Abstract
Crisis Resolution in the Context of Sovereign Debt Restructuring - A Summary of Considerations
I. Introduction
1. This paper discusses a range of economic policy issues that arise when a sovereign debt restructuring is part of the crisis resolution framework. Sovereign debt restructuring will typically take place in very difficult macroeconomic circumstances, requiring a policy response that confronts the widespread disruption that is likely to occur. Efforts to restructure sovereign debt will generally need to be complemented by a wide range of other measures to help a country to emerge from the crisis and to regain sustainability. A successful restructuring will require coordination both with the country’s economic program and between a diverse group of domestic and external creditors. This paper surveys this broad policy agenda as part of the Fund’s work on crisis prevention and resolution, and serves as a companion to the staff’s recent paper on the Sovereign Debt Restructuring Mechanism (SDRM).1
2. Not all crises originate with the sovereign. Crises can emerge from debt difficulties in a number of areas, notably creditor’s fears of generalized insolvency or illiquidity of the banking system, a vulnerable financial position of the corporate sector, as well as doubts about the ability of a sovereign to service or refinance its internal or external obligations. Experience shows that, while a crisis may originate in imbalances or balance sheet mismatches of a specific sector of the economy, it may rapidly spread to others and lead to an abrupt outflow of capital. Corporate insolvency or sovereign difficulties affect the banking system, banking sector problems involve implicit or explicit guarantees of the sovereign, and attempts by resident and nonresident investors to seek safer havens deplete foreign exchange reserves and exacerbate external debt servicing problems. The size of these shocks and the speed with which they occur and spill across sectors pose complicated policy challenges, particularly as crises, regardless of their origins, often result in considerable financial difficulties for the sovereign.
3. A sovereign debt restructuring can therefore be necessary in two broad sets of circumstances. First, the sovereign’s unsustainable debt can itself be the source of the country’s financial difficulties and must therefore be restructured. Second, the crisis can originate in the nonsovereign sector, but the cost of the crisis—both in terms of lost output and the liabilities the government incurs during the crisis—makes the sovereign’s debt unsustainable.
4. Among the many important issues involved in addressing sovereign restructuring cases, this paper discusses three sets of broad considerations.2 First, how to create a framework for assuring that the debt structure that emerges from the debt restructuring is consistent with a country’s adjustment effort and restores balance of payments viability? This requires both determining the appropriate scale and path of fiscal and external adjustment, and developing a framework for coordinating the sovereign’s restructuring agreements with different groups of creditors, where separate restructurings are needed. The SDRM would help improve the prospects for rapid progress toward a restructuring of sovereign debt, and would facilitate the coordination of any restructurings that need to be undertaken outside the framework of the SDRM.
5. Second, how to limit the disruptions associated with a sovereign restructuring on other domestic economic sectors? In general, sovereign restructurings are likely to take place amidst, or precipitate, a broader loss of confidence that could quickly escalate to a full blown currency and banking crisis. Economic policies should be tailored so as to minimize the unavoidable fallout and, where possible, to avert the need to restructure nonsovereign liabilities. In some cases, emergency measures may be required to protect the functioning of the banking system and reduce the rate of general capital flight, to avoid derailing the authorities’ adjustment efforts and impeding the restructuring process.
6. Third, what should be the role of the Fund during a sovereign debt restructuring? A sovereign undertaking a restructuring may face significant financing needs, despite the temporary relief provided by the accumulation of arrears in the event of a default. These circumstances pose particular challenges to the design of Fund programs, and raise difficult choices for the appropriate access to, and use of, Fund resources.
7. This paper addresses these issues and seeks to lay out considerations that have a bearing on the choice of policies in dealing with sovereign debt crises. The paper is organized as follows: Section II summarizes the main channels through which sovereign debt restructurings affect the domestic economy; examines how to define an appropriate adjustment path to help ensure that the new debt structure is sustainable over the medium-term; and explores policies that can help to limit the spill-over effects of a sovereign debt restructuring into other economic sectors. Section III discusses how best to coordinate the restructurings of the various categories of a sovereign’s debt. Section IV addresses issues pertaining to the role of Fund financing during a sovereign debt restructuring. Section V suggests issues for discussion.
II. Policy issues in restoring sustainability
A. The Economic Implications of Debt Restructurings
8. A sovereign debt restructuring (affecting external or domestic debt, or both) can pose significant strain on the domestic economy.3 The strain is usually transmitted through the direct effect on the balance sheets of the domestic holders of debt, and through the reaction of investors and the public to the restructuring as they seek to protect the value of their assets. The precise effects will generally depend on three main factors:
9. First, the losses sustained by domestic holders of government securities. In the first instance, a restructuring would affect their income position depending on the exposure of each holder to debt that is subject to restructuring. Losses will be amplified by the extent to which the restructuring reduces the face value of claims. Mark-to-market investors as well as investors in need of liquidity who had purchased bonds at, or near, face value would suffer immediate losses as the increasing risk premia lead to a fall in bond prices.
10. Second, the degree of asset substitution following the shock to confidence. Crises often lead to rapid demonetization and currency substitution. A loss of confidence in the sovereign’s “promise to pay” could cause money demand to fall sharply and trigger a substitution from domestic currency to foreign currency denominated assets. Moreover, foreign holders of assets, regardless of whether they are subject to restructuring or not, are likely to rush to the exit. Residents will seek to protect their real wealth not only by switching to foreign currency denominated assets, but also by moving capital out of the country. The demand for foreign exchange could quickly mushroom and lead to considerable pressure on the exchange rate.
11. Third, access to capital markets is likely to be reduced. Reflecting the increased country and market risk, new financing is likely to be curtailed, or available only at very expensive terms for both sovereign and non-sovereign borrowers, for a sustained period of time.
12. The banking system and nonfinancial corporate sector are often vulnerable to large swings in exchange and interest rates. As discussed in section III, the critical issue is whether in the run up to a sovereign restructuring, the combination of these effects could weaken public confidence in the banking system to a point that triggers a deposit run. Problems in financial institutions disrupt credit allocation, worsen capital flight, and deepen the crisis.
13. Although it is hard to generalize from recent experience, countries faced with an unsustainable debt burden could typically be expected to go through three phases: (i) an initial chaotic period prior to and in the immediate aftermath of the decision to restructure (which may follow a default) in which the economy is hit by a variety of shocks; (ii) a stabilization phase in which efforts are made to contain the fallout of the restructuring decision; (iii) and a debt-restructuring phase in which the stock of debt is brought back into a sustainable range. These phases may partly overlap in individual country cases. For instance, measures to contain the fallout may already be in place in those cases where a banking or currency crisis preceded the sovereign’s debt restructuring decision (e.g., Ecuador). Also, debt-restructuring negotiations between the sovereign and its creditors may have been initiated prior to the sovereign’s default. Indeed, the fallout of a restructuring decision as well as the reputational damage associated with it could in principle be contained by approaching creditors early, planning a roadmap ahead of the restructuring announcement, and avoiding arrears (Box 1).
14. Nevertheless, there is no stylized model of how sovereign debt crises unfold. Not all of these effects emerge in each debt crisis. The fallout of the restructuring in individual cases depends on the initial conditions, the strength of sectoral balance sheets and the exposure of each sector to risks, as well as the policy response adopted by the authorities. It is to the latter that the paper now turns.
B. Implications for Policies
15. The overall objective in sovereign debt restructuring cases should be to promote the adjustment to a new sustainable equilibrium and avoid the perpetuation of an unsustainable policy mix. Debt restructuring is not a substitute for adjustment. In most cases, sovereigns will need to undertake substantial policy reform and adjustment to provide a credible exit from the crisis. Against the background of intensified pressures in the economy, the policy response needs to:
Restore confidence. A comprehensive set of economic policy measures to deal with the crisis should be announced as soon as possible, as delays and perceptions of policy paralysis would further undermine confidence. These measures should be designed to address the underlying vulnerabilities that produced the conditions of unsustainability.
Limit spillovers to other sectors. Policies should aim particularly at minimizing potential disruption to the domestic financial system, which could exacerbate dislocations in the domestic economy.
Stem capital outflows. Given the uncertainty about market reaction, and the risk that macroeconomic policies may not be sufficiently credible to contain capital flight, consideration may have to be given to complementing the restructuring decision with temporary measures to stem capital outflows, including exchange restrictions and administrative bank measures. Although there is a range of complications to the design of such measures, in some cases they may help provide a respite until the fog of the initial panic reaction clears and corrective measures take hold.
The restructuring should contribute to a credible exit from the crisis. The authorities need to communicate clearly to the public how the restructuring would help the country regain medium-term sustainability. A selective restructuring that is judged not to provide reasonable prospects for sustainability could undermine confidence, as more restructurings would be seen to be necessary in the near future.
C. Defining the Adjustment Path
16. A definitive resolution of the crisis requires reestablishing sustainability. The restructuring proposal should ensure that the resulting debt stock—the sum of restructured, unrestructured, and new debt—is sustainable. The new debt structure will need to be consistent with projected levels of primary fiscal surpluses, growth, real interest rate and exchange rate developments, as well as the timing of renewed capital market access, all of which are interrelated and affected by the realism and acceptability of the restructuring proposal. Therefore, a central element in a sovereign’s debt restructuring strategy should be to draw up a credible plan, which is acceptable to creditors, in which targets are devised and strategies to reach them are formulated (Figure 1).
Coordinating Framework for Sovereign Debt Restructuring
Citation: Policy Papers 2003, 005; 10.5089/9781498329880.007.A001
Recent Sovereign Debt Restructurings
There have been at least five sovereign debt restructuring cases since the late 1990s. Russia and Ecuador restructured most of their debt following a formal default, while Ukraine and Pakistan restructured their debts in the shadow of default. Argentina, which initially planned a voluntary debt exchange, was forced to default following the first leg of its debt exchange. Moldova completed the restructuring of its sovereign debt very recently and it is not covered in this box.
Russia (1998-2000)
In response to a rapidly deteriorating economic situation, resulting mainly from the government’s inability to implement important fiscal reforms, on August 19, 1998 the authorities announced a devaluation of the ruble, a suspension of payments on the sovereign’s domestic debt, and a unilateral moratorium on private sector payments on external liabilities enforced through extensive capital and exchange controls (debt held by households and the central bank was excluded from the exchange restrictions). These announcements, and the subsequent dissolution of the government, triggered a banking crisis (see Box 2).
There was no overall coordinating framework for the restructuring of Russia’s various classes of debt. Domestic debt, London Club debt, and Paris Club debt were restructured in separate processes. Eurobonds were excluded from the restructuring altogether. Domestic ruble debt, amounting to 10.8 percent of GDP (of which 7.8 percent of GDP was held by domestic banks and the remainder by non-residents) was restructured first. An initial restructuring offer for London Club debt in September 1998 failed, triggering protracted negotiations. Eventually, on February 11, 2000, a restructuring agreement was announced by the London Club’s Bank Advisory Committee, under which US$31.8 billion in Soviet era debt was exchanged for US$21.8 billion in new bonds (completed by August 2000). Private creditors pushed for a new Paris Club rescheduling on similar terms but the notion of “reverse comparability” was rejected. The improvement in oil prices took pressure off Russia’s debt service profile. The Paris Club (with exposure of roughly US$37.5 billion) ultimately agreed to a rescheduling but not to a debt reduction.
Ukraine (1998-2000)
While Ukraine’s overall stock of external debt was not particularly large (public external debt stood at roughly 39 percent GDP in 1999), low levels of reserves and a spike in debt-service payments through 2001 made a debt restructuring unavoidable. A selective restructuring of domestic debt held by banks in August 1998 was followed by the restructuring of two bond-like instruments held by non-residents in September and October 1998, and a further restructuring in June 1999. After these piecemeal arrangements, the debt exchange in April 2000 tried to deal more comprehensively with the short maturity of Ukraine’s bonded debt.
In total, the restructuring covered US$2.5 billion of external (Eurobond) debt and US$0.3 billion of domestic debt, representing about 9 percent of GDP. Of this amount, about 1.3 percent of GDP was held by domestic banks. A large part (roughly 50-60 percent) of the Eurobonds was held by retail investors. To limit the outflow of capital, exchange controls were imposed in September 1998, with no deposit freeze.
Pakistan (1999)
Sanctions imposed by bilateral official creditors following Pakistan’s nuclear tests in May 1998 exacerbated the fragile external position, and triggered a debt crisis that had been looming during much of the 1990s. The financial instability that followed led the authorities to impose a deposit freeze on most foreign currency deposits (amounting to roughly US$7 billion). In January 1999, Pakistan concluded an agreement with the Paris Club, reducing debt service by US$3.3 billion, or 5 percent of GDP, in the period up to end-2000. Following an agreement with London Club creditors in June 1999, with a credible threat of default Pakistan launched an exchange of Eurobonds in November, in order to fulfill the comparability of treatment clause included in the Paris Club agreement.
The Eurobonds had a face value of US$608 million (about 1 percent of GDP), of which one third estimated to have been held by residents (11 percent by domestic banks) and the remainder by financial institutions and retail investors in the Middle East. In addition, Pakistan was able to reschedule US$512 million of short-term trade credits and US$415 million of medium-term commercial credits. Domestic debt remained unaffected. Provisions to further tighten capital controls were introduced in June and October 1999.
Ecuador (1999-2000)
When a protracted banking crisis ultimately evolved into a bank run in March 1999, the government declared a bank holiday, deposits were frozen and the exchange rate was floated as pressures on the sucre mounted (no capital controls were imposed.) In September 1999, under continuous financial pressure, the government defaulted on its Discount Brady bonds while staying current on its other bonds. However, holders of the Discount Brady bonds voted to accelerate their claims, forcing Ecuador to default on the other Brady bonds and its Eurobonds. In January 2000, amid political turbulence that resulted in a new government, a new economic plan was announced based on full dollarization of the economy.
On July 27, 2000, almost eleven months following the initial default, Ecuador announced a comprehensive exchange offer for its external debt (with a face value of US$6.5 billion). Domestic public debt maturing between September 1999 and end-2000 (both in domestic and foreign currencies) was restructured separately (with a face value of US$346 million), as were external credit lines to closed banks (US$80 million). The total debt restructured was equivalent to about 50 percent of GDP, of which roughly one tenth was held by residents and the bulk by institutional investors in London and New York. In September 2000, Ecuador reached a rescheduling agreement with the Paris Club.
Argentina (2001-02)
With very high and rising spreads making it increasingly difficult to meet debt-service payments on rolled-over debt, Argentina announced a two-phase approach in late October 2001 to restructure its roughly US$100 billion of domestic and external debt owed to private creditors. Phase 1 was aimed at domestic resident investors and involved the exchange of US dollar and Argentine peso bonds into new government-guaranteed loans (limiting the attractiveness to foreign creditors who have a preference for more marketable securities). The exchange, carried out in December 2001, involved approximately US$42 billion in sovereign debt and US$16 billion in provincial debt. Of the exchanged federal bonds, US$13 billion came from domestic banks’ on own accounts, US$11 billion from their clients, and US$17 billion from pension funds.
By end-December, before Phase 2 could be initiated to restructure the remainder of mainly foreign-held sovereign debt, the financial and political situation had deteriorated considerably, and Argentina announced a moratorium on debt not included in Phase 1. While debt service was to be maintained on the loans issued in Phase 1, the general pesoization of domestic contracts in March 2002 included the loans of Phase 1. Several domestic debt operations were conducted between May-September 2002 (including deposit exchange schemes and bonds issued to banks to compensate them for the asymmetric pesoization of assets and liabilities) but little progress has been made in restructuring foreign-held sovereign debt. On November 14, the World Bank announced that Argentina failed to make an amortization payment of around US$715 million. In the event, the payment was made in January 2003, which enabled the World Bank to reactivate its lending operations to Argentina. A comprehensive solution to Argentina’s debt problems is still pending.
17. Designing a consistent adjustment program is a complicated task because of the trade-offs and the feedback loops between debt restructuring and the adjustment parameters. The adjustment program will have to be put together in the context of a comprehensive and integrated policy response to deal with fiscal and financial vulnerabilities, and the economic disruption associated with the restructuring decision. The central component of the adjustment path is the projected evolution of primary fiscal surpluses. For the sovereign, there may be an incentive bias toward sizeable debt reduction from external creditors rather than adjustment. However, a weak adjustment effort may not be accepted by creditors as it would be seen as an attempt to place the burden for restoring sustainability disproportionately on them, and it is likely to impair the country’s future access to capital markets. On the other hand, an unfeasibly ambitious adjustment effort will not be credible. Given the trade-offs involved, the right balance will need to be found between restructuring terms that creditors can accept and an adjustment effort (and debt service profile) that the sovereign can realistically deliver.
18. Decisions would need to be made about the magnitude of the immediate adjustment as well as the primary fiscal position over the medium term.4 In principle, a degree of front-loading the path of primary surpluses could help signal the authorities’ commitment to restore sustainability. However, front-loading the adjustment may be difficult given the macroeconomic dislocation associated with the restructuring. Guidance on the magnitude of the adjustment planned in the medium-term—importantly, the realism of expenditure and revenue assumptions, future growth rates and the political feasibility of reforms - could be derived from (i) the country’s recent history of budget surpluses/deficits; (ii) cross-country experience of adjustment efforts in similar circumstances; (iii) the feasibility of the measures needed to reach the targeted position in the context of a country’s circumstances; (iv) the level of adjustment that is consistent with the growth objectives; and (v) the size of spending reductions that is sustainable over the medium term.
19. The path of primary surpluses will be based on judgments on a number of endogenous variables. Important assumptions will need to be made about the timing and pace of re-access to domestic and international capital markets, the real interest rate, and the medium-term growth rate of the economy. For a country forced to restructure its debt or emerging from default, any medium-term assumptions would be subject to considerable uncertainty. There are also important interactions between the terms in the debt sustainability equation—the debt stock, interest rates, and growth. For example, the marginal cost of funds to a defaulted sovereign will depend, among other things, on the size of the restructured debt stock. And medium-term growth prospects will depend, among other factors, on the real interest rate on new financing, which is itself a particular source of uncertainty. 5 Judgments concerning these variables could be a source of disagreement between the authorities and creditors, given the sensitivity of the size of debt reduction to the choice of parameters.
20. The fallout associated with the debt restructuring may create additional uncertainty about the sustainable level of future debt-service payments. The calculation of the level of public debt consistent with long-term solvency is derived from steady-state conditions, and initially conditions are anything but that. Three effects deserve particular consideration. First, to the extent that the exchange rate depreciates sharply and overshoots its long-run equilibrium value, the debt-to-GDP ratio will be overstated and may exaggerate the need for debt reduction. In calculating the sustainable level of debt, judgments will need to be made about the likely reversal of the initial overshooting. Second, the total cost of bank restructuring may be significant but uncertain as the deterioration of the banks’ balance sheets materializes over time.6 Third, there is uncertainty about the magnitude of economic dislocation and the social needs in the aftermath of the crisis, given that the act of default or restructuring is likely to exacerbate temporarily the already poor initial macroeconomic conditions.
21. Despite the uncertainties involved, it is particularly important that the restructuring provides reasonable prospects for a definite resolution of a country’s debt problem. The design of the debt-restructuring package must be acceptable to creditors, and the new debt service profile resulting from it should help ensure that a country’s exposure to rollover risks is manageable and that exposure to interest rate and exchange rate risks would not undermine financial stability. In combination with the authorities’ adjustment efforts, the restructuring should provide for a realistic and credible exit from the crisis.
22. In addition to determining a sustainable fiscal adjustment path, restoring credibility requires a monetary policy framework to help anchor inflation expectations. To the extent that reducing the debt burden through inflation is not a desirable option, policies in the aftermath of the crisis need to restore a degree of nominal stability. This is likely to prove a difficult challenge as:
severe fiscal and banking problems and the erosion of international reserves are prone to undermine the credibility of an exchange rate anchor;
the instability of money demand complicates the implementation of monetary targeting strategies, and low interest elasticity of money demand may require unsustainable interest rate changes in an environment dominated by high risk of sharp depreciation and default; and
inflation targeting is confronted with the difficulty of forecasting inflation with any confidence during and in the aftermath of a debt crisis.
Under these conditions, and in view of the likely need for the exchange rate to adjust to a new equilibrium supportive of a resumption of economic growth, monitoring monetary aggregates may still play a useful supportive role in restoring confidence, preventing an acceleration of inflation, and avoiding the exhaustion of international reserves. To the extent that balance sheet vulnerabilities increase the costs associated with exchange rate overshooting, this approach could in some cases be complemented by the temporary imposition of exchange restrictions (see Section HE).
D. Dealing with Banking Crises7
23. In the event of a sovereign debt restructuring, the financial position of banks and other financial intermediaries could be severely affected. In addition to exposure to losses on their holdings of government debt instruments, banks may be vulnerable to a deterioration in credit quality as a result of debt-servicing difficulties in the household and corporate sectors; to exchange rate risk due to currency mismatches on and off balance sheets; and to deteriorating liquidity conditions as a result of deposit withdrawals and the interruption of interbank credit lines. Regardless of the financial position of individual banks, the loss of explicit or implicit government guarantee and the worsening economic outlook could exacerbate depositors’ fears about the availability or value of their deposits in real terms which may trigger sharp deposit withdrawals. Experience indicates that the combination of these factors may contribute to widespread bank insolvencies, which may exacerbate dislocations in the domestic economy and problems with banks’ external obligations (Box 2).
24. Some of these effects also take place in conventional banking crises. However, the restructuring of an unsustainable sovereign debt has an especially destructive effect on banks’ balance sheets and may complicate the resolution of a crisis. One of the most complex set of issues confronting sovereigns that need to embark on a restructuring of their debt concerns the treatment of domestic debt, particularly when a large share of this debt is held by domestic banks. In some circumstances, it may be appropriate to exclude domestic debt from a restructuring, in view of the implications for the financial system and the domestic economy. In cases, however, where domestic banks are among the principal creditors to an unsustainable sovereign, it may be impossible for the sovereign to obtain the debt-service relief that it needs without a reduction in the real debt it owes to the domestic banking system. A reduction in the value of government securities in banks’ portfolios will test the capacity of banks’ equity capital to withstand the balance sheet shock and the associated reduction in earnings.
25. The share of claims on the sovereign in bank assets, the extent of the debt reduction, and the initial level of capital are key parameters in evaluating the impact on banks. In a crisis situation, banks may already be experiencing increased credit risk from the deteriorating position of corporates. There may already be liquidity drains and emerging maturity mismatches. The reduction in value of claims on the sovereign makes the situation worse.8 In these circumstances, it may not be reasonable to expect banks to hold sufficient capital to withstand all shocks that may occur in a systemic crisis of this nature, or to be able to raise additional capital in unsettled market conditions. Therefore, finding the least costly way of obtaining the needed reduction in the sovereign’s real debt burden is a major policy challenge. All options are likely to be very costly, and it may not be possible to avoid imposing losses on depositors.
26. The initial priorities in protecting the banking sector during a sovereign restructuring crisis should be on stabilizing the system and restoring confidence. As is the case in a standard banking crisis, the immediate policy response should aim at stopping bank runs; safeguarding the payments system; and minimizing disruptions to credit flows, in order to preserve at least core banking functions and activities. To be effective, emergency measures need to be backed as soon as feasible by a coherent macroeconomic stabilization framework, and by comprehensive bank restructuring programs. However, in the midst of a crisis, it may be very difficult to ascertain the true solvency conditions of individual banks. This complicates the task of assessing what could constitute a core banking system on which to concentrate scarce resources in order to protect the payment system and preserve financial intermediation services.9
Experience with Banking Crises in Selected Sovereign Restructuring Cases
Russia (1998): Withdrawals of bank deposits started in December 1997 and accelerated in July-August 1998, reflecting contagion from the Asian crisis and inconsistent fiscal and exchange rate policies. Withdrawals affected many banks, including the state-owned Sberbank, Russia’s largest holder of household deposits (85 percent of total deposits).
Following the debt standstill announced on August 17, 1998 (see Box 1), deposit withdrawals continued and between June-September Sberbank lost approximately 19 percent of its domestic deposits. Interbank market activity and the payments system slowed dramatically, and foreign credit lines dried out. The central bank announced a blanket guarantee for all household deposits, and required that deposits held by six large Moscow banks (13 percent of total deposits) be transferred to Sberbank (dollar deposits were also covered, but these would be paid out in rubles due to the lack of foreign reserves). CBR provided considerable support to Sberbank and selected financial institutions, including through the creation of an overnight unsecured loan facility and an easing of required reserves. In subsequent months, the authorities put in place a comprehensive bank restructuring strategy, which involved closing a large number of banks and helping rebuild a core group of viable institutions.
Ecuador (1999): Confidence in the Ecuadorian banks deteriorated rapidly in early 1999, following intervention in the country’s largest bank in December 1998 and the failure of five small banks during the first two months of 1999. Loss of confidence was exacerbated by the slow response of the Deposit Guarantee Agency under the blanket guarantee approved in December 1998. The fear of a systemic meltdown increased after the country’s second largest bank became illiquid. In response, the government declared a banking holiday on March 5-11, 1999 and froze most of deposits for one year. Subsequent political pressure led to the progressive easing of the freeze.
The default on the external debt in September 1999 prompted a second currency crisis and further bank runs in the last months of 1999. The authorities’ decision to dollarize in January 2000 and the announcement of a Fund-supported program succeeded in restoring confidence. In March 2000 deposits were unfrozen. Although some flight to quality occurred, the banking system was able to retain most of the deposits. The cumulative costs of the banking crisis were, however, very large, with estimates in the order of 22 percent of GDP.
Argentina (2001-02): Concerns about the sustainability of the authorities’ exchange rate policy and fiscal stance were behind the initial run on Argentine banks that started in March 2001 and accelerated into a large scale withdrawal of deposits at end-November. This led the authorities to impose withdrawal restrictions (corralito) on December 1, followed by a bank holiday (December 21-26) in the context of the moratorium on Argentina’s external debt announced on December 23. On January 3, convertibility was abandoned, and another bank holiday was declared on January 7-8 as dollar deposits were forcibly pesoized (see Box 1 and 4) and the maturities of time deposits extended.
Deposit losses during 2001 were considerable, amounting to some 22 percent of end-2000 deposits. Following the external debt moratorium, deposits continued to decline in early 2002, reflecting withdrawals of sight and savings deposits and, since March 2002, judicial injunctions freeing time deposits. A new bank holiday was declared during April 22-26 to find a solution, but withdrawals continued in January-August 2002 and amounted to 25 percent of end-2001 deposits. While deposits were concentrated on a few large public and domestic private banks during 2001, deposit losses were generalized to all banks during the first half of 2002, with some flight from private to public banks. During the second half of the year, overall bank deposits stabilized and modestly expanded from the trough of July 2002. In late 2002, a gradual removal of deposit restrictions began. The monthly cash withdrawal limits on the corralito were relaxed in October and the corralito was fully lifted in December. In February 2003, several banks announced the early liberalization of frozen time deposits.
27. However, conventional crisis resolution approaches may not work in cases of sovereign default or where an unsustainable sovereign debt need to be restructured. Sterilizing large-scale liquidity injections (which may exacerbate capital flight and have adverse effects on inflation and the exchange rate) has typically involved the issuance of public debt instruments. Government bonds are also issued for bank recapitalization or in exchange for non-performing loans, or to central banks to compensate for contingent funding needs for deposit insurance. However, the use of domestic public debt instruments may not be effective when a sovereign has defaulted on its obligations, or when an unsustainable debt dynamics generates expectations of impending sovereign default.10 Under these circumstances, the authorities have to address the issue of how to apportion the losses in the banking system between bank shareholders, depositors, and current rather than future taxpayers. This is likely to be a contentious task.
28. Similarly, the extension of a blanket guarantee on deposits—which in a conventional systemic crisis may be necessary to restore confidence, prevent the risk of runs and contagion while weaker banks are intervened, and protect banks’ domestic sources of funding—may not be credible (or affordable) when the sovereign is unable to service its own debt.11 The sovereign’s ability to back these guarantees therefore may hinge critically on its access to external financing to provide needed liquidity and handle deposit withdrawals—particularly if dollarization is high.12 Section IV considers issues regarding the appropriateness of official financing, which may not be an option in cases where the banking system is sizeable. In any case, a credible medium-term fiscal program is ultimately needed to absorb the contingent fiscal costs of such guarantees.
29. Determined actions to put in place a credible framework for macroeconomic policies and structural reforms are necessary to restore confidence and halt pressure on deposits, reserves, or the exchange rate, but they may not always be sufficient. In instances where this strategy fails, the authorities could be confronted with the need to resort to administrative measures to avoid an uncontrollable run on banks. Such measures can serve the dual purpose of preventing a collapse of the banking system and limiting the scope for capital flight and exchange market pressures.13
30. Among the administrative measures that may have to be used to avoid the collapse of the banking system are:
A bank holiday. This has large costs in terms of loss of confidence, but can be effective as a temporary emergency measure (lasting a few days only) to take stock of the problems and provide the authorities some breathing space to formulate a comprehensive strategy for their resolution.
Securitization of deposits. While at least a minimum amount of sight deposits should be paid out in cash to preserve the payment system and minimize losses, securitizing deposits may be less damaging than deposit freezes or outright defaults, since a secondary market in such instruments could provide some liquidity.14 Nonetheless, in a situation of sovereign default, this option may require the use of commercial bank securities as a more credible instrument than government bonds, provided the bank is considered viable in the medium term. To mitigate the impact on financial intermediation, the payments system, and economic activity, securitization could be limited to time deposits (with no exceptions to avoid circumvention), and the securities used should be made transferable and issued in sufficiently small denominations.
Restrictions on deposit withdrawals. Extension of maturities of time deposits (“rescheduling” or “reprogramming”), or limitations on withdrawals of these deposits may be inevitable if deposits cannot be securitized. Deposit restrictions can greatly damage confidence in the banking system; they should be established only for defined time periods, and a strategy to lift the restrictions should be designed and implemented as quickly as possible.15
Generalized deposit freezes. These are very disruptive measures carrying high costs in terms of deepening the crisis, as they directly contribute to curtailing private consumption and investment. Even when deposit freezes are necessary, the strain placed on the operation of the payment system, especially if the freeze lasts for an extended period of time, will exacerbate a liquidity shortage and credit crunch, with adverse effects on the prospects for economic recovery.
Each of these measures imposes changes in the contractual terms between banks and depositors that, at least in principle, could be implemented at par and in the original currency. Nonetheless, these measures will generally entail large losses in real terms.16 In extreme cases, the emergency response to protect at least some parts of the banking system may involve a reduction to the face value of deposits, a compulsory redenomination of foreign currency bank deposits into domestic currency instruments or both. The extent and severity of the potential fallout of these measures depends on their specific characteristics, which should be considered with a view to minimizing the damage to the payment system, the impairment to the underlying claims, and the costs to the government.
31. These administrative measures should be considered with extreme caution and only where no other option is available to contain the crisis. The adverse effect on confidence and the severe real losses involved will substantially hinder the recovery of financial intermediation.
32. Beyond the immediate emergency, a comprehensive strategy for bank restructuring to restore financial sector soundness is critical to regaining confidence. Such strategy should be formulated, announced and implemented as soon as feasible, taking into account that systemic bank restructuring is a complex medium-term process that often strains the authorities’ implementation capacity. Nonetheless, failure to address adequately the restructuring needs of the banking sector may risk triggering a reemergence of the runs. A comprehensive strategy should include setting up appropriate institutional frameworks; intervening and closing nonviable insolvent institutions; strengthening weak but viable institutions through appropriate rehabilitation plans; dealing with value-impaired assets; improving prudential regulations and banking supervision; and promoting transparency in financial market operations.
E. Dealing with Capital Outflows
33. The announcement of a sovereign debt restructuring can trigger a general loss of confidence in the currency. This may result in large-scale capital flight and pressure on the foreign exchange market. In such an environment, conventional policy efforts may fail to restore confidence by themselves,17 and the imposition of exchange controls may be considered to support these efforts. The potential costs of exchange controls in crisis situations need to be weighed against the costs posed by a sharply depreciated exchange rate on the balance sheets of banks and corporates with unhedged foreign currency positions. Nonetheless, a certain degree of exchange rate adjustment and the unwinding of balance sheet overhangs is an important part of the adjustment process which would be postponed by keeping the exchange rate at an unsustainable level.
34. Exchange controls might stanch the hemorrhaging of reserves and provide some assurance to creditors at the negotiating table that the resources needed to service restructured debt obligations in the future were not fleeing the country. In addition, they can, at least for a time, help reduce pressures on the exchange rate, thus mitigating the substantial and widespread adverse effects of excessive depreciation, particularly if liabilities are widely dollarized. To have the desired effects, exchange controls need to be both enforceable and sufficiently comprehensive to prevent large-scale capital outflows and exchange rate depreciation during any initial panic.
35. That said, controls are likely to provide only temporary breathing space. The use of controls in capital account crises, including the few cases where it involved a sovereign debt restructuring (Box 3), has yielded mixed results at best. Experience has shown that controls cannot provide lasting protection, they come at significant costs, and they provide no substitute for addressing the fundamental causes of a crisis. Apart from their well-known negative impact on efficient resource allocation, they provide opportunities and incentives for corruption and may foster a culture of rent-seeking. Possible exceptions to controls and discretion in their administration creates opportunities for evasion.18
Exchange Controls in Sovereign Restructuring Cases
Countries that restructured their foreign debt, either after default or in the shadow of a default, imposed a wide range of controls in the attempt to deal with capital outflows, often in combination with deposit freezes. Some of the administrative restrictions discussed here are not exchange controls in the meaning of the Articles. The scope and sequence of measures depended on individual country circumstances.
Argentina (2001-02). In response to a large-scale bank run in late November, restrictions on cash withdrawal on sight deposits (corralito) and controls on capital outflows were announced on December 1, 2001 (see Box 2). Deposits restricted by the corralito could be used to purchase foreign exchange only for payments of imports of goods and services, profits and dividends, financial obligations (interest and principal) and other current account operations, particularly for trade-related operations. Although peso notes and coins outside the financial system could in principle be used to buy foreign exchange, prior authorization from the central bank was needed for most international transfers (debt service, profit remittances and dividends, purchase of foreign securities and other portfolio investment abroad, foreign exchange sales to nonresidents above certain limits).
Following the moratorium on external sovereign debt (December 23), time deposits were frozen (corralon) in January, while restrictions on sight deposits were relaxed. Other measures in support of the exchange control regime included surrender requirements on export proceeds and new foreign financing, the control of import financing (e.g. minimum maturity financing; restrictions on prepayments) and strict limitations on interbank currency trading. The controls began to be removed in late 2002. In December, the corralito was lifted and in January 2003, all but two exchange restrictions that are subject to the Fund’s jurisdiction under Article VIII were removed.
Ecuador. The country’s banking system had been in crisis for many months before the sovereign’s default in September 1999. In March 1999, a comprehensive deposit freeze was imposed. The deposit freeze restricted most banking funds from moving abroad, even in the absence of explicit capital controls. Nevertheless, substantial pressure on the exchange rate continued, with the sucre depreciating 200 percent during 1999.
Russia. In contrast to Ecuador, Russia intensified existing capital controls at the time it announced its default in August 1998, to prevent large outflows of capital and enforce the 90 day standstill on private (bank and nonbank) debt payments. The freeze on bank deposits was comparatively narrow in scope, aiming to deal with a few larger insolvent banks in the context of bank restructuring (see also Box 2). To support the provision of foreign exchange, the controls included export surrender requirements and advance deposits for import payments.
Ukraine. During its piecemeal restructuring of selected external debt , the government used exchange controls, without resorting to a deposit freeze. The existing regime of capital controls was tightened in August 1998, shortly before the first external debt restructuring. The controls included export surrender requirements, controls on import financing (screening of importers’ application for foreign exchange, limits on prepayments for import contracts, restrictions of foreign exchange loans to importers) and of interbank currency market transactions. The controls were removed gradually, with the last foreign exchange restrictions inconsistent with Article VIII being lifted in August 1999. In contrast to Russia, however, the controls never covered private external debt service, and there were no indications that the Ukrainian banks ran into arrears with foreign creditors as a result of the capital controls.
Pakistan. A default was avoided during Pakistan’s relatively small Eurobond restructuring. The freeze of the country’s large foreign currency deposits in mid-1998, largely owned by nonresidents, was a direct response to the external crisis after May 1998. Although the Eurobond exchange was not finalized until end-1999, existing controls were tightened in June and October 1999 during the debt restructuring negotiations. These included restrictions of certain capital account transactions (purchase of foreign debt securities, money market instruments, or real estate; extension of loans to nonresidents; prior approval for direct investment abroad), export surrender requirements, and the control of import financing.
The evidence on the ability of capital controls to preserve reserves and limit currency depreciation is mixed—and the confluence of other factors at play during a crisis complicates an assessment of their effectiveness. For example, Russia’s adoption of comprehensive controls in August 1998 failed to limit pressure on international reserves and the exchange rate continued to depreciate until early 1999. Russian residents had been moving funds abroad prior to August 1998 despite controls. The default would have likely accelerated such outflows in the absence of controls, and therefore it is difficult to determine the proper counterfactual, i.e. whether at the margin the controls helped limit the scale of capital flight. In Argentina large deposits losses occurred between January and August 2002, despite comprehensive restrictions on withdrawals (see Box 2). Even larger losses might have occurred in the absence of the deposit freeze, however. Similarly, evidence on the impact of controls on the country’s ability to attract new capital inflows is not conclusive. The imposition of exchange controls, particularly controls that impair the private sector’s ability to service its external debt, could be expected to undermine foreign investor willingness to invest for fear of restrictions in the repatriation of future profits. However, a debt crisis alone—including its associated spillovers into the private sector—would tend to deter future foreign investment in the absence of controls, making it difficult to isolate the impact of controls.
36. In the environment of a sovereign debt crisis, the authorities may have to address two principal sources of outflows. First, both residents and nonresidents may want to suddenly substitute their domestic assets (including their bank deposits) with holdings of foreign assets. Second, capital outflows may result from the private sector’s foreign debt payment obligations. In this respect, it may be important to try to limit their distortive effects on the economy, in particular the adverse longer-term consequences for the country’s corporate sector (Box 4).
37. In deciding on the adoption of exchange controls, three issues regarding implementation should be carefully considered. First, the timing. The desirability of an early imposition of controls, and the benefits of preserving value and addressing inter-creditor equity concerns by preventing short-term investors from exiting, should be weighed carefully against the costs of interfering with investors’ rights and impairing private sector’s future access to international capital markets, even for companies that may have had some ability for external borrowing. However, if a determination is made that exchange controls are necessary, they should be implemented as soon as possible as the potential benefits in terms of keeping capital in the country tend to dissipate over time.
38. Second, the scope of controls. Exchange controls may be designed to restrict: (i) the use of current account proceeds, by imposing the repatriation or surrender requirement on payments for exports of goods and services;19 (ii) residents’ outward investments, by reducing the scope for investing abroad and requiring residents to invest their funds domestically; (iii) nonresidents’ repatriation of capital, to avoid the collapse of domestic asset prices, including the currency; (iv) nonresidents’ access to domestic capital markets, thus making it more difficult for them to speculate offshore; (v) where unavoidable, suspend transfers of amortization payments on private foreign debt, and as a last resort, interest payments.
Spillovers into the Corporate Sector
Non-financial firms typically are not major direct creditors of a sovereign. However, a number of channels for spillovers discussed in this paper can trigger widespread financial distress in the corporate sector. These include:
Exchange rate depreciation. Sharp real depreciation can dramatically increase the real burden of servicing foreign currency denominated debt and undermine firms’ net worth. An interest rate defense will also affect firms with short-duration local currency debt.
Limited external market access. Macroeconomic instability can create difficulties accessing capital markets and rolling over short-term debts.
Domestic credit constraints. Banking weaknesses may create a domestic credit crunch, exacerbating the impact of reduced access to external credit.
The magnitude of these spillovers has varied and has been, not surprisingly, more limited in countries with underdeveloped domestic financial systems and limited external market access for corporate borrowers. For instance, Russia’s banking system was not a major source of credit to Russia’s private firms (domestic credit was less than 10% of GDP), and relatively few firms had access to international finance (the external debt of banks and firms was also under 10% of GDP, concentrated in the oil industry). As a result, Russia’s sovereign default—and the associated banking crisis—had a comparatively limited impact on the corporate sector. Some external corporate debt was rescheduled informally, but no systemic framework was put into place.
Ecuador experienced a much more severe corporate crisis, but it is not clear whether the causality went from the sovereign to the corporate sector or the other way around. A rapid expansion of domestic credit to the private sector prior to 1999 contributed to the banking crisis that preceded the sovereign default. External credit to the private sector was small in relation to public sector external debt, but larger than in Russia (16% of GDP in 1998). Domestic credit to the private sector was in excess of 20% of GDP. Sixty percent of these loans were non-performing by 1999. As part of the general financial sector restructuring that followed dollarization, all loans under US$50,000 were automatically restructured. Larger loans were restructured on a case-by-case basis.
Argentina’s sovereign crisis spilled over into the corporate sector. Firms had borrowed extensively from the local banking system (domestic credit was in excess of 20% of GDP), and the private sector’s external debt also exceeded 20% of GDP. Because of the relatively small scale of Argentina’s exports (slightly over 9% of GDP in 2000) and extensive liability dollarization, many firms without export earnings had foreign currency denominated debts. Consequently, the sharp real depreciation that followed the end of the currency board presented a significant threat to the solvency of many firms. Argentina “pesoized” the domestic debts of Argentine firms in order to try to limit these pressures.
The combination of increased real debt burdens and difficulties rolling over short-term debts have often called for some form of corporate debt restructuring after a sovereign default. In general, the approaches followed depend on established domestic systems of creditor rights and bankruptcy legislation to provide the basic framework for restructuring corporations’ external and internal debt. In circumstances of widespread financial distress, a process for encouraging out-of-court agreements in the “shadow” of an adequate formal insolvency system can be more efficient than allowing a large number of firms to enter bankruptcy procedures at the same time. These frameworks typically focus on replicating “out of court” the basic features of bankruptcy regimes— creditors agree to refrain from litigation and provide new financing while restructuring terms are negotiated.
39. To minimize the distortions and disruption of commercial activities associated with the introduction of controls, some exemptions could be considered, including on (i) current account, especially trade-related, transactions; (ii) direct investment flows; (iii) debt service associated with foreign debt contracted after the imposition of controls; (iv) private sector payments on foreign debt to help enterprises retain some access to new foreign credit, including for trade, and avoid legal actions by remaining current on their debt service.
40. Third, the enforceability of controls. In general, the necessary capacity to enforce controls is likely to be better developed in countries with a tradition of exchange controls (where existing controls may only have to be tightened) than in countries that have had liberal financial markets for some time and no longer have the necessary infrastructure in place. Successful enforcement demands a comprehensive information and disclosure system between the exchange control authority and the commercial banks. Additional restrictions may have to be introduced to control circumventions. For example, requiring trade transactions to be conducted on the basis of letters of credit makes it easier to match payments and receipts against shipping and customs declarations, thereby preventing the over- and under-invoicing of trade; and, similarly, requiring documentation of external borrowing could limit the scope for it to serve as a vehicle for capital flight. Regulations may have to be designed to limit residents’ and non residents’ ability to modify leads and lags (in the timing of payments) on the trade account, which can exert non negligible pressures on foreign reserves. All these requirements manifest the distortions that controls impose to the economy.
41. Transparency and a strict enforcement of the controls is crucial for their effectiveness. Existing administrative capacity to apply controls and the regulator’s reputation for strict enforcement can play an important role in containing circumvention efforts. Nonetheless, the more sophisticated a country’s financial market has become, the more opportunities for arbitrage and circumvention it provides, and the more difficult it is to monitor all these transactions closely.
42. The announcement of controls ideally should be accompanied by a time frame for their withdrawal. Controls only serve as a transitory remedy, while the debt restructuring is underway and sound macroeconomic policies are being implemented. To be perceived as such by the public, the controls would have to be presented as an element of the overall restructuring effort with a firm commitment (possibly in the context of a Fund arrangement) to lift them as this process is brought to completion. Otherwise, even the most targeted controls would create significant distortions over time. If extended beyond the period needed to complete the restructuring, the controls could rapidly lose their effectiveness, and may exacerbate rather than reduce the disruptions in the economy.
43. While members are prohibited from imposing restrictions on the making of payments and transfers for current international transactions without Fund approval, Article VI, Section 3 recognizes the right of members to impose capital controls when necessary to regulate international capital movements. 20 The scope of the right of members under Article VI, Section 3 has been interpreted broadly, and includes the authority to permit, prohibit, or limit inward and outward capital movements. The Fund has treated as “capital” any payment or transaction that does not fall within the definition of “payments for current transactions” set out in Article XXX(d). On Fund support for exchange restrictions under the Articles of Agreement, according to Art. VIII, Section 2(a), a member which intends to impose measures that give rise to restrictions on the making of “payments of transfers of current international transactions” (defined in Art. XXX (d)) has to seek the Fund’s prior approval.21 Measures that preclude residents from making payments and transfers for current international transactions, such as interest or moderate amounts of amortization on loans, or that restrict nonresidents from making transfers of balances acquired in recent current international transactions, have been found to give rise to such restrictions. The Fund is prepared to give temporary approval to such restrictions, when applied for balance of payments reasons, provided they are necessary, temporary and nondiscriminatory.22
III. The Overall Coordinating Framework for Sovereign Debt Restructuring
44. A key issue that policy makers need to address in sovereign default cases concerns how to set up a framework for coordinating the restructuring of various forms of debt. In conjunction with determining the appropriate scale and path of fiscal and external adjustment, a framework for coordinating the overall restructuring of the sovereign’s debt needs to be developed so that the restructuring agreements with different groups of creditors combine to produce a payments profile consistent with the adjustment path and a sustainable debt stock. This section discusses these coordination issues.
45. A sovereign that needs to seek a debt restructuring will often have many different kinds of debts outstanding. Some debt will be governed by external law, some by domestic law. Some external debt will be owed to private creditors, some to official bilateral creditors and some to multilateral development banks and the Fund. And currency of denomination may differ. The sovereign will need to decide which external law debts to restructure and the extent to which it will be necessary to restructure other categories of claims, including domestic debt. The sovereign will also need to assess the advantages of a targeted debt restructuring focused on certain categories of claims against those of a more comprehensive restructuring.
46. Although there may be economic reasons for attempting a more targeted restructuring, achieving a satisfactory debt-service profile is likely in many cases to require a comprehensive approach both in order to bring the overall debt-service level and profile to a sustainable position, and to achieve sufficient inter-creditor equity to garner support for the restructuring proposals. Clearly, the benefits of a comprehensive debt restructuring need to be weighed carefully against considerations about the viability of the domestic financial system, balance sheet effects, and the implications for economic activity. In some circumstances this may require a two-stage process, where initially domestic banks would at least receive interest pending the restructuring, while other creditors might be subject to a more comprehensive cessation of payments. Nonetheless, achieving sustainability is the paramount objective, and a comprehensive approach, by spreading the costs of restructuring and improving inter-creditor equity, may improve the chances of reaching agreement on a restructuring on terms that are consistent with a return to medium-term viability. Each creditor group whose debt is restructured may predicate any agreement on assurances that other creditors will face similar debt reductions. Indeed, achieving a given reduction from one group of creditors could be futile if other creditor groups subsequently negotiate more favorable terms—possibly necessitating reopening negotiations if the overall terms of the various restructuring agreements are inconsistent with sustainability. This highlights the desirability to negotiate simultaneously on multiple tracks in a collective framework.
47. There is at present no formal system for coordinating a comprehensive debt restructuring for sovereign debtors. Official bilateral debt is generally restructured in the context of the Paris Club, based on an Agreed Minute, although claims of non-Paris Club countries are restructured separately. 23 Sovereign external debt held by private creditors, particularly commercial banks can be restructured in the context of the London Club (as was the case in Russia’s restructuring - see Box 5). In other cases, no formal creditor committee is formed, or negotiations can take place with different groups of creditors or without any negotiations at all (i.e., through unilateral debt exchange efforts on an instrument by instrument basis). Domestic debt is restructured according to the laws of the sovereign debtor. Against this setting, there is no comprehensive framework in place now that can help guide the coordination of various elements of a restructuring process, achieve predictability, and give some assurance of inter-creditor equity, or sufficient progress toward achieving debt sustainability.24 25
Coordination Issues in Sovereign Debt Restructuring
The countries discussed in Box 1 followed different approaches in the restructuring of their debt.
Russia differentiated between various groups of creditors, and the ultimate restructuring terms for domestic debt (46 different issues of GKOs/OFZs), Paris Club debt and London Club were quite different. Eurobonds were not restructured. The ability of Russia to differentiate among private holders was driven mainly by its strong legal hand: the obligor for a large part of the London Club debt was Vnesheconombank rather than the sovereign; a substantial part of the domestic (GKO) debt could be restructured under local law; and creditors were perceived to have limited recourse to successful litigation strategies. An important coordinating mechanism for restructuring the external privately held debt was the London Club’s Bank Advisory Committee, which had previously restructured the Soviet era debt. As a substantial portion of this debt had been securitized and sold in the secondary market, a separate group was organized—the London Portfolio Managers Association—to represent the interests of institutional investors and to liaise with the Bank Advisory Committee. There was a tension between London Club and Paris Club creditors when the latter refused to accept a restructuring on similar terms.
Ecuador’s dialogue with the bondholder community was limited in the eleven months between the initial default and the announcement of a comprehensive debt exchange in late July 2000. This reflected the authorities’ concern about losing leverage in the negotiations, the diversity of the creditor base (various types of institutional investors, commercial banks, insurance companies and hedge funds, but relatively few retail investors), and problems in forming a representative creditor committee. Relations with foreign creditors were further strained by what was seen as an unfairly advantageous restructuring of domestic debt. The authorities did establish a so-called Consultative Group, consisting of eight institutional holders with large exposure, but only two meetings were held. No meetings were held with minority bondholders. In May 2000, however, following the approval of the Fund arrangement, an open meeting was held with bondholders to discuss Ecuador’s dollarization program, economic recovery measures, and program financing issues. Projections presented at this meeting were disseminated to the wider investor community through the website of the Emerging Markets Traders Association. Ultimately, the restructuring succeeded in achieving 97 percent bondholder participation through the innovative use of exit consents. Nonetheless, Ecuador was forced to settle arrears on holdout bonds, in order to shelter itself from litigation. As in the case of Russia, disagreements between private creditors and the Paris Club over the lack of reverse comparability of treatment emerged after the debt exchange.
Pakistan’s experience reversed the sequence followed in Russia and Ecuador in that a Paris Club rescheduling, requiring a restructuring of private debt on comparable terms, preceded an exchange offer for Eurobonds. As in Ecuador, however, Pakistan avoided a formal meeting with bondholders. While the bonds included collective action clauses, the authorities feared that the required majority restructuring threshold might not be achieved, and that a meeting could help creditors coordinate disruptive legal strategies. The authorities were able to contact key creditors and conduct informal dialogues concerning the bond restructuring proposal. The restructuring was facilitated by the relative homogeneity of the investor base, mainly financial institutions and retail investors in the Middle East. In the end, the exchange enjoyed widespread participation with over 99 percent of all bonds tendered.
Ukraine’s bonds were held by a limited number of investment banks and hedge funds. As in Pakistan, the authorities engaged in informal discussions with investors concerning the terms of restructuring. However, one bond was held by a wide base of retail investors in Asia and Europe, complicating the process of consultation. The authorities retained four investment banks to identify and market the exchange to retail investors. In contrast to the Pakistan bond exchange, Ukraine employed a novel use of the collective action clauses embedded in three bonds governed by Luxembourg law. By requiring a qualified majority of proxy votes before calling a formal meeting, Ukraine ensured that the proposed new payment terms would be accepted. Consistent with the need to smooth out a liquidity hump more than reducing the debt stock, the new bonds in the April 2000 exchange were offered on attractive terms: no debt forgiveness or reduction in principal, no interest-free grace period, amortization starting in 2001 and cash payment of all accrued but unpaid interest. Ukraine was briefly technically in default while the exchange offer was open. Bondholders, however, chose not to accelerate their claims or initiate legal proceedings, and Ukraine managed to achieve a very high level of participation - 99 percent of old bonds were tendered in the exchange.
Argentina’s progress in restructuring its foreign held debt following the interruption of the two-phase approach in December 2001 has been limited. More recently, however, the authorities have met with institutional creditors, as well as with some groups of retail bondholders. Efforts to lay out a broader debt restructuring strategy have been complicated by the significant uncertainty regarding Argentina’s macroeconomic strategy and future payment capacity. In addition, delays in advancing the dialogue with creditors have led to litigation. To date, around thirty lawsuits have been initiated, including some class action suits in New York. In February 2003, the authorities appointed an external advisor to assist in managing relations with creditors and in preparations for debt restructuring negotiations.
48. To the extent that countries in need of debt restructuring seek a Fund arrangement, the macroeconomic framework underlying the arrangement can help anchor deliberations between the debtor and its creditors. A Fund-supported program typically provides a signal to creditors and the international community that the authorities have embarked on a coherent strategy to deal with the crisis and a framework for monitoring the implementation of the needed policy adjustment measures. The magnitude of the needed debt reduction is the difference between the existing debt level (taking into account the aforementioned valuation issues) and the sovereign’s long-term sustainable debt level. The macroeconomic framework will inform some of the parameters and the underlying policies as discussed in Section II.C above. In all cases, the baseline fiscal path will involve balancing the restructuring terms that creditors can accept and the adjustment that the sovereign can realistically deliver without undermining prospects for future growth.
49. Restoring sustainability is generally likely to require efforts to coordinate the actions of different creditor groups. While creditors collectively have an interest in a restructuring that results in the resumption of debt service and a viable debt-service path— indeed, the NPV of debt-service payments following a debt restructuring can be higher as the debt overhang is eliminated, real interest rates fall and growth prospects are improved26— each creditor would prefer a restructuring that minimizes its own financial concessions. Early consultation with a broad range of creditors can help the sovereign understand the different interests of creditors, the types of restructurings most likely to attract broad support, and whether or not some instruments can be excluded from the restructuring. As it proceeds with the restructuring, it is important for the sovereign to share all relevant information and to explain how it proposes to treat different classes of creditors, the adjustment effort that it plans to undertake, and how the concessions it seeks from different groups of creditors will combine to achieve the debt-service relief that is ultimately needed. 27
50. The Fund’s policy regarding the “good faith criterion” in its lending into sovereign arrears to private creditors has special relevance for the modalities of the sovereign’s dialogue with its creditors, and is an important step toward the development of a code of conduct concerning the renegotiation of sovereign debt. The policy is flexible with regard to the form of the collaborative dialogue. Dialogue and coordination can take place within an informal structure or through a formal representative committee, depending on the complexity of the individual case and on the timely organization of a sufficiently representative committee.
51. The Fund’s assessment of a member’s good faith efforts is guided by the following general principles:
The debtor should engage in early dialogue, which should continue until the restructuring is complete. In cases where a representative committee of creditors has been established, a collective framework for negotiations should be established.
The member should share relevant, non-confidential information with all creditors or a steering committee on a timely basis. This information would include an explanation of the adjustment program and the financial circumstances that justify a restructuring, as well as a comprehensive picture of all domestic and external claims, including official claims, and the broad process by which restructuring would restore medium-term viability. Creditor committees can also play a useful role in assessing confidential information, provided that confidentiality can be preserved.
The member should provide creditors with an early opportunity to give input on the design of the restructuring strategy. This could help address the specific needs of different types of investors, thereby increasing the likelihood of a high participation rate.
52. Fund support of a program that lays out the overall payments profile consistent with the country’s monetary and fiscal program can therefore help to coordinate the restructuring of the claims of different groups of creditors. It allows each creditor group that is participating in the restructuring to assess the overall payments envelope and the proposed treatment of different classes of claims. A Fund program also provides a framework for monitoring whether the sovereign is taking the steps to implement the needed policy adjustment measures. In addition, financing reviews under the lending into arrears policy provide a framework to check that sufficient progress is made toward restoring sustainability.
IV. Role of Fund Financing
53. While a program supported by the Fund may provide a framework for the resolution of the crisis, an important question concerns the appropriate use of, and scale of access to, Fund resources in sovereign debt restructuring cases. In order to ensure the revolving character of the Fund’s resources, the Fund can only lend if it has adequate assurances that the member’s balance of payments position will be sufficiently strong to allow repurchase. This precludes Fund financing in cases where the member’s debt situation is unsustainable, unless this situation is addressed.28 For a member with an already precarious debt burden, a Fund decision to withhold financial assistance in the absence of a needed debt restructuring may lead the member to seek such a debt restructuring, possibly under the provisions of the SDRM (Box 6).
54. In those cases where a program supported by the Fund is in place, the goal should be to resolve the balance of payment problem of the member and achieve medium-term external sustainability while fostering sustainable economic growth. This means supporting a restructuring of the sovereign’s overall debt that restores a viable balance of payments position, allows for the lifting of any exchange controls as soon as possible, and establishes a sustainable debt situation in which prolonged use of Fund resources is not contemplated and repayment of existing Fund exposure is assured.
55. Under the Articles, Fund financing can only be used to meet a country’s balance of payments need. A member can have a balance of payments need “because of its balance of payments or its reserve position or developments in its reserves” (Article V, Section 3 (b) (ii)).29 A Fund arrangement provides general balance of payments support, which will typically imply a replenishment of the central bank’s reserves. This augmentation in gross reserves can serve to enhance the sovereign’s foreign exchange coverage over short-term claims, or be used for exchange rate intervention within the confines of the program’s parameters.
56. The financing needs of a sovereign undertaking debt restructuring may be severe. While a standstill on payments is likely to provide a sovereign with some limited breathing space, this could well be offset by a deterioration in the private capital account, putting pressure on already very low levels of reserves (Russia, for example, lost 40 percent of its reserves in the month prior to default). Some of the key needs for external support that could be provided by IFIs are the following:
The Role of the Fund in the Debt Restructuring Process
Fund policies and decisions are guided by the provisions of its Articles. Fund resources are to be made temporarily available to members under adequate safeguards to help them resolve problems in their balance of payments. These safeguards require conditioning the availability of Fund resources on the adoption of appropriate adjustment policies and preclude Fund financing in situations where the member’s debt is unsustainable, unless the situation is addressed. Inevitably, Fund decisions play a role in a number of dimensions of the restructuring process:
The Fund’s judgment about debt sustainability and the availability and scale of financing it is willing to provide in the absence of a debt restructuring could influence a member’s decision as to whether to seek a debt restructuring and, possibly, suspend payments to conserve its international reserves.
While the member is negotiating a restructuring with its creditors, it will normally be working with the Fund on the development of an appropriate policy framework. To merit Fund support, a program should be economically coherent, properly financed, and consistent with a return to balance of payments viability and debt sustainability. Such a framework can provide an anchor both for the formulation and implementation of policies and for debt restructuring negotiations.
An important element of a Fund-supported program will be a fiscal and external adjustment path consistent with a return to debt sustainability. A program will help establish an envelope of resources available for debt service by the sovereign and, in this respect, help coordinate the restructurings of the claims of different creditor groups as noted in section III.
Fund support for a member that is in arrears on its obligations to private external creditors would be guided by the lending into arrears policy, including the Fund’s assessment of the member’s good faith efforts to reach a collaborative agreement with its creditors. Progress toward a restructuring agreement will be monitored in programs by financing reviews.
The Fund would assess the consistency of the restructuring agreement with the adjustment path in the member’s economic program and whether the resulting medium-term payments profile is consistent with a return to sustainability. These judgments would have a bearing on the Fund’s decisions regarding further financial support to the member and may therefore need to be taken into account by the member and its creditors in reaching agreement.
Trade Financing. Debt restructuring could lead to an asymmetric deterioration in access to trade credit. Exporters, in principle, should benefit from any depreciation in the exchange rate, enhancing their profitability—although this may be offset to the extent they rely on imported inputs, or their financing is affected by a disruption to the domestic banking system. For importers, the situation is likely to be more difficult, as domestic banks lose access to external credit lines, and external financing dries up due to perceived increased credit risk arising from the uncertain environment. Fund financing may have a role to play in helping finance needed imports of goods and services and prevent a collapse in output.
Containment of balance sheet effects. In an environment of high uncertainty about the future value of domestic currency assets, monetary policy alone may not be effective in stemming capital outflows. Interest rates that might attract foreign inflows by compensating for the risk of further depreciation or confiscation could be extremely high, with negative effects on balance sheets and real activity. IFI financing may help prevent a sharp overshooting of the exchange rate and contain pressures on sectoral balance sheets.
Targeted support for the banking sector. External liquidity could enable domestic monetary authorities to operate a dollar discount window for banks (e.g. Korea ‘97), enabling private banks to stay current on their external obligations. Also, to the extent that the source of capital flight is founded in concerns about the credibility of a guarantee backing, for instance, foreign currency denominated deposits, the provision of external liquidity could allay these concerns.
Minimizing the collapse of investment. In an environment where investment is already depressed (due to a deterioration in balance sheets and a loss of access to foreign credit), IFI financing directly or indirectly to the government could mitigate crowding out the private sector in an environment where sources of investment financing are extremely scarce. In cases where the balance of payments need is exacerbated by large fiscal imbalance, such financing may also reduce the need for central bank financing in the absence of other budget financing sources.
Maintaining access to priority financing. Limited reserves may be used to service claims of creditors whose continued support is essential for economic stabilization. To achieve a more viable debt-service profile the sovereign may seek access to the Fund or other IFI lending facilities with longer maturities, as appropriate.
57. There are limits, however, to what the Fund can and should finance in a debt restructuring context. Fund resources should not be used to comprehensively underwrite liabilities of the financial or corporate sector, or to finance payments to private creditors on debt subject to restructuring except in a limited, temporary context. Fund or other IFI lending should be considerate of the need not to increase excessively the rigidity of the country’s debt stock (by providing preferred credit) to an extent that future repayment capacity is endangered. Moreover, there may be cases where foreign support is clearly insufficient and undesirable to back up a guarantee.30 Also, while limited provision of external liquidity might help overcome a dollar shortage and nip a panic in the bud, sustained large-scale interventions to accommodate large capital outflows may be ineffective. Indeed, the Fund is prohibited from financing large or sustained outflows of capital (Article VI, Section 1).
58. Judging the appropriate access level in debt restructuring cases will pose particular challenges. These are related to the different nature of the balance of payments need described above, and the high degree of uncertainty surrounding macroeconomic variables. In addition, the basic incentive problem described in section II.C - i.e. the trade-off between adjustment and debt reduction - also applies to the consideration of new financing, to avoid supporting insufficient policy adjustment or excessive debt-service payments. These highly uncertain circumstances would generally imply limited financing within normal access limits to help safeguard Fund resources, given the precarious debt situation until the process of restructuring is completed and the limited prospects for regaining access to private market financing. This notwithstanding, there may be cases that warrant exceptional access.31 Such circumstances could arise in those cases where the member is not in a position to make significant net repurchases to the Fund or where economic activity is excessively and directly hampered by a shortage of foreign exchange. In these rare cases, the justification of exceptional access in the context of a debt restructuring should be particularly strong. The need for exceptional financing to support a strategy for limiting economic disruption associated with debt reduction has to be balanced against the need to avoid adding to the debt burden or adversely affecting the debt composition of an already over-indebted sovereign.
59. During the period the member is negotiating a restructuring with its creditors, the conditions of the Fund’s policy on lending into arrears will apply, whether or not the member has invoked the SDRM. In addition to an assessment of the good faith criterion mentioned above, the Fund could continue to make disbursements dependent on financing reviews. These reviews are a way of ensuring that the negotiations with creditors converge on the objective of reestablishing sustainability. Should such objective prove elusive, the adjustment parameters of the program would need to be reconsidered to avoid an interruption in Fund financing.
V. Issues for discussion
60. This paper provides a preliminary overview of the complex issues arising in the resolution of crises that require a sovereign debt restructuring. Staff will continue its work on these aspects in parallel to developing the operational design of the SDRM. In particular, further work will focus on the best approaches to the resolution of systemic banking problems, and the appropriate use of exchange controls to supplement and facilitate restructuring.
The paper discusses issues involved in determining a reasonable adjustment path following a debt restructuring decision and in the midst of a highly uncertain economic environment. Directors may wish to comment on the overall framework in determining this path and the appropriate scope (comprehensive vs. targeted) for sovereign debt restructuring to ensure a return to medium-term sustainability.
The paper outlines the challenges in protecting the banking system in restructuring cases where domestic banks have significant holdings of claims on the sovereign. Do Directors share the assessment on the limited effectiveness of conventional crisis resolution measures in these cases? Directors may wish to comment on measures that could be considered to forestall deposit runs.
Recognizing that administrative measures are disruptive, how do Directors see the balance of costs and benefits in using them as a last resort and in the context of a well-defined time frame? Are there other strategies in addition to the ones discussed in the paper that Directors consider appropriate?
The paper discusses circumstances under which the temporary imposition of exchange controls may reduce the scope for capital flight in the context of a sovereign debt restructuring. Directors may wish to comment on the appropriate design and use of controls in these cases.
Directors may wish to comment on the role of a Fund-supported program under the lending into arrears policy as a coordinating framework in cases where claims of different groups of creditors need to be restructured.
In sovereign debt restructuring situations, financing constraints are particularly acute. Directors may wish to comment on the appropriate access to, and use of, Fund resources in these circumstances.
“The Design of the SDRM—Further Considerations,” EBS/02/201, November 27, 2002.
This paper does not attempt to address conceptual considerations regarding the assessment of sustainability, or the incentives that surround the decision by a member to undertake a debt restructuring.
See “Sovereign Debt Restructuring and the Domestic Economy: Experience in Four Recent Cases,” SM/02/67, February 21, 2002.
where “d” is the public debt to GDP ratio, “p” is the primary fiscal balance to GDP ratio, “λ” is the growth rate of the economy, and “r” is the real interest rate paid on public debt. In general, a higher debt burden can be supported: (i) the higher the rate of economic growth; (ii) the lower the real interest rate; and (iii) the higher the primary fiscal balance.
One means of assessing them would be to look at the borrowing costs of other countries with broadly similar macro-economic characteristics (public and external debt post-restructuring, fiscal and external imbalances etc). In addition, the results of studies on market access by emerging markets can be used to gauge re-entry rates for the sovereign post-restructuring. See “Assessing the Determinants and Prospects for the Pace of Market Access by Countries Emerging from Crises,” EBS/01/157, September 6, 2001.
The costs of recapitalization associated with the banking crises in recent restructuring cases can be large—public debt levels have risen in some cases by 15 percent of GDP or more. See “A Framework for Managing Systemic Banking Crises” (forthcoming Board paper).
This section deals with a set of complex issues where the work of MAE has played a prominent role in advancing the staff’s thinking. References include: Financial Sector Crisis and Restructuring: Lessons from Asia, IMF Occasional Paper No. 188; Building Strong Banks Through Surveillance and Resolution, IMF 2002, and MAE’s forthcoming Board paper on “A Framework for Managing Systemic Banking Crises”.
See “Involving the Private Sector in the Resolution of Financial Crises—The Restructuring of International Sovereign Bonds—Further Considerations,” EBS/02/15, January 31, 2002.
A core of the banking sector could be gauged on the basis of considerations related to the branch network, the value of transactions processed through the payments system, the number of depositors, and the balance sheet conditions and prospects for the banks involved. In practice, this could involve difficult judgments given the high uncertainty on the conditions of banks associated with a sovereign debt crisis.
The implications of the fiscal costs of bank restructuring for debt dynamics and debt sustainability are of particular concern in financial markets where the volume of deposits and the size of the banking sector are large in relation to GDP.
See “Deposit Insurance: Actual and Good Practices,” IMF Occasional Paper No. 197, 2000.
In a highly dollarized economy, the central bank’s role as a lender of last resort is constrained by the size of international reserves relative to deposits in the banking system. A limited coverage may undermine the credibility of a strategy for liquidity support. External financing of the deposit guarantee may be necessary to restore the credibility needed to stem dollar bank runs.
Administrative bank measures may be a more effective means of stemming capital flight in cases where the enforceability of exchange restrictions is weak.
Redemption of bonds on the secondary market may take place at a (heavy) discount.
The mobility of transaction accounts’ within the banking system and the possibility of limited withdrawals should be considered to minimize payment system disruptions. Depositors may also be allowed to use restricted deposits as means of payment within the banking system. This option, while allowing payments, may create additional pressures on the system as depositors move restricted deposits to the strongest banks.
The adoption of administrative measures affecting creditor rights may give also rise to legal or constitutional challenges that can result in increased uncertainty.
Particularly where there is limited room to defend the exchange rate with interest rate increases, due to concerns about the adverse implications on fragile balance sheets, including, those of banks, where a maturity mismatch toward short-term liabilities is typical.
For an overview, see “Capital Controls: Country Experience with their Use and Liberalization,” IMF Occasional Paper No. 190.
This measure is not an exchange restriction under the Fund’s Articles, as it does not constitute a restriction on the making of payments and transfers for current international transactions. It could, however, limit the scope for residents to undertake capital transactions.
This paper does not discuss the broader issue relating to the vulnerability of domestic enterprises to litigation that may arise as a result of the default that occurred following the imposition of capital controls.
The concept of “restrictions” is not defined in the Articles, but the Fund’s guiding principle to determine a restriction is the involvement of a direct governmental limitation on the availability or use of exchange as such, Decision No. 1034-(60/27) June 1, 1960.
Ibid, see also Decision No. 955—(59/45), October 23, 1959.
It should be noted that Paris Club restructurings are not aimed a priori at achieving a sustainable debt profile but, rather, at providing a country with cash flow relief during the “consolidation period” (i.e. on claims falling due during the period of the Fund-supported program).
Sometimes a Paris Club agreement precedes an agreement with private creditors, e.g. in the case of Pakistan, and sometimes it does not, as in the case of Ecuador.
The SDRM would establish a framework for the restructuring of the claims of private creditors governed by external law, and—depending upon its coverage—possibly also the claims of official bilateral creditors. In addition, there are a number of features of the proposed mechanism that could help facilitate the coordination of the restructuring of debts that would be restructured outside the mechanism.
In these circumstances, an orderly and prompt restructuring can create value for both creditors and debtors.
See “Fund Policy on Lending into Arrears to Private Creditors—Further Consideration of the Good Faith Criterion,” (SM/02/248, July 31, 2002).
See “Access Policy in Capital Account Crises—Modifications to the Supplementary Reserve Facility and Follow-Up Issues Related to Exceptional Access Policy,” (EBS/03/20, January 14, 2003) and “Access Policy in Capital Account Crises,” (SM/02/246, July 30, 2002).
See “Need as a Condition for the Use of Fund Resources,” (SM/94/299) for a more detailed discussion of the legal and economic considerations surrounding the concept of “need”.
In some emerging markets, for instance, deposits to GDP stand in the range of 50-100 percent of GDP, well in excess of available IFI financing. By comparison, the average ratio of (annualized) Fund financing to GDP between 1993-2001 was 1.4 percent of GDP for arrangements within the access limits and 3.3 percent of GDP for exceptional access arrangements. The recent arrangement for Uruguay lies at the extreme end of the range with annualized access to GDP (as projected at the time of approval) at 10.5 percent of GDP.
Circumstances that might warrant exceptional access to Fund resources were discussed by the Board in the context of the paper on access policy in capital account crises, BUFF/02/159, September 20, 2002.