtains a judgment is removed from the contractual limitations on its claim, and thus will not be subject to collective action mechanisms.49 Views among practitioners and academics diverge on this issue,50 and the application of this doctrine in the context of collective action mechanisms has not been tested in courts. If, however, such arguments were to prove viable, the contractual framework for addressing collective action problems could be seriously weakened.

28. While holdout creditors have litigated against sovereigns in a few cases, overall, creditor litigation has not been a roadblock to private-sector debt resolution. The prime example of very disruptive litigation is the litigation stemming from Argentina’s 2001 default,51 but, in almost all cases, creditors were not successful in seizing assets from the debtor or disrupting ongoing debt restructurings.52

B. Non-Bonded Debt and Other Complications

Sovereign debtors have borrowed in a variety of instruments that do not provide for majority restructuring for payment terms or provide creditors with other leverage, highlighting gaps in the current framework.

29. Debt instruments other than international bonds generally do not have majority voting provisions for modifying payment terms and continue to account for a significant share of debt in many LICs.

  • Syndicated bank loans require consent of all members of the syndicate to amend payment terms. Obtaining unanimous consent is further complicated by the fact that bank lenders may either transfer their interest in the bank loan to a secondary buyer or sell a participation in the loan to another creditor.53 As such, each creditor may have different fiduciary concerns and risk tolerance—relative to a borrower dealing with a small group of bank lenders. Syndicated loans contain amendment provisions for non-payment terms that, coupled with the legal technique of exit consents, can facilitate high participation by rendering the old debt difficult to enforce.54 However, loan terms are not standardized and vary greatly: loans will differ on the majorities needed for amendment and what terms may be amended with agreement from less than all creditors.

  • Domestic law-governed debt generally does not include majority restructuring provisions. While sovereigns can use collective action clauses in domestic law-governed bonds, as Euro-Area members have done since 2013, there are no established market practices in this area. As noted above, the IMF has focused on international sovereign bonds, since holdout creditors possess greater legal leverage.55 Nevertheless, legal scholars have raised questions about whether the “local-law advantage” can still be invoked where domestic law-governed bonds already include CACs and existing domestic law constrains the sovereign in exercising such local-law advantage.56

  • Arbitral awards against sovereigns can be significant and can be equivalent to a sizable share of GDP. These claims may be based on underlying sovereign debt claims but could also arise from other sources, such as bilateral investment treaties. Creditors holding arbitral awards may take steps to enforce their arbitral award and, under the doctrine of merger discussed above, could possibly be found not be bound by CACs even if their claim originates from a bond contract. In some instances, the original creditor may choose not to enforce the claim itself but instead to sell the claim on the secondary market, which may further complicate a restructuring. Finally, such contingent liabilities may pose some uncertainty for a sovereign’s debt sustainability analysis, particularly if such disputes are long standing.

  • Contingent liabilities such as government-guaranteed private debt may further complicate a debt restructuring since the sovereign may need to assume such liabilities and/or include such claims within the perimeter of its debt restructuring.

30. Foreign law debt contracted by sub-sovereign entities and state-owned enterprises (SOEs) poses unique challenges. Vulnerabilities in the sub-sovereign sector pose risks to the sovereign even if such debt is not explicitly guaranteed.57 Indeed, in recent sovereign debt restructurings such as Argentina (2020) and Ukraine (2015), sub-sovereign entities undertook restructuring alongside the sovereign. With respect to contractual mechanisms for restructuring, a limited survey suggests that inclusion of CACs varies among foreign law-governed sub-sovereign bonds: English law-governed bonds generally include clauses that allow a majority of holders to amend payment terms, while the practice under New York law-governed bonds is not uniform. Some, for example, include “enhanced” single-limb CACs or other forms of CACs (e.g., series-by-series), while others do not include CACs at all. Moreover, as discussed below, some of this debt is collateralized with revenue streams from natural resources.

31. While some SOEs may be subject to the restructuring provisions of national insolvency law (or special resolution rules), others may not. The application of insolvency law can ensure that creditors share the burden of the resolution of SOEs, thereby limiting the public cost of rehabilitation.58 However, SOEs in many countries may not be subject to the generally applicable corporate insolvency law or special resolution rules. Moreover, even if some SOEs are subject to an insolvency law or special insolvency rules, the sovereign may be unwilling to use the insolvency procedure in practice given the strategic importance of such SOEs in the economy.

32. Collateralized debt of both sovereigns (in particular, LICs) and sub-sovereign entities can increase the duration and cost of debt resolution.59 In an event of default, a secured creditor may enforce the collateral and receive proceeds from it, diminishing incentives to participate in a restructuring. Even creditors who hold collateral that may not easily be enforced (as explained above) are likely to retain more leverage than unsecured creditors, and burden-sharing issues may draw out debt restructuring negotiations, which in turn may delay fresh external financing. Delays in completing restructurings weaken investment and growth and hence capacity to repay, which benefits neither the borrower nor creditors as a whole. As discussed above, providing collateral that is an essential asset of the debtor (such as shares in a strategic company) or that relates to essential payment streams (such as oil revenues) can give the creditors holding such collateral significant bargaining power, further complicating debt restructurings.

33. Negative pledge clauses (NPCs)—which disincentivize certain collateralization—vary in coverage and have not always been vigorously enforced. NPCs typically protect the financial interest of an existing creditor by limiting (or sometimes precluding) the ability of a borrower to grant security over specified assets or revenue streams to secure a prospective lender, thereby ensuring that its claims are not subordinated to other debt. NPCs (or permitted lien clauses), which are an undertaking by the borrower in favor of the lender, are fairly standard in commercial law contracts and also typically appear in bonds as well as the loan documentation of MDBs. They normally provide for some carve-outs or exceptions to triggering the clause, and these exceptions are extensively negotiated within the private sector.60 The drafting of NPCs may vary depending on the lender/borrower, including in terms of covered transactions, assets, borrowing entities, and remedies and are not always vigorously enforced.61 In addition, the failure of borrowers to transparently disclose and report their secured borrowing activities, coupled with weak governance or low capacity in public debt management or in a borrower’s ability to ensure that it meets its obligations to its various lenders, can create information asymmetries among creditors.

C. Information Asymmetries

Information asymmetry between debtors and creditors adds friction to debt restructurings by preventing common understandings of the perimeter of the operation and how each claim will be treated, as well as in cases where economic uncertainty is high. Enhanced debt transparency including on treatment of claims in the restructuring is needed.

34. Debt restructurings can be further complicated when there is a lack of clarity as to how particular claims will be classified, heightening inter-creditor equity concerns. Ambiguity (from either the borrower or the creditor) regarding the classification of a claim as official or private can create uncertainty over the perimeter of debt to be covered in a restructuring and the terms of treatment (which partly depends on the perimeter), affecting the prospects for reaching agreement. Claims on SOEs, state-owned banks, or debt from development banks (such as national development banks or export-import banks) may also be classified as sovereign or private, depending on the extent of linkages between the entity and the sovereign or whether the entity lends on the direction or on behalf of the sovereign. Clear upfront principles for the treatment of different claims, as well as clear communication by the debtor and creditors with respect to claims in a particular restructuring, can help mitigate inter-creditor equity concerns.

35. Debt restructuring in the context of macro-uncertainty (such as volatile commodity prices) may make outcomes more uncertain. There are clearly challenges when there is major macro-uncertainty on commodity prices or the ability of the sovereign to design a predictable and credible adjustment path. The scope for differences of view between debtors and creditors on repayment capacity may widen, making it more difficult to reach agreement. These issues are addressed in a forthcoming paper on the role of state-contingent debt instruments and value-recovery instruments in sovereign debt restructurings building on previous IMF work.

36. While much has been achieved, significant improvement is also needed in the area of debt transparency. The lack of debt transparency hampers rapid resolution of debt distress and undermines risk assessments by creditors and other stakeholders. Data gaps differ between bonds and loans regarding information available on debt amount, terms of financial contract, and holder of debt.62 On bonded debt, information on debt amounts and contract terms is generally more transparent, as many prospectuses are issued publicly (although many bond indentures are not publicly available), unless bonds are issued as a private placement. Information on bondholders are available partially through commercial platforms such as Bloomberg, but sovereigns do not have access to central securities depositories (CSD) data on bondholder identities, and given secondary market activity, bondholder identities may change frequently. As a result, at the time of a restructuring, sovereigns often have to hire a firm to collect information on bondholders; information on holders of domestic debt may be available from the CSD often run by the central bank or other government-owned entity. The limited information on bondholders, together with the diverse and atomized creditor base can delay and complicate restructuring negotiations. In addition, for bank loans and other commercial loans (e.g., resource-backed loans), as well as certain official-sector loans, information on the debt amount and terms of contract including collateralization features, remains opaque (e.g., undisclosed debt in Mozambique).63 In some instances, side letters defining additional requirements outside the contracts exist that exacerbate the lack of transparency. Finally, the imposition by lenders of confidentiality clauses also creates information asymmetries and a lack of transparency.

Section III. Reform Options

With a view to facilitating high participation in debt restructurings involving private-sector debt, there are a number of reform options that could address some of the challenges identified. The following section presents tools that can be considered to improve the current framework in the areas of contractual, legislative, IFI policies, debt transparency, and capacity development.

A. Enhanced Contractual Approach

The enhanced contractual approach generally remains appropriate, but it can be strengthened to limit disruptive holdout behavior. IMF staff has looked into the further use of trust structures, state-contingent features, standardized majority restructuring provisions for payment terms in loan agreements, and strengthened NPCs as potential avenues to make the framework more robust.

37. Enhanced CACs have contributed to achieving high participation in recent restructurings. The first uses of the enhanced CACs in the Ecuador and Argentina restructuring demonstrated the parties’ ability to use those clauses to restructure without major complications, while modifying specific points to suit their needs (see Box 3, above). ICMA is reviewing the continued appropriateness of the design of enhanced CACs in light of these recent restructurings to determine whether further refinement of the standard language is required or if the market should be left to incorporate case-specific modifications, and IMF staff will continue to work closely with issuers and market participants on this review. Additional options to further strengthen the contractual framework for sovereign debt restructurings are discussed below.64

Trust Structures

38. The issuance of bonds under trust structures provides additional protections against disruptive holdout enforcement actions.65 International sovereign bonds are typically issued under either fiscal agency agreements (FAAs) or trust structures. Under an FAA, the fiscal agent serves as an agent of the issuer, and its main responsibility is making principal and interest payments to the bondholders. Under trust structures, however, a bond trustee acts on behalf of, and has a number of responsibilities to, bondholders as a group. Trust structures provide additional protections against holdout creditors because they put limitations on individual creditor enforcement actions and require the pro rata distribution of the proceeds of litigation among all bondholders. In particular, trust structures differ from FAAs in that (i) only the trustee can commence litigation to collect accelerated amounts and (ii) proceeds of any such amounts collected are shared amongst all bondholders. These features prevent (and disincentivize) minority bondholders from litigating to disrupt an orderly restructuring. Short of switching to trust structures, bonds issued under FAAs could be amended to include sharing clauses, which achieve sharing of proceeds among creditors on a ratable basis.66

39. While trust structures continue to be prominent in international sovereign bonds issued under New York law, issuances under English law almost exclusively use FAAs. A number of large emerging market issuers under New York law, such as Mexico and Chile, have switched to trust structures from FAAs in recent years. The share of new international sovereign bond issuances since October 1, 2014 using trust structures is approximately 33 percent (in nominal principal terms), of which approximately 92 percent are issued under New York law and approximately 8 percent under English law. As legal practitioners have noted, the continued preference for FAAs, particularly among lower-income countries that issue under English law, may reflect the (slightly) higher costs associated with trust structures.

Sub-Sovereign Debt

40. Further work is required to deepen understanding of the challenges posed by sub-sovereign debt. IMF staff has been monitoring the inclusion of CACs in international law bonds issued by sub-sovereign governments and large SOEs, as well as their collateralization, to better understand trends and practices in this area. It will continue to encourage countries to subject SOEs to a robust general insolvency regime, in line with international best practices, which can effectively ensure that creditors contribute to the resolution of the financial problems of SOEs, thereby limiting the public cost of rehabilitation. Sub-sovereign entities should also be encouraged to include enhanced CACs in their international bonds.

Model Majority Restructuring Clauses for Payment Terms in Syndicated Loans

41. In order to address the lack of majority restructuring provisions for payment terms in syndicated loans, consideration could be given to developing model clauses focused on specific areas of reform—similar to the enhanced CACs.67 In particular, including in syndicated sovereign loan agreements standardized provisions that allow a qualified majority of lenders to agree to amend payment terms (that currently require unanimity) could facilitate restructuring of loans issued under such agreements and provide more predictability than the current approach of relying on the use of exit consents. Preliminary feedback from market participants was mixed; while some indicated a model clause might be a helpful reform, others highlighted that the impact may be limited, given that a significant portion of sovereign loans are not syndicated but bilateral, unlike international bonds which are held by many investors. Moreover, there may be substantial creditor pushback against the introduction of such clauses due to the limitation of voting power and possibly for regulatory reasons (e.g., where holding a minority stake in loans might affect their capital requirements). In any event, and as was the case with the development of enhanced CACs, further consultation and cooperation between the official and private sectors would be essential to develop such model loan provisions and ensure their widespread adoption. As such a reform would take considerable time, its effectiveness in facilitating sovereign debt restructurings in the near term would be limited.

Negative Pledge Clauses

42. The excessive use of collateralization that may pose challenges to a sovereign debt restructuring could be addressed by improving transparency and strengthening NPCs. These measures could include: (i) enhanced authorization processes and accountability in borrower governments for the creation of collateralized borrowings; (ii) greater transparency and reporting in connection with collateralized borrowings; and (iii) a concerted effort, particularly among official bilateral lenders, their state-owned banks and development agencies, to only sparingly and very selectively structure collateralized loans. In addition, lenders should not seek security as a stopgap for inadequate financial and commercial due diligence on their part, or as a stopgap that enables investment in a project or initiative that is not otherwise financially sound. Encouraging creditors to uniformly respect other lenders’ NPCs (and related provisions) may be an additional step towards minimizing the excessive creation of secured debt in the first place. Further, while lenders and borrowers negotiate NPCs dependent on each transaction, there is a perception among market participants that there are loopholes in NPCs that could be tightened. In some agreements, NPCs do not capture certain types of transactions (domestic currency transactions) or certain entities (central government versus public sector) or all borrowing (bonds versus loans). Strengthening such clauses to capture more transactions, so as to encourage a more sparing use of collateralization, could be considered.68

State-Contingent Features

43. State-contingent features in sovereign debt instruments can help protect the sovereign from downside risk, especially to cover situations involving natural disasters. Many countries may face a higher frequency and greater intensity of natural disasters than in the past, disproportionally affecting those which are more vulnerable, including small island economies. These events have significant economic and humanitarian costs. Bond instruments with state-contingent features—extending debt service obligations when natural disasters hit—would provide the authorities breathing space and allow them to focus on the humanitarian needs and the recovery efforts.69 In the event a full and deep restructuring is needed, the extra time will be extremely helpful from a capacity perspective. More widespread issuance of bonded debt with these features would allow investors to diversify risks across more countries. Higher demand and a deeper market would in turn reduce the new issuance premium, the liquidity premium, and in general help with pricing. This might require a coordinated issuance effort, and a hard look at the regulatory treatment of these instruments (i.e., to ensure that lenders do not face disincentives to their use). Still, the greatest benefits would only be realized if all creditors, including bilateral official lenders, were to accept such clauses. Indeed, such lenders could lead market development by adopting such clauses (perhaps subject to comparability of treatment requirements, to preserve their seniority).

44. There are challenges to a similar application of state-contingent features to cover commodity price shocks. The concerns are similar to those regarding natural disasters—there are many small countries which rely heavily on one commodity, and many of these also suffer from weak capacity and would benefit from breathing space in the wake of a shock. In addition, commodity prices are beyond the control of most countries, easing any fears of index manipulation that could be present in, for example, GDP-linked bonds. However, unlike idiosyncratic natural disasters, lenders may also be exposed to a downturn in commodity prices, potentially limiting their desire to allow debt relief to borrowers in a shock scenario. Still, to encourage the use of such clauses, an approach similar to that taken with natural disasters could be applied to a narrow set of commodities and small countries.

45. Despite some success, challenges remain in the use of value recovery instruments in restructuring cases. VRIs may not always help achieve high creditor participation because, as noted above, differences in perceptions between the creditors and the debtor seem to make these instruments expensive compared to fixed-value instruments. VRIs may be more useful in the restructuring of SOEs and PPPs, which are more similar to corporate bankruptcies (where the use of such instruments is common). In designing new VRIs it will be important to minimize measurement issues, avoid lagging indicators, and structure payouts properly (including through floors and caps).

B. Targeted Legislative Options

There is a question about the desirability of broader use of targeted legislative measures to complement the contractual approach by limiting holdout creditor recovery as proposed in some quarters, but closer analysis would be needed of the desirable parameters of such measures so as to limit the impact on creditors’ rights and to avoid undermining the secondary market.

46. Taking into account the recent adoption by some advanced economies of targeted legislation to complement the contractual approach by limiting holdout creditor recovery, there is a question of whether the wider use of these tools is desirable. This could be most effective in the key jurisdictions under whose law most bonds are issued—New York and England—but need not be limited to them (and indeed, as noted above, the European Parliament has called for adoption by EU members of a regulation based on the Belgian variant of these laws). New York has no such laws at present, and while the UK already has such legislation, its scope is limited to the pre-2004 debt of HIPC-eligible debtor countries as discussed above.

47. Importantly, any such legislation would need to be carefully tailored so as to limit the impact on creditors’ rights and avoid undermining the secondary market in sovereign debt.70 Appropriate parameters would thus need to be considered with respect to the coverage of such legislation, including the types of debtors, types of claims, guidance on “acceptable” level of creditor profit, and identification of problematic creditor behavior. Consideration could also be given to tying these limitations to restructuring cases where the official sector has provided contributions, either with new financing or debt relief. Each of these parameters has been weighed differently in the three laws currently on the books as discussed above. Further analytical work and market consultations would be required to determine the appropriate balance.

C. Policies of International Financial Institutions

International financial institutions can play a role through their lending policies in supporting orderly and speedy sovereign debt restructuring.

48. IFIs and regional institutions can contribute to the orderly and speedy sovereign debt restructuring of private-sector claims through several channels. These include (1) the conditions under which IFIs (and particularly the IMF) will lend to countries with unsustainable debts and (2) the support of debt restructurings that require upfront financing of credit enhancements, buybacks, or cash “sweeteners”. As noted below, the IMF has an extensive work program on sovereign debt which includes a review of certain policies applicable to both channels. Consideration could also be given to a review of relevant policies of other IFIs.

49. The conditions under which the IMF may lend to countries whose debt is deemed unsustainable on a forward-looking basis can create incentives for sovereign debtors and their creditors to engage in orderly and speedy debt restructurings. If a sovereign’s debt is deemed unsustainable, the IMF is precluded from lending unless the member is taking steps to restore debt sustainability.71 Post-default, the IMF’s Lending into Arrears policy (LIA) ensures that the IMF lends into arrears only if (i) prompt IMF support is considered essential for the implementation of the member’s adjustment program and (ii) the member is pursuing appropriate policies and is making a good faith effort to reach a collaborative agreement with its creditors. Pre-default, the IMF requires assurances that a credible process is in train that the debt restructuring will be successful in restoring debt sustainability consistent with the IMF-supported program. IMF practices and policies in both areas will be reviewed in 2021 to ensure their effectiveness. The review will examine the IMF’s role in pre-default debt restructurings, where there is presently a well-established practice but not a formal policy. It will also examine recent experiences with the LIA policy, including a more fragmented creditor base, which raises questions on how the expectation of a representative creditor engagement through creditor committees in complex cases should be applied.

50. In some circumstances, it may make sense for IFIs to support debt restructurings, within their mandates, by providing financing that allows members to offer financial incentives to their creditors. Depending on the IFIs’ particular mandate, such support could take the form of financing to sovereigns for debt buy-backs, purchases of high-quality collateral that “enhance” a debt exchange offer or cash sweeteners. Some IFIs (but not the IMF) may also be able to offer partial guarantees on newly issued debt. Support of this type was offered by the IMF and World Bank in the context of the “Brady deal” restructurings of the 1990s (but discontinued thereafter, e.g., the IMF policy for Debt or Debt Service Operations) and by the European Financial Stability Facility (EFSF) for the 2012 Greek debt restructuring. The economic rationale for this form of support is that it can raise the value of the new instruments offered to private creditors, and hence the prospects for high participation, for a given level of debt relief from the perspective of the debtor.72 At the same time, this type of financial support can create significant risks for IFIs, and would reduce IFI resources available for important purposes such as social support or boosting reserves. Whether and in what circumstances IFI financing should be used to offer such support hence requires a careful review.

D. Debt Transparency

The international community should further support and pursue the initiatives that enhance debt transparency.

51. The international community should further support and pursue the initiatives that enhance debt transparency; the IMF and the World Bank will jointly work on this as part of the multi-pronged agenda on debt.73 The main responsibility for transparency lies with the sovereign and its creditors. The IMF supports debt transparency through IMF policy (e.g., the LIC-DSF operationalized in July 2018 and the ongoing reviews of the market access countries DSA and the debt limits policy emphasizing full debt disclosure), data provision and surveillance (Article VIII, Section 5), capacity development (strengthening member’s debt data recording, monitoring, and reporting) and supporting other stakeholders’ transparency initiatives (such as those of the IIF).

E. Capacity Development

Continued capacity development to strengthen ex ante debt management capacity remains critical.

52. The provision of capacity development assistance by the international community to debtors will continue to be critical. On the front end, increasing capacity in debt management expertise and disseminating best practices will help sovereigns better manage their debt risks and reduce the need for debt restructurings. Timely and comprehensive debt recording and strong governance safeguards for the debt authorities are critical for this. The IMF and the World Bank have delivered extensive technical assistance to countries where debt management has been identified as a macro-critical risk, and several ongoing initiatives supported by the G20 are in place to help continue this work across LICs and EMs.74

Section IV. Possible Responses to a COVID-Related Systemic Crisis

Should a COVID-related systemic sovereign debt crisis materialize that cannot be effectively addressed by the existing resolution toolkit, additional measures would be needed to bolster the current framework as a last resort on a time-bound basis.

53. The recent COVID-19 pandemic has highlighted that, in the case of a widespread systemic crisis, the current framework may be pushed to its limits. As highlighted by the June 2020 update to the World Economic Outlook, global public debt is projected to surge by 19 percentage points in 2020 to reach an all-time high (see Figures 11 and 12).75 More prolonged macroeconomic and financial impacts would raise the potential for the pandemic to be followed by a wave of sovereign debt problems in EMDEs. The potentially large impact of the pandemic on the solvency of these borrowers may require deep restructurings, implying large losses for creditors, and potentially involving protracted and difficult negotiations. In extreme scenarios, these impacts could also have financial stability implications.76

Figure 11.
Figure 11.

Debt-to-GDP Ratio

(Average, % of GDP)

Citation: Policy Papers 2020, 043; 10.5089/9781513557472.007.A001

Figure 12.
Figure 12.

Debt-to-GDP Ratio – DSSI-Eligible Countries

(Average, % of GDP)

Citation: Policy Papers 2020, 043; 10.5089/9781513557472.007.A001

54. As a result, a deep or systemic sovereign debt crisis may require additional instruments both to create incentives for creditors to agree to the needed deep restructurings and to soften the impact on them. Since contractual reforms would require time before they can affect the debt stock, such additional instruments could only be either of a statutory or financial nature. Statutory instruments could include use of the targeted “anti-vulture fund” legislation discussed in Section II as well as other domestic and international law tools, on a time-bound basis, as discussed below. Furthermore, within their mandates, the IMF and other IFIs could provide targeted financing for financial incentives that could raise participation in debt exchanges, as they did in the restructurings that resolved the debt crises of the 1980.

55. IFI-financed cash or credit enhancements could narrow the gap between the debt relief required for debtors to regain solvency and the losses that creditors are willing or able to accept. As a result, when the required debt relief is very deep, such financing could make debt deals more feasible. At the same time, this approach is likely to be subject to significant limitations. First, it is constrained by the financial resources that IFIs could make available, and the fact that competing demands on these resources will be particularly intense during a systemic crisis. Second, IFI financing of financial incentives provided by debtors implies that claims on these debtors will migrate from the balance sheets of private investors to those of IFIs. The more extensive this migration, the smaller the cushion of private claims and the more difficult it will be to protect the IFIs’ de facto preferred creditors status. Hence, while IFI financing of financial incentives provided by debtors could contribute to resolving a widespread systemic debt crisis, it is not an instrument that can be used on a very large scale.

56. A further option could be time-bound legislative or executive actions that could be used as a last resort. In addition to targeted legislation of the kind discussed above that limits recovery or seizure of sovereign assets under specified circumstances, legislation that focuses on the timing of lawsuits (rather than the amount of recovery or ability to enforce a judgment) could in principle also be considered for this purpose.77 A further option might be actions at the national or international level that immunize specified sovereign assets from judicial actions such as attachment.78 At the international level, the key precedent in this area is the U.N. Security Council Resolution passed in 2003 to create immunity for Iraq’s oil assets against attachment and other forms of legal action (see Box 6).79 While CACs and other contractual terms establish the rights of debtors and creditors, these options would limit or delay in a time-bound manner creditors’ enforcement of those contractual rights.80 This both reduces incentives to hold out and limits the ability of holdouts to derail restructuring during a systemic crisis through disruptive enforcement actions at the expense of other creditors who may be inclined to accept a restructuring deal.

U.N. Security Council Resolution in Iraq’s Restructuring

In 2003, Iraq’s outstanding sovereign debt stock exceeded US$140 billion, which was owed to a very diverse creditor base consisting of sovereign bilaterals, commercial banks, companies and trade creditors. Iraq derived virtually all of its foreign currency earnings from the sale of oil. Creditors could therefore have impeded Iraq’s economic recovery by attaching Iraqi oil shipments in the international markets and seizing the cash proceeds from the sale of that oil.

The U.N. Security Council Resolution No. 1483 of May 22, 2003 encouraged a prompt restructuring of Iraq’s debt and immunized all Iraqi oil sales, as well as the cash proceeds from the sale of that oil, from “any form of attachment, garnishment, or execution.” Since Resolution 1483 was adopted under Chapter VII of the U.N. Charter, it was binding on all member states and all UN members were required to incorporate these immunities for Iraqi assets into their domestic law (for instance, an Executive Order in the US). The legal immunities for Iraqi assets provided by Resolution 1483 lasted until 2011, and those under U.S. Executive Order until 2014. Under the protection of these legal immunities, Iraq successfully restructured most of debt stock on terms that gave Iraq debt relief of at least 80 percent.

57. Despite the potential necessity of their use in response to a systemic crisis, measures of the kind discussed above raise important legal and policy issues and would require careful tailoring. If overused, such an approach could undermine basic principles inherent in a commercial law system regarding the enforceability of contractual rights and increase the ex ante costs of debt issuances. Thus, such measures would be expected to be used as a last resort and on a time bound basis to address the unique challenges posed by the crisis.

Section V. Conclusions and Next Steps

While the framework generally remains appropriate, it can be further strengthened. The IMF’s work program on sovereign debt will support this.

58. On balance, the experience to date, while not extensive, suggests that the contractual framework for sovereign debt resolution with private creditors remains generally appropriate. Compared with previous periods, recent sovereign debt restructurings have generally proceeded smoothly, with shorter average duration and higher average creditor participation, and were largely preemptive, mainly due to the use of CACs. No ex-post litigation with private creditors has arisen when CACs were used. While creditor coordination has raised challenges, experience so far suggests these challenges can be managed, as reflected in the recent Argentina and Ecuador restructurings.

59. Notwithstanding the above, the framework can be further strengthened along a few key lines, with a view to closing key gaps in the system that could pose challenges:

  • CACs. The IMF will continue to promote the inclusion of enhanced CACs in international sovereign bonds and periodically update the international community on the inclusion of enhanced CACs in international sovereign bonds. Sub-sovereign entities should also be encouraged to include enhanced CACs in their foreign law-governed bonds.

  • Trust Structures. Sovereigns should be encouraged to issues bonds under trust structures.

  • Model Majority Restructuring Loan Clauses for Payment Terms. Official and private sectors should cooperate to develop model majority restructuring clauses for payment terms in loans and encourage their widespread adoption.

  • Negative Pledge Clauses. Official and private sectors should join forces in promoting greater disclosure about the use of collateral, enhanced authorization process for borrowers, increased awareness among borrowers and lenders alike to respect NPCs (and permitted lien obligations) of other lenders, and more rigorous enforcement of NPCs so as to disincentivize the excessive proliferation of new collateralized debt.

  • Transparency. In line with ongoing international initiatives and reviews of IMF policies, debt transparency should be further enhanced and both creditors and the sovereign should be encouraged to clarify the perimeter of claims upfront.

  • State-Contingent Features. Increased use of state-contingent features, particularly to protect debtors against downside risks, such as natural catastrophes, could be considered.

  • Insolvency Regime. SOEs should be subject to a robust general insolvency regime in line with international best practices.

  • Targeted Statutory Tools in Limited Circumstances. The desirability of wider application of targeted statutory tools of the kind already in place in a few countries to complement the contractual approach (i.e., “anti-vulture fund” legislation) could be further explored to limit holdout creditor recovery in specified circumstances, though they should be carefully designed to limit the impact on creditors’ rights and avoid undermining the secondary market.

  • Capacity development. IFIs and relevant public and private sector entities should continue to provide technical assistance and training in debt management and debt data reporting to enhance members’ capacity.

60. Should a COVID-related systemic sovereign debt crisis requiring multiple deep restructurings materialize, the current resolution toolkit may not be adequate in addressing the crisis effectively and additional instruments may need to be activated at short notice. Since contractual reforms would require time to become effective, such instruments could only be either of a financial or statutory nature. The former could include IFI financing of cash or credit enhancements that lowers the risk, and hence increases the value, of the assets offered to creditors without reducing debt relief from the perspective of the debtor. However, to avoid undermining the de facto preferred creditor status of IFIs, the scale of such financing must necessarily remain limited. The latter could in principle include both targeted domestic law tools and international law options (such as a U.N. Security Council resolution), which could be used to limit creditor recovery or the timing of suits or to immunize specified assets from attachment by creditors. These instruments raise significant legal and policy issues, however, and would be expected to be used only as a last resort and on a time-bound basis to address the unique challenges posed by the crisis.

61. Going forward, the IMF has a rich work program on sovereign debt in which the IMF will review its relevant policies and collaborate closely with the World Bank, when necessary.

  • Explore ways to enhance the market-based approach and the sovereign debt resolution architecture, including through the greater use of state-contingent debt instruments.

  • Strengthen ex ante debt management through continued IMF and World Bank technical assistance.

  • Review of the debt limits policy.

  • Review of debt sustainability analysis for market access countries (MAC DSA).

  • Continue with the multi-pronged approach to addressing debt vulnerabilities, jointly with the World Bank.

  • Review of the arrears policies.

References

  • Anthony, Myrvin, Impavido, Gregorio and van Selm, Bert, 2020, “Barbados’ 2018–2019 Sovereign Debt Restructuring – A Sea Change?”, WP/20/34.

    • Search Google Scholar
    • Export Citation

  • Arslanalp, Serkan, Bergthaler, Wolfgang, Stokoe, Philip and Tieman, Alexander, 2018, “The current landscape”, in: Sovereign Debt: A Guide for Economists and Practitioners, (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Asonuma, Tamon and Trebesch, Christoph, 2016, “Sovereign Debt Restructurings: Preemptive or Post-Default”, Journal of the European Economic Association, Volume 14, Issue 1, pp. 175–214.

    • Search Google Scholar
    • Export Citation

  • Asonuma, Tamon, Chamon, Marcos, Erce, Aitor, and Sasahara, Akira, 2019, “Costs of Sovereign Defaults: Restructuring Strategies, Bank Distress and the Capital Inflow-Credit Channel”, IMF Working Paper No. 19/69.

    • Search Google Scholar
    • Export Citation

  • Asonuma, Tamon, Niepelt, Dirk and Ranciere, Romain G., 2018, “Sovereign Bond Prices, Haircuts and Maturity”, NBER Working Paper 23864.

    • Search Google Scholar
    • Export Citation

  • Bardozzetti, Alfredo and Dottori, Davide, 2013, “Collective action clauses: how do they weigh on sovereigns?”, Bank of Italy, Economic Working Papers No. 897.

    • Search Google Scholar
    • Export Citation

  • Borensztein, Eduardo, and Panizza, Ugo, 2009, “The Costs of Sovereign Default”, IMF Staff Papers, Volume 56, Issue 4, pp. 683–741.

    • Search Google Scholar
    • Export Citation

  • Bradley, Michael and Gulati, Mitu, 2013, “Collective Action Clauses for the Eurozone”, Review of Finance, pp. 1–58.

  • Buchheit, Lee and Gulati, Mitu, 2012, “Restructuring a Sovereign Debtor’s Contingent Liabilities”, SSRN.

  • Buchheit, Lee and Gulati, Mitu, 2019, “Sovereign Debt Restructuring and U.S. Executive Power”, Capital Markets Law Journal, Volume 14, Issue 1, pp. 114–130.

    • Search Google Scholar
    • Export Citation

  • Buchheit, Lee and Gulati, Mitu, 2020, “The Argentine Collective Action Clause Controversy”, Capital Markets Law Journal (forthcoming 2020).

    • Search Google Scholar
    • Export Citation

  • Buchheit, Lee and Hagan, Sean, 2020, “From Coronavirus Crisis to Sovereign Debt Crisis”, Financial Times.

  • Chung, Kay and Papaioannou, Michael, “Do Enhanced Collective Action Clauses Affect Sovereign Borrowing Costs?”, IMF, WP/20/162.

  • Cruces, Juan J. and Trebesch, Christoph, 2013, “Sovereign Defaults: The Price of Haircuts”, American Economic Journal: Macroeconomics, 5 (3): 85–117.

    • Search Google Scholar
    • Export Citation

  • Eichengreen, Barry, Kletzer, Kennet and Mody, Ashoka, 2003, “Crisis Resolution: Next Steps”, National Bureau of Economic Research, Working Paper No. 10095.

    • Search Google Scholar
    • Export Citation

  • Hagan, Sean, 2020, “Sovereign debt restructuring: The centrality of the IMF’s role”, Peterson Institute for International Economics Working Paper 20–13.

    • Search Google Scholar
    • Export Citation

  • International Monetary Fund and World Bank, 2018, G20 Notes on Strengthening Public Debt Transparency (Washington).

  • International Monetary Fund and World Bank, 2020a, Collateralized Transactions: Key Considerations for Public Lenders and Borrowers (Washington).

    • Search Google Scholar
    • Export Citation

  • International Monetary Fund and World Bank, 2020b, Public Sector Debt Definitions and Reporting in Low-Income Developing Countries (Washington).

    • Search Google Scholar
    • Export Citation

  • International Monetary Fund and World Bank, 2020c, Staff Note for the G20 International Financial Architecture Working group (IFAWG)—Recent Developments on Local Currency Bond Markets in Emerging Economies (Washington).

    • Search Google Scholar
    • Export Citation

  • International Monetary Fund and World Bank, Implementation Update on the Joint IMF-WB Multi-Prong Approach for Addressing Debt Vulnerabilities (forthcoming).

    • Search Google Scholar
    • Export Citation

  • International Monetary Fund, 1999, Orderly & Effective Insolvency Procedures, Legal Department (Washington).

  • International Monetary Fund and World Bank, 2003, Report of the Managing Director to the International Monetary and Financial Committee on a Statutory Sovereign Debt Restructuring Mechanism (Washington).

    • Search Google Scholar
    • Export Citation

  • International Monetary Fund and World Bank, 2013, Sovereign Debt Restructuring – Recent Developments and Implications for the Fund’s Legal and Policy Framework (Washington).

    • Search Google Scholar
    • Export Citation

  • International Monetary Fund and World Bank, 2014, Strengthening the Contractual Framework to Address Collective Action Problems in Sovereign Debt Restructuring (Washington).

    • Search Google Scholar
    • Export Citation

  • International Monetary Fund and World Bank, 2015a, Progress Report on Inclusion of Enhanced Contractual Provisions in International Sovereign Bond Contracts (Washington).

    • Search Google Scholar
    • Export Citation

  • International Monetary Fund and World Bank, 2015b, The Fund’s Lending Framework and Sovereign Debt—Further Considerations (Washington).

    • Search Google Scholar
    • Export Citation

  • International Monetary Fund and World Bank, 2016, Second Progress Report on Inclusion of Enhanced Contractual Provisions in International Sovereign Bond Contracts (Washington).

    • Search Google Scholar
    • Export Citation

  • International Monetary Fund and World Bank, 2017a, IMF Policy Paper, State-Contingent Debt Instruments for Sovereigns (Washington).

  • International Monetary Fund and World Bank, 2017b, Third Progress Report on Inclusion of Enhanced Contractual Provisions in International Sovereign Bond Contracts (Washington).

    • Search Google Scholar
    • Export Citation

  • International Monetary Fund and World Bank, 2019, Fourth Progress Report on Inclusion of Enhanced Contractual Provisions in International Sovereign Bond Contracts (Washington).

    • Search Google Scholar
    • Export Citation

  • International Monetary Fund and World Bank, 2020a, The Evolution of Public Debt Vulnerabilities in Lower Income Economies (Washington).

  • International Monetary Fund and World Bank, 2020b, World Economic Outlook Update (Washington).

  • International Monetary Fund and World Bank, IMF Staff Discussion Note, Role of State-Contingent Debt Instruments and Value Recovery Instruments in Sovereign Debt Restructurings (forthcoming).

    • Search Google Scholar
    • Export Citation

  • International Monetary Fund and World Bank, IMF Board Paper, Issues in Restructuring of Sovereign Domestic Debt (forthcoming).

  • Krueger, Anne, 2002, “A New Approach to Sovereign Debt Restructuring” (Washington: International Monetary Fund).

  • Loan Market Association, 2019, Guide to Syndicated Loans & Leveraged Finance Transactions (London).

  • Mihalyi, David, Adam, Aisha and Jyhjong, Hwang, 2020, “Resource-Backed Loans: Pitfalls and Potential”, Report 27, Natural Resource Governance Institute, February 2020.

    • Search Google Scholar
    • Export Citation

  • Sobel, Mark, 2020, “Argentina and Creditors Enter New Round”, OMFIF.

  • Sturzenegger, Federico, 2004, “Tools for the Analysis of Debt Problems”, Journal of Restructuring Finance, Volume 1, Issue 1, pp. 201–223.

    • Search Google Scholar
    • Export Citation

  • Walker, Mark and Cooper, Richard, 2017, “Venezuela’s Restructuring: A Realistic Framework”, SSRN.

  • Walker, Mark and Chong, Alice, 2020, “Collective Action Clauses Reexamined: Thank you Argentina”, SSRN.

  • Weidemaier, Mark, 2019, “Restructuring Italian (Or Other Euro Area) Debt: Do Euro CACs Constrain or Expand the Options?”, UNC Legal Studies Research Paper.

    • Search Google Scholar
    • Export Citation

  • Weidemaier, Mark, 2020a, “Judgments > CACs”, Credit Slips.

  • Weidemaier, Mark, 2020b, “Venezuela, Lebanon, and Tools to De-Fang 𠆘Rush-In’ Creditors”, Credit Slips.

  • Zandstra, Deborah, 2017, “New Aggregated Collective Action Clauses and Evolution in the Restructuring of Sovereign Debt Securities”, Capital Markets Law Journal, Volume 12, No. 2, pp. 180–203.

    • Search Google Scholar
    • Export Citation

  • Zettelmeyer, Jeromin, Trebesch, Christoph, and Gulati, Mitu, 2013, “The Greek Debt Restructuring: An Autopsy”, Economic Policy, Volume 28, pp. 513–563.

    • Search Google Scholar
    • Export Citation
1

In 2002, the IMF proposed a sovereign debt restructuring mechanism (“SDRM”), which would have been a statutory sovereign bankruptcy process but did not garner the required membership support. Following this, the IMF endorsed the contractual approach. See: Krueger, Anne 2002, “A New Approach to Sovereign Debt Restructuring” (Washington: International Monetary Fund); IMF 2003, Report of the Managing Director to the International Monetary and Financial Committee on a Statutory Sovereign Debt Restructuring Mechanism (Washington).

3

Strengthening the Contractual Framework to Address Collective Action Problems in Sovereign Debt Restructuring, October 2014 (the “2014 Paper”). The four progress reports on inclusion of enhanced CACs were published in September 2015, December 2016, December 2017, and March 2019, respectively.

5

Debt restructurings are typically associated with significant output costs but these costs are lower in cases where restructuring is preemptive or where an agreement is reached quickly after a default. Several channels can lead to lower GDP growth after a debt restructuring—which average about 2 percentage points each year on average-including reduced sovereign access to market financing, spillovers on trade, investment, and productivity, and reduced corporate access to foreign credit (Sturzenegger 2004, Borensztein and Panizza 2009). However, output costs are smaller after a preemptive restructuring in which an agreement is reached in a pre-default context. Preemptive restructurings are quicker (only 1 year compared to 5 years in post-default restructurings), have a smaller NPV haircut, and see a quicker re-access to international capital markets (Asonuma and Trebesch, 2016). Even post default, output costs can be reduced when debtors and creditors are able to reach a quick agreement (Asonuma et al. 2019). The quicker agreement seems to be even more important for attenuating output costs after a default than a smaller NPV haircut.

6

Preemptive and post-default cases accounted for 38 and 62 percent of privately-held external debt restructurings, respectively, in 1978–2010, and the average duration of preemptive and post-default cases was 1.0 and 5.0 years, respectively (Asonuma and Trebesch 2016). These numbers are based on a definition of restructuring events that treats overlapping restructurings involving the same debtor but different debt instruments as separate restructurings. Treating such overlapping restructurings as one event would lead to longer average durations. The fact that the average duration of restructurings since 2014 have been relatively short is robust to these differences in definitions.

7

See IMF Staff Discussion Note, Role of State-Contingent Debt Instruments and Value Recovery Instruments in Sovereign Debt Restructurings (forthcoming).

8

Some GDP-linked bonds have suffered from measurement issues that have led to unexpectedly low payouts.

10

As of August 11, 2020, 43 of the 73 eligible countries (or 59 percent) have made formal requests for the DSSI (based on creditor and debtor data) and requests continue to come in. The G20 estimates debt service deferral under the initiative likely to be on the order of US$ 5.3 billion. Of the remaining DSSI eligible countries, about a quarter have informally indicated interest to participate or are still considering.

11

Some eligible sovereigns also have no debt to the private sector falling due during the deferral period.

12

Belize also recently reprofiled its international sovereign bond through a consent solicitation.

13

Out of the 50 countries with outstanding debt, 25 fully relied on loans at end-2018; see also: IMF Policy Paper, The Evolution of Public Debt Vulnerabilities in Lower Income Economies, February 2020.

14

SOE debt can be part of transparent on-lending from the government. However, SOEs’ debts are generally off-budget, and can consist of debt used to finance regular business activities but have been in some cases incurred on a collateralized basis. With SOEs’ financial reporting cycles often operating at long lags, their financial position can lack in transparency.

16

Arslanalp Serkan, Bergthaler Wolfgang, Stokoe Philip and Tieman Alexander, “The current landscape”, in: Sovereign Debt: A Guide for Economists and Practitioners, IMF, 2018.

17

See forthcoming IMF Board Paper on Issues in Restructuring of Sovereign Domestic Debt.

18

Importantly, this data misses direct (non-syndicated) lending, for which no systematic data sources exist. Data sources based on new reports, government documents and contractor websites suggest that the volume of commodity-backed lending to developing countries is substantial, but the coverage of this data is insufficient to identify trends. For example, a recent report identified 52 commodity-backed loans to sub-Saharan Africa and Latin America totaling $164 billion between 2004 and 2018, (Mihalyi, Adam and Hwang, “Resource-Backed Loans: Pitfalls and Potential”, Report 27, Natural Resource Governance Institute, February 2020).

19

Generally speaking, escrow accounts in the lender’s jurisdiction are the most readily enforceable (but typically cover only a small portion of the loan), followed by assets located outside the borrowers’ jurisdiction (e.g., equity shares in a company). Movable assets (e.g., oil cargoes) can also be subject to enforcement actions. Assets within the borrower’s jurisdiction are typically harder to enforce.

20

For key considerations regarding entering into collateralized transactions, see the January 2020 G20 IFA note Collateralized Transactions: Key Considerations for Public Lenders and Borrowers. As recognized in a recent IMF Policy Paper, comprehensive sovereign debt restructurings that have taken place in a few LICs have been protracted, incomplete and non-transparent. For instance, an inadequate first restructuring agreement that raised the NPV of the loan through the imposition of fees required Chad to restructure twice (2015, 2017) in circumstances involving a commercial collateralized lender. For the Republic of Congo, the restructuring that began in early 2018 remains incomplete. See: IMF Policy Paper, The Evolution of Public Debt Vulnerabilities in Lower Income Economies, February 2020.

21

The three creditor committees were a first group representing small bondholders of international bonds with and without the enhanced CAC (Argentina Creditor Committee, ACC), a second group representing large bondholders also holding both types of bonds (Ad-Hoc Committee, AHC), and a third group representing holders of bonds without the enhanced CACs only (Exchange Bondholders, EB).

22

Differences in views/approaches among creditors also reflected the diversity in the holdings of domestic-law debt, both in USD, which is being restructuring alongside foreign law-governed debt and in pesos (not being restructured).

23

The three creditor committees were the “Ad Hoc Group”, comprising major institutional holders of Ecuador’s external sovereign debt, a “Steering Committee” (the “Minority Committee”), and an ad hoc group of holders of the 2024 bond.

24

On August 10, 2020, the authorities announced that the consent solicitation and the invitation to exchange had been approved by a majority of eligible bondholders, and that they had obtained a high participation rate in the exchange, with around 98.3 percent of the aggregate principal amount agreeing to the exchange.

25

For purposes of this paper, “new issuances” exclude reopenings of previous issuances or take-downs under programs established prior to October 1, 2014. All shares are calculated in terms of total nominal principal amount. In 2014, the IMF also endorsed revised pari passu provisions that make clear the sovereign has no obligation to make ratable payments to holdout creditors in a restructuring.

26

The figures presented in this paper are based on information available to staff through the Perfect Information database (i.e., a commercial data service). The sample includes international sovereign bonds issued between October 1, 2014 and June 30, 2020, except Euro-Area sovereign issuances (as they are required by law to include Euro Area-specific CACs), China’s domestic issuances under Hong Kong law, and GDP warrants. There may also be international sovereign bond issuances (e.g., private placements) that have not been captured by the database relied upon by staff.

27

As a share of the nominal principal amount, about 45 percent of the total stock outstanding of international sovereign bonds are governed by English law and about 52 percent by New York law.

28

Since 2014, such clauses have also been included in 91 percent and 88 percent of New York and English bonds, respectively, and 91 percent of emerging market sovereign issues and 96 percent of frontier market issues. Pari passu clauses are generally incorporated as a package with the enhanced CACs.

29

This agreement will translate into one of several upcoming amendments to the European Stability Mechanism (ESM) Treaty. As such, approval of such CACs remains subject to the signing and ratification of the amendment agreement to the ESM Treaty by each ESM Member.

30

See Eichengreen, Barry, Kletzer, Kenneth, and Mody, Ashoka, “Crisis Resolution: Next Steps”, National Bureau of Economic Research Working Paper No. 10095, November 2003; Bardozzetti, Alfredo and Dottori, Davide, “Collective action clauses: how do they weigh on sovereigns?” Bank of Italy, Economic Working Papers No. 897, January 2013; Bradley, Michael and Gulati, Mitu, “Collective Action Clauses for the Eurozone”, Review of Finance, 2013, pp. 1–58; Chung, Kay and Papaioannou, Michael, “Do Enhanced Collective Action Clauses Affect Sovereign Borrowing Costs?”, IMF, WP/20/162.

31

Chung, Kay and Papaioannou, Michael, “Do Enhanced Collective Action Clauses Affect Sovereign Borrowing Costs?”, IMF, WP/20/162; note that the empirical analysis sample is from September 2014 to March 2020 and this did not include the period when market started to show price differentiation in Ecuador and Argentina restructurings in 2020.

32

Ecuador’s 2024 bond provision allowed modifications of reserved matters with the consent of 75 percent of the aggregate amount of outstanding principal. A higher per-series voting threshold may be seen as advantageous in a restructuring for creditors, because it is easier for creditors to acquire a sufficiently high percentage of the bond series to block the deal.

33

For instance, two of the Exchange Bonds maturing in 2033 have higher coupons, as well as amortization features that require Argentina to make 20 semi-annual installments of principal payment over 10 years before maturity.

34

The NPV haircuts were 42 percent in Ecuador and 36.2 percent in Argentina.

35

Over 98 percent of creditors consented to the Ecuador exchange, resulting in 100 percent participation after the use of CACs. Ministry of Economy and Finances of Ecuador, “The Republic of Ecuador Announces Results of its Invitation to Exchange” Press Release, dated August 10, 2020. Over 93 percent of creditors consented to the Argentine exchange, resulting in over 99 percent participation (Press Release, dated August 31, 2020).

36

Minimum participation threshold is designed to assure creditors that the debtor would only proceed with the debt exchange only if a qualified majority of creditors decide to participate.

37

Argentina’s offer included a complex “minimum participation condition” which could be met at much lower participation levels (as little as 42 percent prior to the activation of CACs and about 60 percent after activation of CACs).

38

It refers to a legal technique that can be used in a debt restructuring where a majority of bondholders, when accepting the exchange offer and exiting from the existing debt instruments, can vote to modify the nonpayment terms of these existing debt instruments to make these instruments less attractive to hold and more difficult to enforce.

39

In Ecuador, bondholders who tendered certain eligible bonds received 86 percent of the accrued and unpaid interest on such bonds covering a specified time period, in the form of a zero-coupon bond maturing in 2030. In Argentina, bondholders who tendered certain eligible bonds received a USD 1.00 percent 2029 Bond or Euro 0.500 percent 2029 Bond (depending on the type of New Bonds the bondholder will receive pursuant to the offer), in an aggregate principal amount determined by reference to interest accrued and unpaid on such eligible bonds, covering a specified time period.

40

For a more detailed discussion, see Walker, Mark and Chong, Alice, “Collective Action Clauses Reexamined: Thank you Argentina”, SSRN, August 2020.

42

In New York, sovereign debtors sought to use a state statute that codified the common-law doctrine of champerty as a similar defense to lawsuits. See N.Y. Judicial Law §489. However, courts adopted a narrow interpretation, excluding the business model of distressed debt funds from the scope of prohibited activity. See, e.g., Elliott Assocs., LP v. Banco de la Nacion, 194 F.3d 363 (2d Cir. 1999); Turkmani v. Republic of Bolivia, 193 F.Supp.2d 165 (D.D.C. 2002). Further, the statute was subsequently amended in 2004 to eliminate the defense for debt purchases with an aggregate purchase price of more than USD 500,000.

43

The law’s constitutionality was upheld by the Belgian Constitutional Court against a challenge by NML Capital. Belgium, C.C., May 31, 2018, n°61/2018.

45

As an example, the composition of Venezuela’s debt means that an eventual debt restructuring would be extremely complex, with significant risk of disruptive holdout behavior: (i) in addition to its bonded debt and loans, there are other significant liabilities, including those arising from arbitral awards (investment disputes), foreign exchange allocation to the private sector, and intra-government debts (e.g., PDVSA’s debt with the central bank); (ii) PDVSA’s bonds do not contain CACs, and a share of its debt is in the form of US-issued promissory notes; (iii) a significant share of the debt owed to official bilateral creditors is likely to be collateralized with receivables from oil exports; and (iv) Venezuela’s oil cargoes and its assets abroad, most notably CITGO as well as other PDVSA refineries and oil-related receivables, could be available for attachment by commercial creditors.

46

Many countries have a debt portfolio that includes bonds (issued post-2014) with enhanced CACs and bonds (issued prior to 2014) with series-by-series or two-limb aggregated CACs. In Argentina’s 2020 exchange, bonds issued under the 2005/2010 indenture (which did not include enhanced CACs) were allowed to remain under the 2005/2010 indenture, given preference of bondholders. For a discussion of this issue, see Sobel, Mark, “Argentina and creditors enter new round”, OMFIF, June 2020.

47

These arguments were made by a group of bondholders in recent Venezuela litigation (Casa Express Corp. v. Venezuela, US District Court for the Southern District of New York). Judgment is still pending on this case.

48

This means that the parties’ legal obligations under an accelerated bond are extinguished by a court judgment on liability—the debtor’s obligation to repay principal and interest becomes merged in the judgment, and the creditor obtains the right to enforce its judgment through the judicial enforcement mechanisms, as opposed to relying on the contractual terms.

49

Weidemaier, Mark, “Venezuela, Lebanon, and Tools to De-Fang ‘Rush-In’ Creditors”, Credit Slips, February 17, 2020. The rights of bondholders holding a monetary judgment derives from the judgment itself, so arguably a withdrawal of acceleration would no longer affect these bondholders.

50

Some legal commentators have supported this view – see Weidemaier, Mark, “Judgments > CACs”, Credit Slips, February 12, 2020.

51

As reported in past papers, federal courts in New York interpreted the pari passu clause contained in the defaulted bonds of Argentina (issued prior to Argentina’s 2001 default but not tendered in its 2005 and 2010 debt restructurings) to require ratable payments to restructured bondholders and holdout creditors. This essentially prohibited Argentina from making payments on its restructured bonds unless Argentina paid in full the principal and interest owed and past due on the original unrestructured claims. Argentina settled the claims with the holdout creditors in 2016. See IMF 2013 and IMF 2016. As noted above, the vast majority of bonds issued since 2014 include revised pari passu clauses that make clear “ratable” payments to holdout creditors are not required. Moreover, New York courts have since clarified that a sovereign’s decision to pay some creditors but not others in and of itself does not give rise to a breach of the pari passu clause, and some other acts by the sovereign are necessary. See White Hawthorn, LLC et al. v. the Republic of Argentina and Ajdler v. Province of Mendoza.

52

For example, a small group of creditors in the Ecuador (2020) restructuring filed a complaint in federal court in New York the week the exchange was due to close. Ecuador extended the deadline by one business day to allow time for the judge to rule. Plaintiffs’ request to further delay the closing of the exchange was denied, with the judge ruling that the lawsuit was unlikely to be successful on the merits. Contrarian Emerging Markets L.P. et al. v. The Republic of Ecuador, 20 Civ. 5890-VEC (SDNY 2020).

53

This means that the bank lender remains the lender on record, but all economic rights and risks would have been transferred to the entity which entered into the participation agreement. In general, the bank lender may have to obtain the consent of the participant before agreeing to amendments to payment terms.

54

Syndicated loan agreements generally provide that certain amendments may be taken by the majority (or super-majority) of the lenders from time to time (often referred to as the majority or super-majority lenders provisions). The majority lenders will usually be defined as those members of the syndicate at the relevant time which (taken together) hold a specified percentage of the total commitments under the senior facilities (typically two-thirds by commitment). Supermajority voting might also be included in an agreement where a higher consent threshold is deemed appropriate for particularly significant amendments or waivers, such as for proposed changes in the security structure. However, certain decisions are considered ‘all lender’ decisions. These generally include (amongst other things) changes to the principal amount of any senior facility, the amounts payable by way of interest or fees and/or the maturity of any senior facility. Guide to Syndicated Loans and Leveraged Finance Transactions, Loan Market Association.

56

See Weidemaier, Mark 2019, “Restructuring Italian (Or Other Euro Area) Debt: Do Euro CACs Constrain or Expand the Options?”, UNC Legal Studies Research Paper.

57

See Buchheit, Lee and Gulati, Mitu, 2012, “Restructuring a Sovereign Debtor’s Contingent Liabilities”, SSRN.

59

Collateral-like features (which are not legally collateral but give the creditor similar features as collateral such as oil pre-payment agreements) may have similar implications.

60

Exceptions are generally limited to two types: first, liens on a property which is the sole security for payment of the purchase price of such property (so called acquisition financing); and second, liens arising in the ordinary course of banking transactions to secure a debt of not more than one year.

61

If the clause is breached, creditors may pursue several remedies depending on the contractual terms and applicable law. For instance, violation of the NPC may constitute an event of default and the creditor may decide to accelerate and pursue litigation. Creditors may also seek an injunction to prevent a breach. The creditor may require the debtor to grant security equally and ratably to it or grant an equivalent security. Other possible solutions to an NPC violation will vary depending on the circumstances and the clause but may include early repayment by the borrower of all outstanding financing to the lender; or the borrower seeking to restructure the relevant transaction to make it unsecured or repaying early the violating transactions.

62

See joint IMF-WB notes for the G20s on Strengthening Public Debt Transparency (June 14, 2018) and on Public Sector Debt Definitions and Reporting in Low-Income Developing Countries (January 31, 2020).

63

In Mozambique, two large previously unreported external loans were revealed to IMF staff in April-June 2016. The two loans, amounting to US$1.15 billion (9 percent of GDP at end-2015), were contracted in 2013 and 2014 by two SOEs with government guarantees, allegedly for maritime projects.

64

Sovereigns have not pursued voluntary liability management operations to replace bonds without enhanced CACs with bonds with enhanced CACs due to a number of concerns such as cost.

65

See paragraphs 23–25 of Progress Report on Inclusion of Enhanced Contractual Provisions in International Sovereign Bond Contracts, September 2015. The bonds in both the recent Ecuador and Argentina restructurings were all issued under trust structures.

66

See Buchheit, Lee and Hagan, Sean, “From Coronavirus Crisis to Sovereign Debt Crisis”, Financial Times, March 26, 2020.

67

Developing entire standardized agreements is a related proposal that is broader in scope, but such a proposal presents several challenges—most importantly, that the costs of doing so would likely outweigh the benefits it might bring. Model agreements may help to streamline efforts in producing legal documentation. However, the main transaction documents in a debt restructuring do not lend themselves towards standardized templates, given that many elements are borrower and deal specific. The main impediments to a successful and efficient debt restructuring relate to issues on commercial terms, intercreditor equity and effective debtor-creditor communication, among others. While legal documentation is important, it is not determinative of the speed and success of a debt restructuring. Moreover, given that an extensive consultation process with a wide range of market participants is required, it will likely take significant time to develop model agreements, which detracts from their benefits in an upcoming sovereign debt restructuring scenario.

68

The IMF has occasionally decided to set conditionality to prevent the contracting of new collateralized debt in certain IMF lending arrangements dependent on the member specific circumstances and when critical for achieving the program goals. The IMF may consider this issue in the broader forthcoming review of the IMF Debt Limits Policy.

70

Given their small size, distressed debt funds cannot contribute much liquidity to the second market. While they offer other creditors an opportunity to off-load their exposures at very low prices, the benefits of this option are ambiguous (sovereigns would have lost market access, and the conditions under which mainstream investors may want to sell to holdouts may have been created by the holdouts themselves). At the same time, however, it is critical that anti-holdout legislation does not discourage the trading of sovereign debt for reasons unrelated to holdout behavior.

72

This is because debt relief from the debtor perspective ought to be evaluated at discount rates that do not embody a crisis risk premium, whereas the risk premia that the market uses to discount the value of newly issued debt instruments in a restructuring could be very high. In these circumstances, it may make sense for the debtor country to offer creditors cash, short dated instruments, or partially collateralized long instruments, rather than only risky long instruments. However, this requires upfront financing. So long as IFIs charge an appropriate risk premium to the debtor country, providing such financing could be an efficient use of IFI funds.

73

IMF-World Bank Staff Note on the Implementation Update on the joint IMF-WB Multi-Prong Approach for Addressing Debt Vulnerabilities (forthcoming).

74

For example, the Debt Management Facility is a trust fund administered jointly by the World Bank and the IMF for LICs that provides customized advice on sovereign debt management. Since 2009, the Debt Management Facility has supported over 280 TA missions in about 80 countries and 14 subnational entities; see https://www.worldbank.org/en/topic/debt/brief/debt-management-facility.

76

It is sometimes argued that a large number of sovereign debt restructurings in a short period of time could overwhelm the system. While such a bunching of restructurings could be problematic, a bigger concern is that creditors with claims on many countries undertaking restructurings could face large aggregate losses. While the likelihood that this could threaten financial stability in advanced economies is lower than in the 1980s (when most sovereign debt claims were held by a small number of large banks), this could make sovereign debt restructurings more protracted and difficult. The willingness and ability of commercial banks to accept and write off such losses was the main reason for the long time that it took to resolve the debt crisis of the 1980s.

77

See, e.g., Buchheit, Lee and Hagan, Sean, “From Coronavirus Crisis to Sovereign Debt Crisis”, Financial Times, March 26, 2020, which proposes amending sovereign immunity laws to explicitly provide judges with the discretion to stay suits in certain situations—for example where the IMF has judged debt to be unsustainable—as a deterrent to disruptive litigation.

78

See, e.g., Buchheit, Lee and Gulati, Mitu, “Sovereign Debt Restructuring and U.S. Executive Power”, Capital Markets Law Journal, Volume 14, Issue 1, January 2019, pp. 114–130; Hagan, Sean, “Sovereign debt restructuring: The centrality of the IMF’s role,” Peterson Institute for International Economics Working Paper 20–13, July 2020.

79

U.N. Security Council resolutions are binding on all members, and each member would be required to enact into their domestic laws the immunities set forth in the Resolution. If assets are concentrated in particular jurisdictions, laws or executive actions providing immunity for assets within those jurisdictions could also be designed to have a similar effect in those jurisdictions as a U.N. Security Council resolution. Conversely, international law options could in principle focus on issues other than immunity for assets.

80

Under the laws of many countries, including the U.S. and U.K., sovereign assets already enjoy broad immunity from attachment, with select exceptions, such as assets used in connection with commercial activity (see. e.g., 28 USC s.1602 et seq.). In this regard, steps to broaden sovereign immunity via additional measures would be particularly useful only for countries with significant commercial assets abroad.

Same Series

Other IMF Content

Other Publishers

  • Collapse
  • Expand
Cover Policy Papers
The International Architecture for Resolving Sovereign Debt Involving Private-Sector Creditors—Recent Developments, Challenges, And Reform Options
Author:
International Monetary Fund
Volume/Issue:
Volume 2020: Issue 043
Publisher:
International Monetary Fund
ISBN:
9781513557472
ISSN:
2663-3493
Pages:
52
DOI:
https://doi.org/10.5089/9781513557472.007
  • <sup><xref ref-type="fig" rid="RA01fig01">Figure 1.</xref></sup>
    View in gallery
    Figure 1.

    EMDE External Public and Publicly Guaranteed Debt by Creditors 1/

  • <sup><xref ref-type="fig" rid="RA01fig02">Figure 2.</xref></sup>
    View in gallery
    Figure 2.

    DSSI Countries: External Bonded and Other Non-official PPG Debt, 2018

    (in billions of USD)

  • <sup><xref ref-type="fig" rid="RA01fig03">Figure 3.</xref></sup>
    View in gallery
    Figure 3.

    SOE and Sub-National International Bond Issuance

  • <sup><xref ref-type="fig" rid="RA01fig04">Figure 4.</xref></sup>
    View in gallery
    Figure 4.

    Government Debt Securities Issues by Jurisdiction, end 2017

  • <sup><xref ref-type="fig" rid="RA01fig05">Figure 5.</xref></sup> <sup><xref ref-type="fig" rid="RA01fig05">Figure 5.</xref></sup>
    View in gallery
    Figure 5.

    Developments in EM Local Currency Government Debt

  • <sup><xref ref-type="fig" rid="RA01fig06">Figure 6.</xref></sup>
    View in gallery
    Figure 6.

    Public Sector Collateralized Bonds and Syndicated Loans

    (US$ billion)

  • <sup><xref ref-type="fig" rid="RA01fig07">Figure 7.</xref></sup>
    View in gallery
    Figure 7.

    Collateralized Bonds and Loans as a share of total Syndicated Bonds and Loans, average 2002–17

  • <sup><xref ref-type="fig" rid="RA01fig08">Figure 8.</xref></sup>
    View in gallery
    Figure 8.

    Ecuador: Price Developments of Eurobonds under Restructuring

    (in percent)

  • <sup><xref ref-type="fig" rid="RA01fig09">Figure 9.</xref></sup>
    View in gallery
    Figure 9.

    Argentina: Price Developments of Eurobonds under Restructuring

    (in percent)

  • <sup><xref ref-type="fig" rid="RA01fig10">Figure 10.</xref></sup>
    View in gallery
    Figure 10.

    Low Income Country Eurobond Issuers: Minimum Bond Purchase for Holdouts 1/ 2/

    (billion US$)

  • <sup><xref ref-type="fig" rid="RA01fig11">Figure 11.</xref></sup>
    View in gallery
    Figure 11.

    Debt-to-GDP Ratio

    (Average, % of GDP)

  • <sup><xref ref-type="fig" rid="RA01fig12">Figure 12.</xref></sup>
    View in gallery
    Figure 12.

    Debt-to-GDP Ratio – DSSI-Eligible Countries

    (Average, % of GDP)

Figure 1.

EMDE External Public and Publicly Guaranteed Debt by Creditors 1/
Figure 2.

DSSI Countries: External Bonded and Other Non-official PPG Debt, 2018

(in billions of USD)
Figure 3.

SOE and Sub-National International Bond Issuance
Figure 4.

Government Debt Securities Issues by Jurisdiction, end 2017
Figure 5.

Developments in EM Local Currency Government Debt
Figure 5.

Developments in EM Local Currency Government Debt
Figure 6.

Public Sector Collateralized Bonds and Syndicated Loans

(US$ billion)
Figure 7.

Collateralized Bonds and Loans as a share of total Syndicated Bonds and Loans, average 2002–17
Figure 8.

Ecuador: Price Developments of Eurobonds under Restructuring

(in percent)
Figure 9.

Argentina: Price Developments of Eurobonds under Restructuring

(in percent)
Figure 10.

Low Income Country Eurobond Issuers: Minimum Bond Purchase for Holdouts 1/ 2/

(billion US$)
Figure 11.

Debt-to-GDP Ratio

(Average, % of GDP)
Figure 12.

Debt-to-GDP Ratio – DSSI-Eligible Countries

(Average, % of GDP)