This paper provides an updated review of Fund-supported programs undertaken during the global financial crisis. It follows a series of previous reviews during 2009–12 that assessed program design and outcomes during the surge in Fund supported programs since 2008.
The review covers experience during 2008–15 for 32 arrangements financed from the Fund's general resources account (GRA). It covers 27 countries for which arrangements were approved during September 2008–June 2013, with two years or more of program performance.
This paper provides an updated review of Fund-supported programs undertaken during the global financial crisis. It follows a series of previous reviews during 2009–12 that assessed program design and outcomes during the surge in Fund supported programs since 2008.
The review covers experience during 2008–15 for 32 arrangements financed from the Fund's general resources account (GRA). It covers 27 countries for which arrangements were approved during September 2008–June 2013, with two years or more of program performance.
Fund-supported programs helped chart a path through the global financial crisis that avoided the counterfactual scenario many initially feared, involving a cataclysmic meltdown of the global economic system. Given the radical differences between the 2008 crisis and its predecessors, decisions were made amidst significant uncertainty about shocks, transmission channels, and policy responses. Program outcomes helped inform the design of later programs, and contributed to broadening the array of feasible policies over time by strengthening frameworks and reducing the risk of contagion.
Nominal exchange rate adjustment was less central to the adjustment strategy than in previous program episodes. “Internal devaluation”, which relies on domestic price adjustment, proved hard to achieve within a short period, owing in part to domestic rigidities as well as partner countries’ weak growth and low inflation. Vulnerabilities accumulated ahead of the crisis remain to be fully resolved in many cases.
Regional financing played an important role in Euro Area programs. Given the welcome and growing importance of regional financing arrangements (RFAs), clearer operational guidance for the Fund’s interaction with RFAs in the context of Fund-supported arrangements would be helpful for delineating responsibilities.
Context. The Fund approved support under the GRA facilities and instruments of SDR 420 billion during 2008–13, including both precautionary arrangements and arrangements from which members made drawings, of which nearly SDR 119 billion of these resources was drawn during the period. Fund-supported programs responded to differing needs among members, including emerging markets affected by capital flow reversals and credit disruptions after the September 2008 collapse of Lehman Brothers; small countries whose external financing and exports declined during the global crisis; the Euro Area program countries; and a few countries in the Middle East and North Africa (MENA) region that tried to address deep-seated structural and fiscal issues in an external environment strained by the global financial crisis.
Objectives. When the global crisis broke, many feared outcomes such as a second Great Depression or a cascading of contagion-fueled crises around the world. Fund-supported programs were undertaken in an environment where avoiding such outcomes was seen as crucial across the world. Program design in many cases was undertaken in the context of the extraordinary uncertainty that existed after the collapse of Lehman Brothers and during the Euro Area crisis. From an operational perspective, as in previous crises, IMF-supported programs provided balance of payments support, helping to smooth the pace of adjustment and restore investor confidence. The focus of the programs varied according to country circumstances but they generally shared a few common features. Programs focused on restoring external viability through adjustment in internal or external prices; improving competitiveness and productivity by addressing product and labor market rigidities; restoring fiscal sustainability through adjustment and restructuring; re-capitalizing banks and promoting non-financial private balance sheet restructuring; and strengthening financial supervisory and regulatory frameworks.
Outcomes. At a broad level, Fund-supported programs helped the global economy avoid the most feared outcomes. The Fund helped to chart a way through the crisis, using experience to inform future program design and contributing to the strengthening of frameworks and firewalls that gradually broadened the array of feasible policy choices over time. Even with Fund financial support, a significant adjustment was unavoidable for some countries as earlier imbalances were unwound. Without Fund financial support, even more rapid adjustment would have been necessary to close financing gaps, creating disruptions for program countries and risking a further intensification of the crisis. Fund-supported programs helped to smooth the adjustment and gain time for addressing problems. They helped the world and most program countries weather the effects of the crisis, cushion output, reduce imbalances, and stabilize financial systems. They helped equip a range of emerging economies and small states to handle the collapse of trade and financing flows in 2008–09; provided the Euro Area with time to build firewalls; and supported reforms and confidence in the MENA economies after the 2011 Arab Spring. About ¾ of the program countries have regained market access, and a third have substantially reduced their reliance on IMF financing.
While a few countries adjusted relatively quickly, in many cases underlying vulnerabilities remain, debt is still elevated, and the restoration of market access has occurred amidst easy global financial conditions with its durability to be tested. Unemployment remains high and growth generally tepid, reflecting weak global demand, limited exchange rate adjustment, continuing deleveraging, and a reduction in potential growth notwithstanding structural reforms. In part, this picture reflects the fact that recovery from financial crises tends to occur over protracted periods and adjustment is particularly difficult for all countries in a weak global environment.
External Adjustment. External imbalances and currency misalignments in recent crisis program cases were at least as large as those in previous programs. Nonetheless, recent programs both in countries belonging to currency unions and in those with independent exchange rates featured more rigid exchange rates, partly reflecting authorities’ existing exchange rate regimes as well as a recognition that large and abrupt currency adjustments could destabilize balance sheets with currency mismatches. Greater exchange rate rigidity implied a greater reliance on domestic price adjustment to restore competitiveness. In practice, internal devaluation proved difficult to achieve, and the desired recovery in growth and exports did not materialize for most countries.
Fiscal Policy and Public Debt. Programs typically sought to reduce fiscal deficits to lower public debt ratios over the medium term, taking into account available financing. While deficits were generally reduced in line with program objectives, the negative short-term effect on output was greater than envisaged in programs featuring large fiscal consolidations, in part owing to larger than expected fiscal multipliers. This, together with bank recapitalization costs and other factors that dampened activity, some of which were particularly hard to predict given the circumstances, led in several cases to a larger than expected rise in debt-GDP ratios during the program period. Where public debt significantly exceeded high risk thresholds, it was generally restructured through private sector involvement and, in a few cases, official sector involvement. Concerns about bank-sovereign linkages and cross-border contagion sometimes delayed or limited public debt restructuring, adversely affecting growth and credit intermediation.
Structural Reforms. Many programs featured extensive structural reforms, which in general were necessary for successful internal devaluation, focused on core areas of the Fund’s responsibilities, and may have reflected the Fund’s growing emphasis on the structural challenges to growth. Structural conditionality may, however, have resulted in reform fatigue in some cases and the growth payoffs from structural reforms in the near term appear to have been modest and less than envisaged.
Private Sector Balance Sheets. Households and corporates in many program countries entered the crisis with debt sustainability challenges. During the crisis, the private sector increased its saving (deleveraged) to rebuild stressed balance sheets, dampening aggregate demand. Program design generally identified balance sheet strains, but the drag on growth and the implications for key policy goals such as fiscal adjustment turned out to be more severe than envisaged, underlining the notion that recovery from financial crises tends to be protracted. Only modest progress was made toward the accelerated repair of balance sheets, reflecting moral hazard concerns about debt write-offs, potential fiscal costs, and gaps in insolvency and foreclosure frameworks.
Financial Regulation/Supervision. The crisis revealed an excessive buildup of risks in bank balance sheets, often as a result of gaps in supervisory arrangements. Once the crisis hit, progress in resolving insolvent financial institutions and recapitalizing systemic ones was relatively slow, which amplified financial market volatility, curtailed bank lending, and kept up borrowing costs for both sovereign and private borrowers. The delay in setting up a banking union in the Euro Area with a single supervisory-regulatory framework, resolution mechanism, and safety net was costly. Macro-prudential regulations were not a core feature of recent programs and, where adopted, generally focused on addressing banks’ foreign currency risks.
Regional Financing Arrangements (RFAs) and Currency Unions. Fund-supported programs that involved collaboration with RFAs benefited from RFAs’ regional expertise and an expanded financing envelope. The recent experience in the Euro Area could be useful for further building guidelines for future Fund-RFA collaboration, while recognizing that institutional frameworks and practices differ across cases. In the case of Fund arrangements for members belonging to currency unions, Fund program design took into account the fact that union-wide policies can have an important bearing on the member’s economic situation. When changes in such policies were warranted for program success, the Fund typically sought them through commitments or through its surveillance advice.
Considerations for Future Program Design
External Adjustment. Greater exchange rate adjustment helps address external gaps with a less adverse impact on output, though where foreign currency liabilities are substantial steps are needed to mitigate the impact of currency depreciation on balance sheets. For countries where nominal devaluation or depreciation is not a realistic option, for example those in currency unions, program design should recognize that the alternative route of internal devaluation is very demanding, requiring ambitious macroeconomic adjustment and structural reforms sustained over a period that can well exceed the standard 3–4 year period of Fund-supported programs. The policies of the currency union as a whole affects prospects for external adjustment by individual members: for example, internal devaluation is harder to achieve if inflation is very low and external demand is weak.
Fiscal Policy and Public Debt. Fiscal consolidation is generally key for adjustment, and its appropriate pace and size should reflect macroeconomic objectives, available financing, and debt sustainability. Program design should take into account the effects of fiscal consolidation on output. Where these effects are projected to be large, with consequences for program sustainability, it would be appropriate to seek additional financing to accommodate a more gradual consolidation; where public debt is high, timely debt restructuring may also be needed.
Structural Reforms. Structural conditionality remains important for achieving reforms necessary for adjustment and long-term growth. It may need to be more extensive where programs include broad-based reforms design to support internal devaluation, but should pay regard to authorities’ implementation capacity. At the same time, modest short-term growth dividends from supply-side structural reforms should be reflected in realistic and prudent program assumptions.
Private Sector Balance Sheets. Balance sheet strains and the associated drag on credit and activity take time to resolve. While an early start to reforms is important to support recovery over the medium term, program design should not typically anticipate large near-term benefits for economic activity. Early priorities for tackling private debt overhangs include attention to legal frameworks and out-of-court settlement options, prudential measures to incentivize debt write-offs and restructuring when needed, and the creation of markets or institutions to handle distressed debts. In crisis circumstances, the benefits of policies designed to write down debts may exceed the adverse impact on public balance sheets and moral hazard considerations. Steps to address balance sheet data gaps can help identify vulnerabilities and transmission channels, and inform decisions on the merits and costs of exchange rate depreciation. Regarding financial regulation and supervision, sustained and proactive steps are important toward strengthening institutional frameworks to help prevent the build-up of risks.
RFAs and Currency Unions. The G-20 principles for cooperation between the Fund and RFAs provide a helpful foundation for developing more updated and operational guidelines in this area. The guidelines could ensure clear roles for various partners, such as the Fund’s responsibility for macroeconomic and debt sustainability analysis, as well as critical and parsimonious conditionality that does not overburden national implementation capacity. Consistent with ongoing practice, where changes in currency-union-wide policies are important for program success, the Fund should provide advice through its surveillance as warranted or, when necessary (including for financing assurances), seek commitments on prospective implementation of necessary union-wide policies; alternatively, program design would need to be based on larger adjustment and financing, or Fund involvement be postponed.
Overview of Crisis and Program Outcomes
1. In the early 2000s, the global economy experienced strong growth, loose monetary policy in advanced economies, and a boom in international capital flows. Euro accession led to easier domestic financing conditions for several Euro Area countries, accompanied by financial liberalization.2 Aggregate demand and credit growth were generally robust across the globe, boosting property and other asset prices. In several fast growing countries, rapid wage and price inflation eroded competitiveness and current account deficits emerged amidst fixed or managed exchange rates. Indeed, three years before their respective programs, all but one of the 27 countries covered in this review had some form of fixed or heavily managed exchange rate arrangement. Large current account deficits were justified in part by convergence with creditor countries, but rather than financing productivity-enhancing investment, foreign saving frequently financed asset booms, including real estate (Ireland, Iceland, Latvia, and Portugal), and supported government consumption (Greece).
2. The demise of Lehman Brothers in September 2008 triggered a sudden loss of confidence in many countries, a spike in counterparty risk, and a collapse in activity. In several economies, capital inflows stopped or reversed, financial systems came to an abrupt halt, and external stability problems emerged. Countries with strong trade linkages with crisis-hit countries experienced sharp falls in external demand that exposed underlying vulnerabilities. As the crisis persisted and global growth remained weak, economies less directly exposed to the initial shock also started to experience balance of payments difficulties (Figure 1).
3. As troubled economies sought official financial support, the demand for Fund resources rose to unprecedented levels. The Fund approved financial arrangements under the General Resources Account (GRA) in the amount of SDR 420 billion during 2008–13. Arrangements from which members made drawings accounted for SDR 167 billion, of which nearly SDR 119 billion was drawn during this period. utstanding IMF credit, including both GRA and PRGT resources, rose from under SDR 10 billion in 2007 to over SDR 90 billion in 2013, declining with repayments to nearly SDR 60 billion as of June 2015.
4. The International Monetary and Financial Committee (IMFC) has requested reviews of Fund-supported programs undertaken since the global financial crisis. The first such review took place in September 2009 amid the surge in IMF lending after the collapse of Lehman Brothers and was followed by updates to the Executive Board in 2010–12. A general conclusion was that program objectives and recommended policies in the first wave of arrangements approved in 2008–09 (mainly for non-Euro Area Europe) were largely appropriate; by contrast, the challenges faced by the second wave of arrangements approved in 2010–11 (focused on Euro Area members) were greater and it was less clear that program financing and policies were adequate to meet them.
5. This paper responds to the request for an updated review of crisis programs. It reviews the performance of Fund-supported programs since the outbreak of the global financial crisis with a view to drawing lessons for future program design. It takes into account the conclusions of the previous reviews, subsequent developments, and the 2014 Independent Evaluation Office (IEO) report on the IMF’s response to the crisis.3 The review seeks to assess program countries’ progress toward addressing external and internal imbalances and whether program objectives and policy recommendations were appropriate given the initial imbalances and structural problems. The assessment is mindful of the role played by the authorities’ ownership and implementation of program policies, external demand conditions, and other countries’ policies on the effectiveness of Fund-supported programs. It is also cognizant of the context in which programs were designed, including the extraordinarily high degree of uncertainty that existed after the collapse of Lehman Brothers and during the Euro Area crisis. Experience in the recent program cases is compared with the experience of countries that did not request Fund support and with that of countries that made use of Fund resources in earlier programs.4
B. Analytical Country Groupings
6. The crisis affected countries through different transmission channels and over different timeframes. Some were affected directly through a disruption of credit flows, others indirectly by trade and financial spillovers, and yet others only generally by the weak global growth environment that made external adjustment harder. An analytical grouping of countries along these lines corresponds broadly to regional groupings as follows:
A group mostly comprising European emerging markets required Fund financial support in 2008–09 when capital flows dried up at the start of the crisis. This group includes Georgia, Hungary, Iceland, Latvia, Ukraine (2008 and 2010), and Armenia, Belarus, Bosnia and Herzegovina, Mongolia, Romania, Serbia, and Sri Lanka (requests in 2009).5 Later arrangement requests from Moldova (2010) and Kosovo (2012) were similar in character.
Several small highly open economies had domestic vulnerabilities that were exposed by the disruption they experienced through their trade, tourism, and financial linkages with the United States and crisis-affected countries. These economies were the Seychelles (2008), Dominican Republic, Maldives (2009), Antigua and Barbuda, Jamaica (2010), and St. Kitts and Nevis (2011).6
The Euro Area crisis countries of Greece (2010, 2012), Ireland (2010), Portugal (2011), and Cyprus (2013) faced problems that were partly associated with the global crisis but whose more immediate origin was their public and private balance sheet vulnerabilities and the accumulation of large current account imbalances within the Euro Area.
Some Middle East and North African (MENA) countries faced fiscal and structural vulnerabilities that were heightened by the global crisis (Pakistan, 2008) or strained by the economic dislocations associated with the 2011 Arab Spring (Jordan, 2012; Tunisia, 2013).
C. Program Design
7. Program design had to take into account particularly difficult initial conditions. Coming out of the Great Moderation many countries had large current account deficits, overvalued exchange rates, and high public and private debt. External adjustment was constrained by weak external demand, limited scope for exchange rate adjustment under existing exchange rate regimes and, in several cases, concerns that large or sudden exchange rate depreciation could lead to balance sheet disruptions arising from large foreign exchange liabilities.7 Balance sheet adjustment was also complicated by the fact that where the public sector, households, and corporates all sought to deleverage, the reduction in spending undercut domestic demand and contributed to raising the real burden of debt further.8 With domestic financial systems being important creditors of the private and public sectors, debt restructuring options seemed constrained by the impact they might have on banks’ balance sheets and, therefore, financial intermediation. Low global inflation limited the extent to which inflation could reduce real debt burdens.
8. In addressing these challenges, program design needed to find the right balance between large upfront adjustment and more gradual approaches. Program design was conducted in the context of often unprecedented circumstances, such as a particularly weak global economy and large adjustment needs, and an environment of unusual uncertainty in the early stages of the crisis and intervening episodes. In general, unlike earlier programs, the recent crisis programs leaned toward more gradual approaches, seeking to restore competitiveness through modest depreciation or internal devaluation rather than substantial exchange rate correction, to strengthen debt sustainability through fiscal consolidation rather than debt restructuring, and to address banking sector concerns through recapitalization rather than closing insolvent banks.9 Large imbalances and limited exchange rate flexibility led to protracted adjustments that entailed large financing.
9. The programs co-financed with RFAs and those with members of currency unions highlighted special considerations for the design of Fund-supported programs. Under currency unions, effective control over several policy levers—monetary, exchange rate and financial policies— lies with union-wide institutions rather than the individual country. For programs designed in close collaboration with an RFA, it is important to clarify the roles of various institutions, including the Fund’s responsibility for aspects such as the macroeconomic framework and debt sustainability analysis, and to ensure that conditionality applied by the Fund and RFA remains parsimonious and macro-critical when considered in aggregate.
D. Program Outcomes
10. Fund-supported programs during 2008–13 helped the global economy avoid the feared counterfactual outcome of an even deeper and more severe crisis. Programs helped the global economy and most program countries weather the effects of the crisis. The decline in output from its previous artificially elevated levels was cushioned, imbalances were reduced, and financial systems stabilized. A range of emerging economies and small states were able to handle the collapse of trade and financing flows; the Euro Area gained time to build firewalls against contagion; and reforms and confidence in the MENA economies were shored up after the 2011 Arab Spring. As of early 2015, about ¾ of the program countries had regained market access (Figure 2), and a third had substantially reduced their reliance on IMF financing. Only 5 of the 27 program countries required successor arrangements. Program outcomes reflected the fact that the focus of program design varied across countries based on the problems countries faced. Good progress was made, for example, in tackling financial sector strains in Iceland, Ireland, and Latvia, while effective external adjustment was achieved in Fund-supported programs for Armenia, Latvia, Seychelles, and Sri Lanka. Appropriate fiscal adjustment was achieved by Cyprus, Jordan, and Greece (2012), and broad structural reforms were successfully implemented by Armenia, the Dominican Republic, Hungary, Jamaica (2013), and Seychelles.
11. The Fund helped to chart a way through the crisis, using experience to inform future program design and contributing to the strengthening of frameworks and firewalls that gradually broadened the array of feasible policy choices over time. Program design was tested by new risks and uncertainties. Given the radical differences between the 2008 crisis and its predecessors during previous Fund experience, the Fund like other institutions faced significant uncertainties about the shocks, transmission channels, and appropriate policy responses. The Fund was able to identify and share solutions from the formulation and implementation of Fund-supported programs in this environment. Program outcomes helped inform the design of later Fund-supported programs; for example, program design changed over time to include a stronger emphasis on addressing the challenges to internal devaluation and increasingly focused on restarting credit intermediation. The experience with Euro Area programs informed the Fund’s recommendations for strengthening firewalls and developing banking union. More generally, the Fund’s engagement in supporting programs across 27 countries provided it with the expertise needed to inform, advise, and help in the coordination of a global response.
12. While a few countries adjusted relatively quickly, in many cases underlying vulnerabilities remain. The restoration of market access has occurred amidst easy global financial conditions with its durability still to be tested and debt is relatively elevated. Unemployment remains high and growth has generally disappointed, reflecting weak global demand, little exchange rate adjustment, large fiscal multipliers, and a reduction in potential growth notwithstanding structural reforms. In part, this picture reflects the protracted nature of the recovery from financial crises and the difficulty of achieving adjustment in the context of a weak global environment.
13. Growth projections for the world and key countries were revised down serially over time, as reflected in the WEO and market consensus forecasts. The growth disappointments reflected the difficulty of making forecasts amidst uncertain global demand conditions following the crisis, drags from private sector deleveraging on demand, and unforeseen downward revisions to potential growth. Growth shortfalls also reflected larger than expected multipliers in the context of ambitious fiscal adjustment, modest short-term dividends from structural reforms, limited progress in restoring competitiveness, and, in some cases, worsening security conditions and political turmoil. For program countries too, growth fell short of projections, especially in the Euro Area programs, although not in the Caribbean and small states (Figure 3). This pattern seems consistent with the 2014 IEO finding that current-year forecasts in exceptional access programs were initially optimistic.10
14. Inflation outcomes were mixed relative to projections. Fund projections were not consistently too high or low, for both program and non-program countries. Inflation turned out to be broadly in line with projections for small states programs, higher than programmed in the Euro Area, and lower than programmed in other emerging markets cases.
15. Social benefit spending was generally protected, but income inequality trends varied. Traditional indicators such as the expenditure-based Gini coefficient show some narrowing of inequality, suggesting program design was effective in protecting low-income groups. Proxies for income disparities based on estimates of labor and capital income shares show that for small states and non-European emerging markets, labor income rose relative to capital income during the program period. In Euro Area programs, however, capital income shares generally rose during the programs while labor income shares have fallen over time, suggesting some redistribution away from labor. Many programs were successful in safeguarding social expenditures at pre-program levels, or in increasing social spending (Caribbean and small states), although Euro Area programs saw a small decline in such spending-to-GDP ratios in the context of large fiscal adjustments (Figure 4; discussed further in the Fiscal Adjustment section below).
16. Program design was generally tailored to country circumstances, sharing similar features across countries that faced similar challenges and circumstances. Programs targeted different degrees of fiscal adjustment depending on the size of initial fiscal imbalances and the strength of economic activity, being largest in the Euro Area programs (Figure 5). External current account adjustment objectives varied across programs, being largest in small states, where initial imbalances were large. Structural reform content also varied across programs, as measured by the number of structural benchmarks and structural performance criteria, which were more numerous in Euro Area programs needing to achieve internal devaluation or in successor arrangements, where reforms were intensified to tackle unaddressed challenges. These observations echo the findings of the 2011 IMF Review of Conditionality for Fund-supported programs during 2002–11.11
17. Program design in the recent crisis programs differed from previous program episodes in response to different circumstances as well as lessons from previous experience. As in the past, programs focused on restoring external and fiscal sustainability, improving competitiveness through structural reforms, addressing balance sheet problems in the financial sector, and strengthening financial supervision and regulation.
18. Access to Fund financing reflected country circumstances. The large size of some recent arrangements, notably in the Euro Area, reflected relevant members’ adjustment challenges, financial development, and close integration with global financial markets. The latter consideration also gave rise to concerns about systemic contagion that motivated a modification in the Fund’s exceptional access policy (IMF, 2013). Access as a percent of GDP was largest in Euro Area programs, commensurate with relatively ambitious fiscal and structural reform objectives. The 2011 Review of Conditionality also noted that programs with higher and more front-loaded access generally involved countries experiencing capital account crises, which tend to impose larger balance of payments pressures.
19. Program design in the context of currency union members recognized the important influence of union-wide policies on the economic situation of the member. Where the success of a Fund-supported program depended on policies implemented by a regional central bank or other union-wide authorities, the Fund sought assurances on how this policy change would be implemented. In some cases, this took the form of conditions to be implemented by the union wide institution (St. Kitts and Nevis, Antigua and Barbuda). In others, such as for financing assurances in the Euro Area programs, formal commitments were stated in Eurogroup press releases and reflected in staff reports.12 Finally, some policy changes at the union wide level that were supportive of program goals (banking union, monetary policy) were not made conditions or safeguards commitments and instead were part of the advice provided in the regular surveillance discussions with EU institutions on Euro Area policies in the context of Article IV consultations with member countries.13 The different approaches reflected, in part, the nature of the intended policy changes: the cases which used conditions referred to regulatory and supervisory actions within the purview of the union-wide institution that affected the program country but not other members of the currency union.
F. Technical Assistance
20. IMF technical assistance (TA) played an important role in supporting policy implementation in recent programs. On average, TA delivery to program countries nearly doubled compared to the pre-program year, measured in person-years (Figure 6). TA increased most significantly for Euro Area programs (up from a very low pre-program base) and for small states. The largest program TA recipients were Antigua and Barbuda, followed by Greece, Cyprus and Ukraine. Among the Euro Area programs, Ireland was an exception that received virtually no TA during its Fund-supported program, reflecting its strong institutional capacity. Fiscal expertise accounted for most of the increase in TA to Euro Area countries and to small states. While TA rose less markedly for other emerging markets and MENA countries, it was refocused on fiscal TA for emerging markets and on financial TA for MENA countries. Overall, delivery of technical assistance was closely related to the ambition of the countries’ reform agendas, as proxied by the number of conditions per program year (Figure 7).
Recent Experience with the Flexible Credit line (FCL), Precautionary Credit Line (PCL), and Precautionary and Liquidity Line (PLL)
In 2009–10, the Fund introduced the Flexible Credit Line (FCL) and the Precautionary Credit Line (PCL) for countries with very strong (or “sound” in the case of the PCL) economic fundamentals, institutional policy frameworks, and policy implementation records. The new instruments were part of a broader reform of the IMF toolkit in response to the 2008 crisis. The instruments aimed to bolster confidence in the member qualifying for these instruments, reduce the stigma of Fund support by involving reduced or no ex post conditionality, and strengthen crisis prevention by offering Fund support ahead of a crisis.
Since the creation of the instruments, three countries have used the FCL (Colombia, Mexico, and Poland) and two countries have used the PCL or its successor the Precautionary and Liquidity Line (PLL) (FYR Macedonia and Morocco). Of the five countries, only FYR Macedonia has ever drawn upon its precautionary facility. The nature of the vulnerabilities that motivated their request varied across the 5 countries including vulnerability to commodity price shocks (Colombia and Morocco), significant non-resident portfolio investment (Mexico and Poland), and exposure to the crisis in the Euro Area (Morocco and Poland). By definition, the extent of vulnerabilities was greater in the PCL/PLL cases, but all countries had relatively strong fiscal frameworks, robust institutions, and (except for Macedonia and Morocco) were helped by their flexible exchange rates. During the crisis, Colombia and Poland avoided a contraction altogether and Mexico quickly recovered from a one-year recession (Box Figure 1 below). In addition to stronger fundamentals, aspects of these countries’ policies contributed to this outcome, as documented in FCL reviews (IMF (2014d)):
First, these countries sought out additional external buffers (FCL, Central Bank swaps) before there was a balance of payments need. Coming early for insurance helped limit contagion and a capital account crisis.
Second, they had the space to pursue counter-cyclical fiscal policy. Cyclically adjusted fiscal positions widened by about 3 percent of GDP in all three FCL users. Access to the FCL helped reassure markets amid the additional borrowing.
Third, despite larger external buffers and smaller imbalances, they allowed exchange rate adjustment throughout the crisis. Real exchange rates depreciated peak to trough by 25 percent, more than double the adjustment in the CPR sample.
While the degree to which these policies represent implementable lessons for the CPR sample is constrained by the fact that the available policy tool kit is itself a function of fundamentals, institutions, and existing imbalances, the experience of the FCL users underscores the importance of building buffers in good times and the capacity of such buffers to mitigate the impact of shocks in a crisis.
Program countries generally entered the crisis period with large external current account deficits and exchange rate overvaluation that partly owed to rapid credit growth before the crisis. Program objectives were tailored to country-specific circumstances, but large initial external imbalances, together with weak global demand and limited exchange rate flexibility, implied a protracted and painful adjustment in many cases and limited the progress in restoring external viability. In general, greater exchange rate adjustment helps to address external gaps with a less adverse impact on output. If foreign currency liabilities are substantial, steps are needed to mitigate the impact of currency depreciation on balance sheets. For countries where nominal exchange rate adjustment is not a realistic option, for example those in currency unions, program design should recognize that internal devaluation may require large macroeconomic adjustment and deep structural reforms sustained over long periods that can exceed the standard 3–4 year period of Fund-supported programs, and also require large and continued financing. The policies of the currency union as a whole can importantly influence the economic situation of individual members.
21. Program countries entered the crisis with larger external imbalances than countries typically had in earlier Fund-supported programs. Easy global financial conditions in the run up to the crisis had fueled rapid credit growth that widened current account deficits and moved exchange rates away from fundamentals. Exchange rates, which were generally fixed or heavily managed, often had overvaluations well over 10 percent (Figures 8 and 9).14 When the crisis came and triggered a sharp rise in global risk aversion, most countries experienced a sudden stop in capital flows that was large compared to earlier programs.
B. Objectives and Outcomes: Addressing Balance of Payments Gaps through Financing and Adjustment
22. Programs relied on a mix of adjustment and official financing to address balance of payments shortfalls. The balance between adjustment and financing that was sought varied significantly across the sample, reflecting countries’ differing access to exceptional support from partner countries, the extent to which economic shocks were expected to be temporary or permanent, and the scope for rapid exchange rate adjustment (Figure 10).
23. Financial support from other official creditors often occurred alongside Fund financial support. The relative “burden sharing” between the Fund and other official creditors depended on circumstances that differed from program to program (Figure 11). Some cases, such as those in the Euro Area, lent themselves more directly to burden sharing due to the presence of financing partners with access to substantial reserve currency funding capacity. In many of these cases, the financing need was particularly large, and constraints on the Fund’s own balance sheet increased the need for other creditors. While the Fund’s proportionate contribution to official sector financing varied considerably, Fund financing in absolute terms was large. Over half of the recent programs entailed exceptional access to Fund resources (19 out of the 32 programs).15 Euro Area programs were some of the largest in the Fund’s history, both in SDR terms and as a percent of quotas, reflecting the size of their financial systems, their adjustment challenges, and their close integration into global capital markets. Access expressed as a share of financing need was comparable to many non-Euro Area programs (Figure 10).
24. The extent of external current account adjustment varied across programs. In general, the Euro Area and emerging market program cases adjusted by more than projected, and small states by less, while adjustment in MENA programs was roughly in line with projections (Figure 12). Compared to earlier program cases, external adjustment was larger in Euro Area and emerging market programs, but smaller in the small states and MENA programs. External debt outcomes ran counter to expectations given the current account adjustment. Despite larger-than-programmed current account adjustment, external debt as a percent of GDP in the Euro Area cases remained higher-than-programmed, largely due to weaker output. Meanwhile, small states achieved initial objectives of reducing external debt levels: additional borrowing to finance higher-than-programmed current account deficits was offset by substantial debt relief for the Seychelles. External debt outcomes disappointed in other emerging markets and MENA countries.
25. The sustainability of the external current account adjustment remains to be tested. Even where countries have regained access to financing, current account deficits often remain above debt-stabilizing levels (Figure 13). Real exchange rates have typically adjusted little, and current account adjustments have instead reflected a compression of demand, suggesting external gaps may re-emerge as activity recovers.
C. Exchange Rate Policies
26. While programs designed around fixed or managed exchange rates are not new, the limited reliance on exchange rate adjustment to complement expenditure reduction as a means of adjustment in recent crisis programs is a difference from the approach in previous episodes. Fund-supported programs often seek adjustment in the form of both demand management policies (such as fiscal consolidation) to reduce domestic absorption and expenditure switching policies (such as nominal exchange rate adjustment) to redirect production toward tradable sectors. As noted, before the crisis, all but one of the recent program countries had some form of fixed or heavily managed exchange rate (Figure 14). In the years since the crisis broke, about half of the sample has moved toward some greater flexibility, but the actual variation in nominal exchange rates remains far below that of earlier program cases (Figure 15). Only in Iceland and the Seychelles did the nominal exchange rate depreciate significantly.
27. The different approach to exchange rate policy compared with earlier programs reflects a number of factors. For program countries in currency unions, currency union membership was taken as given for program design purposes (Euro Area and East Caribbean Currency Union programs). Outside this group, however, most programs sought to maintain pegs at close to pre-program levels (Latvia) or pursue managed, gradual depreciation (Jamaica). In these cases, program design around more stable exchange rates reflected, in part, recognition of the balance sheet risks associated with large or abrupt exchange rate depreciation (Dominican Republic, Georgia, Iceland, Jamaica, Latvia, and Ukraine) as well as responded to the authorities’ strong commitment to the peg (Jordan and Latvia).
D. Monetary Program Design
28. In several cases, exchange rate stability provided a nominal anchor for monetary programs. Inflation outcomes were close to program projections in a global environment of weak growth, low inflation, and depressed commodity prices in the aftermath of the global financial crisis (Table 1). Programs designed around pegged exchange rates and inflation targeting regimes saw smaller inflation surprises than regimes based on monetary targets such as net domestic asset or base money ceilings. Overall, operational monetary targets were met more consistently than under earlier programs, possibly because large inflation surprises in the latter complicated monetary management.
Monetary Program Design and Outcomes
Program launch at time (T).
Armenia, Georgia, Hungary, Iceland, Moldova, Romania, and Serbia.
Armenia, Georgia, Hungary, Iceland, Moldova, Romania, and Serbia.
29. Exchange rate stability was achieved through a combination of monetary tightening and currency intervention. Several programs saw sizeable increases in real interest rates to stem capital outflows, most notably in floating rate regimes experiencing significant real depreciation pressures (Figure 16). As global liquidity conditions eased, floating rate regimes often benefited from larger than expected capital inflows: in these cases, intervention to moderate currency appreciation led to larger reserve accumulation than planned (Figure 17). By contrast, programs featuring soft pegs consistently fell short of program goals for reserve cover, suggesting less success in attracting capital inflows and more prolonged exchange market intervention to support the currency. The sustained shortfall in reserve cover in these latter cases suggests that intervention was partly sterilized, limiting the associated monetary tightening.
30. The relative rigidity in exchange rates led programs to rely on domestic price adjustment to restore competitiveness. But “internal devaluation” proved very hard to achieve. This experience is in line with historical experience that internal devaluation is rarely achieved, and then typically in the context of supportive trends in partner country growth and inflation, favorable relative movements in the anchor currency, and flexible domestic markets (Box 2). The recent program cases achieved a real effective exchange rate depreciation of just 12 percent over the program period compared with an average of 48 percent in earlier programs. In recent program cases, real effective depreciation during the program period averaged only 7 percent for countries that maintained exchange rate pegs in a currency union or relative to an external anchor currency and 20 percent for those with more flexible exchange rate arrangements. 16
31. In practice, only limited progress was made in improving competitiveness. Labor market reforms were attained ahead of product market reforms, with the benefits of lower labor costs for competitiveness being, therefore, blunted by limited adjustment in producer prices and supply response because of barriers to new entry. In Euro Area programs, for example, a reduction in wage costs relative to trading partners was reflected in a depreciation of ULC-based real exchange rates, but the improvement in price competitiveness was more modest as evidenced by only small depreciations in CPI-based real effective exchange rates (Figure 18).
32. While internal devaluations in recent programs achieved only modest real exchange rate adjustment, where they did do so there is some evidence of a nascent growth impact. While trend growth generally declined during recent programs and fell short of end-program projections, the decline and shortfall were, on average, smaller for the countries that made more progress toward internal devaluation. This suggests that where trend growth fell short of expectations it owed in part to internal devaluation not being achieved (Figures 19 and 20). It is possible, therefore, that sustained pursuit of an internal devaluation strategy over periods longer than the typical program period could deliver growth dividends, provided that financing is available to accommodate the slower pace of adjustment.
33. Capital flow management measures (CFMs) on capital outflows were imposed in response to crisis conditions in Iceland (2008) and Cyprus (2013).17 The experience with CFMs both informed and was informed by the development of the Fund’s institutional view on the liberalization and management of capital flows. The CFMs were put in place by the member before the start of the programs to help stem capital outflows when crisis was imminent, with a view to being eliminated as conditions stabilized. In Cyprus, a roadmap was adopted for the relaxation of restrictions on cross border flows with steps that would depend on progress in rebuilding domestic financial intermediation by restructuring the banking sector and restoring depositor confidence and bank liquidity. In the event, CFMs were eventually removed with little impact on markets and the banking sector. In Iceland, unwinding CFMs proved more difficult than anticipated owing to the size and complexity of the problem, and the liberalization strategy had to be updated successively over the course of the program.
Promoting Internal Devaluation
Historical experience suggests that few countries have achieved a significant depreciation of the real exchange rate without a nominal depreciation. Indeed, since the 1990s, there have been only 16 episodes of emerging markets with fixed exchange rates that have achieved CPI-based real effective depreciations exceeding 15 percent. In two cases, this was achieved through domestic deflation (Antigua and Vietnam), with gains in other cases reflecting either higher inflation in partner countries or fortuitous depreciation of the peg currency, so that the national currency was able to depreciate in nominal effective terms without exiting the peg.
Historical experience indicates a number of factors that seem critical for achieving internal devaluation. Where current account imbalances were initially small and exports were already large in relation to GDP, the country was better placed to achieve internal devaluation with less negative output consequences. A flexible domestic economy is also helpful, particularly labor markets; moreover, relative prices are easier to adjust when global growth and partner inflation are higher. Internal devaluation has also been more successful when countries have small debt stocks and substantial policy space (especially fiscal). In more typical cases, internal devaluation may only be achievable slowly, requiring sustained high levels of program ownership and implementation capacity as well as access to large and continued financing.
Latvia succeeded in achieving a large reduction in unit labor costs and restoration of output. Unit labor costs fell by 25 percent in 1 year, while the CPI-based real effective exchange rate fell only modestly. The resulting large increase in profits in the tradable sector generated a significant supply response as new firms entered. Output had contracted substantially at first (24 percent peak to trough), but recovered subsequently thanks to stronger exports. Notable contributory factors included the authorities’ strong commitment to achieve the adjustment even if the costs were relatively high, a substantial rise in labor productivity, and, as a small open economy with flexible product and labor markets, the ability of the tradable sector to generate a rapid supply response (Blanchard, Griffiths, and Gruss, 2013).
Internal devaluation through labor and product markets reforms was a key objective of Euro Area programs. Greece’s program (2010 and 2012) aimed to do so through a series of labor market measures, such as cutting nominal general government wages and benefits, reducing the minimum wage, and reforming collective bargaining, as well as broader measures such as state enterprise reform and divestment, cutting red-tape, and lowering barriers to entry to promote domestic competition. The Portuguese program was initially anchored on internal fiscal devaluation, achieved by rationalizing the wage bill. Other structural reforms included liberalization of the non-tradable sector and labor reform. Competitiveness was also an issue for Ireland at the outset, but high price and wage flexibility helped the goods and labor markets adjust relatively quickly.
Fiscal consolidation helped support external adjustment, responded to financing constraints, and was necessary to support goals for reducing debt-GDP ratios over the medium term. Fiscal deficits were generally reduced in line with program objectives. In some programs with large fiscal consolidations, the negative impact on output in the short term was, however, larger than anticipated and combined with other factors to weigh down output. The cumulative effect was to raise debt-GDP ratios by more than expected in the short term. Where the near-term impact on output of large fiscal consolidation is likely to be large and protracted with consequences for program sustainability, it is desirable to examine the scope for slower fiscal consolidation—requiring additional financing—and to restructure the debt if it is not deemed sustainable with high probability.
A. Background and Fiscal Objectives
34. Many program countries were characterized by high levels of public debt.18 In some cases, high debt reflected large borrowing in the period leading up to the crisis (Greece 2010), in some it reflected the costs of financial sector interventions (Ireland), and in others it reflected large cyclical deficits that emerged during the crisis and prior to the program (Cyprus, Greece 2012, Ireland, Jordan). Programs typically sought to reduce fiscal deficits, both to ease short-term fiscal and external financing pressures and to put public finances onto a sounder medium-term footing. In order to help reduce high public debt-GDP ratios, some programs targeted primary fiscal surpluses. One in five programs supported public debt restructuring where indebtedness had risen to a point that could not realistically be addressed by fiscal adjustment alone.
35. In general, program policies sought to balance the benefits of stronger fiscal positions against the impact of consolidation on output. Programs with larger initial cyclically-adjusted primary deficits and stronger activity as measured by the output gap sought greater fiscal consolidation (Figures 21 and 22). 19 On average, programs sought to strengthen primary fiscal balances by about 3 percentage points of GDP, in total, over a three-year period. Programmed fiscal adjustment was larger in the Euro Area (averaging 5½ percentage points of GDP of primary balance adjustment) and smaller in MENA programs (about 2½ percentage points of GDP).
36. Programs typically sought fiscal consolidations through expenditure cuts (Figure 23). Many program countries faced very high and seemingly unsustainable public spending ratios, exceeding 50 percent of GDP in most Euro Area program countries. Much of this spending was related to energy subsidies (MENA), wages (Euro Area, MENA, and small states), and social benefits, including pensions (Euro Area). Many program countries also had weaknesses on the revenue side, such as in tax administration, that made a heavy reliance on higher revenue unlikely.20 Program design was generally mindful of protecting social safety nets. A quarter of programs had the explicit objective of poverty reduction21, and around half of the programs included some form of social protection in their conditionality. 22
B. Fiscal Outcomes
37. Fiscal consolidation outcomes were mixed. Primary fiscal balances strengthened by an average of 1¾ percent of GDP during the 3 years after program approval, relative to program goals of about 3 percentage points over three years. Fiscal outcomes were closest to program objectives in the Euro Area and MENA, but fell short in other regions (Figure 24). In the Euro Area, where program fiscal balance targets were met, it was typically in the context of disappointingly weak economic activity, implying a much larger than programmed cyclically adjusted fiscal correction, while in small states original program targets were missed due to expenditure overruns. Except for small states, program fiscal consolidation was larger in the crisis program cases, on average, than in non-program countries, where fiscal deficits typically significantly exceeded projected levels. Capital expenditure in relation to GDP rose in the Euro Area cases but the experience elsewhere was more mixed. Programs were generally effective in protecting social benefit spending, both in relation to total expenditure and relative to pre-program periods, and in contrast to the experience in non-program countries.23 Social spending as a share of total expenditure increased across regions. Social spending was broadly preserved as a percent of GDP although with some variation across regions, declining slightly in the Euro Area programs and increasing slightly in MENA and the small states. 24
38. In some cases, the contractionary effect of fiscal consolidation on output may have contributed along with other factors to raising debt-GDP ratios by more than expected in the short run. Fiscal consolidation was necessary to reduce debt-GDP ratios over the medium term. While progress in strengthening fiscal balances reduced the pace of new borrowing, it did not reduce nominal debt levels. Moreover, some countries implementing large fiscal consolidations experienced sizeable declines in activity reflecting larger-than-anticipated fiscal multipliers as well as reversals of output from artificially inflated levels and weaker activity associated with weak global demand, political uncertainty and incomplete reform implementation (IMF, 2013b). As a result, despite fiscal adjustment, debt-GDP ratios rose more than expected over the program period in some countries (Armenia, Bosnia and Herzegovina (2009), Greece (2010), Iceland, Ireland, Latvia, Maldives, Portugal, and Romania; Figures 25 and 26, Annex I). This trend was exacerbated in some countries by bank recapitalization costs, which amounted to 19 percent of GDP on average, or 60 percent of the short-term rise in the debt-GDP ratio, and the reclassification of the debt of public enterprises that were previously outside of the general government perimeter (Portugal).
39. As the contractionary effect of fiscal consolidation on output became evident, many programs slowed the pace of consolidation (Armenia, Greece, Hungary, Latvia, Portugal, and Ukraine; see Figure 27). In small states, strong initial primary fiscal positions (mainly from pre-crisis programs) allowed fiscal policy to play a somewhat countercyclical role, financed by earlier than expected market access, interest savings from debt operations, and concessional financing provided in successor arrangements.
C. Debt Sustainability
40. Debt reduction measured by public debt-GDP ratios fell short of medium-term program expectations in three-quarters of programs (Figure 28). The shortfall reflected fiscal adjustment that was somewhat smaller than programmed, as well as disappointing growth outcomes in many cases. Other contributory factors included upward revisions to baseline primary deficits, currency depreciations in a few cases (Belarus, Jamaica (2010), Pakistan, Seychelles (2008), and Ukraine (2010)), and lower-than-expected asset sales and higher bank recapitalization costs (Greece) (Figure 29).25 Debt surprises were particularly large in non-restructuring cases, notably for several countries that needed successor arrangements (Greece (2010), Mongolia, Serbia, Seychelles (2008), and Ukraine (2008, 2010)). In these cases, public debt exceeded medium-term program projections by more than 20 percent of GDP. Among Euro Area programs, debt-GDP ratios exceeded targeted levels for Greece (2010) and Portugal (2011).26
41. About one case in five featured sovereign debt restructuring. Of 32 programs, seven included either a debt reprofiling through maturity extension with no face value reduction (Antigua and Barbuda (2010), Cyprus (2013), and Jamaica (2010 and 2013)) or a debt operation with a face value debt reduction (Greece (2012), Seychelles (2010), and St. Kitts and Nevis) (Box 4). These cases represented the majority of program cases in which initial public debt was most elevated. While there is no specific threshold for debt restructuring, featured countries had debt levels exceeding 90 percent of GDP (Figure 30). In three other advanced economy cases, relatively high debt ratios were assessed to be sustainable with a high degree of probability (Iceland, 2008), or concerns about systemic spillovers precluded the consideration of debt restructuring options (Ireland and Portugal).27
42. Among countries with high public debt levels, disappointments in growth and public debt outcomes were smaller in cases that included debt restructuring. Debt-GDP ratios were, on average, higher-than-programmed in the non-restructuring cases, while growth was lower (Box 3). By comparison, restructuring cases achieved more substantial growth turnarounds and declining debt-GDP ratios, broadly in line with initial program expectations. It should be noted that some of the non-restructuring group countries had lower debt ratios than the restructuring group ahead of their respective programs so that debt reduction was less urgent, and they also had stronger pre-program growth so that the scope for improvement during the program period was more limited.28
43. Market access, as signaled by sovereign ratings and bond spreads, took time to recover. Sovereign ratings deteriorated in all cases during the first program year, suggesting that perceptions of debt sustainability were slow to improve. Ratings improved subsequently in most program countries, often reflecting external developments. Sovereign bond spreads in Euro Area program cases remained elevated until mid-2012, however, before they declined helped by decisive policy actions by the ECB. Narrowing spreads contributed favorably to debt dynamics in most countries, although typically not by enough to be offset by the adverse impact of growth disappointments.
44. Concerns about contagion were a critical consideration in decisions about debt restructuring in the Euro Area programs. For Greece (2010), the Fund could not assess that debt was sustainable with high probability, an assessment normally necessary for exceptional access to Fund resources. However, the risk of systemic international spillovers, given the lack of firewalls at the time to insulate other members of the Euro Area, provided a justification for not requiring an upfront debt reduction operation as a condition for a Fund arrangement with exceptional access (Figure 31 shows European banks’ exposure to Greece at the time). A “systemic exemption” clause was added to the exceptional access policy and was subsequently invoked for Ireland (2010), Portugal (2011), and again in Greece’s 2012 successor arrangement29 In programs outside the Euro area, contagion was not seen as a sufficient concern to warrant using this clause.
45. The systemic exemption proved critical in avoiding defaults on private claims at the outset of the crisis and provided breathing space to build firewalls. But it could not on its own prevent contagion. The counterfactual consequences of an earlier debt restructuring are unknown, and some have argued they could have included pushing the global financial crisis into a deeper, more acute stage. Progress in building firewalls was also critical in allowing the ECB to move against contagion risks within the Euro Area. However, by holding debt restructuring in abeyance the systemic exemption did not by itself foster resumed market confidence. There was a continuing rise in CDS spreads following Euro Area program announcements (Figure 32). Only after two years of heightened uncertainty across the Euro Area did market confidence begin to return, largely in response to the ECB’s commitment to do “whatever it takes” to save the Euro. With little evidence that the systemic exemption approach, taken alone, helped prevent contagion, the Fund is considering proposals to reform its exceptional access lending framework, including the systemic exemption, and increase the general flexibility to deal with cases where debt is deemed sustainable but not with high probability.30
46. Overall, the experience reveals the problems for Fund-supported programs when debt sustainability is not secured upfront. Debt restructuring, when it came, was often too little too late.31 By the time private sector involvement (PSI) was considered in Greece in 2012, the implied haircuts for remaining creditors were large by the standards of other pre-default cases even though they were insufficient to restore debt sustainability with high probability.32 In Jamaica, debt restructuring options were constrained by financial stability considerations, and despite very high initial debt levels, the 2010 and 2013 debt exchanges eschewed principal haircuts, limiting the NPV debt reduction to 15 percent. In Seychelles (2008) and St. Kitts and Nevis (2011), debt restructurings involved sizable haircuts, but in both cases the restructurings were undertaken several years after staff first assessed debt to be unsustainable.
Restructuring vs. Non-Restructuring Programs
Public debt-GDP ratios and growth outcomes in restructuring cases (which all had high initial public debt-GDP ratios) are compared with those in non-restructuring cases where initial debt was high.
Restructuring sample: Seven programs included debt restructurings, either with no face value reductions (i.e. debt reprofilings) or with face value reduction (Table 1).
Non-restructuring sample: top ten initial public debt/GDP cases, namely, Greece 2010 (145 percent of GDP), Seychelles 2008 (136 percent), Portugal (111 percent), Ireland (87 percent), Sri Lanka (86 percent), Jordan (82 percent), Hungary (73 percent), Iceland (67 percent), Pakistan (58 percent), and Maldives (53 percent).
Total eligible debt to be restructured in the debt operation.
Figures do not include past due interest.
Figures 1 and 2 show that relative to non-restructuring cases, the restructuring cases observed better debt-GDP dynamics (debt was put on a declining path by year 4 after the program) and growth turnarounds were stronger. Moreover, outcomes mirrored program expectations more closely in restructuring cases than in non-restructuring cases, where growth and debt both disappointed.
Structural conditionality was mainly focused on core areas of the Fund’s responsibilities, and was generally important to achieve key program goals. An extensive reform agenda was motivated by the need to address underlying macroeconomic imbalances, including through internal devaluation, and to boost long-term growth. Implementation rates of structural conditionality were generally high, although some programs exhibited reform fatigue. The growth payoffs from structural reforms in the short term were likely modest, and less than programs may have envisaged, suggesting a need for program design to be prudent about expectations in this regard. An analytical framework for assessing the prospective payoffs from structural reforms could also help inform expectations. Where reforms are macro-critical but outside the Fund’s traditional areas of responsibility, further consideration is needed on how to effectively leverage expertise in other institutions. In cases where ownership of structural reforms is incomplete, offsetting adjustments in other policies (debt reduction, exchange rate depreciation) may be needed to help address underlying imbalances.
A. Program Goals
47. Structural reforms featured importantly in the recent Fund-supported programs. The average annual number of structural reform conditions (prior actions, structural performance criteria, and structural benchmarks) per program year rose throughout the global crisis (Figure 33).33 By 2013, the average was close to the previous peak in 2003–05. The number of structural reform conditions varied across individual countries and program groupings (Figure 34). The 2011 Conditionality Review noted an increase in the number and depth of structural conditions in Euro Area programs (Greece), and high levels of structural conditionality were also present in some programs outside the Euro Area (Bosnia and Herzegovina, Jamaica, Tunisia, and Ukraine).34
48. The high number of structural reform conditions relative to earlier programs appears to have been warranted in light of the nature of the challenge facing the recent crisis cases.35 Labor and product market reforms were needed to foster price and wage flexibility in order to restore competitiveness through internal devaluation. Supply-side conditions were, accordingly, more numerous in countries with weaker initial structural conditions and more rigid exchange rates (Figure 35).36 In MENA countries, structural reforms tried to address vulnerabilities highlighted during the Arab Spring, especially those linked to inequality, youth unemployment, and social issues. Many programs with the highest structural conditionality were successor arrangements launched later in the crisis period (Bosnia and Herzegovina, 2012; Greece, 2012; and Ukraine, 2010). In these cases, high structural conditionality also reflected lessons from earlier programs and occurred alongside a general effort by the Fund to address issues relating to emerging global challenges facing the membership, such as potential growth, job creation, inequality and other social issues.37
49. Implementation of structural conditionality was generally strong. On average, about 70 percent of structural benchmarks and performance criteria were met without delay (Figure 36). Implementation rates were highest in small states programs. They were lower in the later years of programs, and programs with the largest number of structural conditions, possibly pointing to reform fatigue in some cases (Figure 37). However, there were important exceptions. Some programs with high structural conditionality had very strong implementation rates (Jamaica, 2013), suggesting country ownership and implementation capacity played a critical role in program outcomes. Implementation rates were also generally lower for structural reforms in the financial sector, reflecting specific challenges discussed below.
50. In general, structural conditionality focused on areas within the Fund’s core responsibilities.38 The majority of structural conditionality related to the fiscal and the financial sectors (Figure 38). Fiscal reforms accounted for over half of all structural conditionality in recent crisis programs. Their focus reflected the particular issues facing countries: tax policy and revenue mobilization was most prominent in MENA countries and small states, public financial management in other emerging markets, and a range of issues in Euro Area programs (Figures 39 and 40). Conditionality in other areas mainly covered labor and product market reforms and was more prevalent in countries with rigid exchange rate regimes (Figure 41), including Euro Area programs (Greece, 2012; Portugal), as well as energy sector reform (Pakistan, 2008; Ukraine, 2008 and 2010). In cases where conditionality was applied in areas outside of the Fund’s traditional areas of responsibility, implementation delays were about 20 percent longer, on average. The 2011 Conditionality Review underlined the importance of better scrutinizing the macro-criticality of structural reforms and justifying conditionality in program documents by identifying clear links to program goals, a message that remains relevant.
51. Fund structural conditionality was sometimes accompanied by separate agreements with other official creditors. In Euro Area programs, these agreements were reflected in a Memorandum of Understanding (MOU) between the member and Euro Area institutions as a basis for official financing commitments. The MOU included additional reforms that, together with the Fund’s conditionality, may have increased the strain on the authorities’ implementation capacity.
52. The near-term growth dividend from supply-side structural reforms appears generally to have fallen short of expectations (Annex II). While program documents do not isolate the growth payoff assumed to accrue specifically from structural reforms, programs that included the most supply-side structural conditionality also projected the largest increases in trend growth relative to historical averages (Dominican Republic, Greece, Jamaica (2010), and Portugal).39 These reforms focused on labor and product markets and the business climate.40 The expected medium- to long-term growth dividend from structural reforms assumed in the programs appears broadly in line with empirical evidence. But, for the short term, while the typical view in the literature is that supply side reforms have a very small, possibly even negative, impact on growth,41 in several programs a growth dividend was implicitly expected as early as the second program year. The disappointing growth performance in recent programs cannot, however, be pinned on disappointing structural reform dividends alone as it also reflects factors such as weak global conditions, fiscal consolidation, balance sheet stress, and shrinking bank credit.
Addressing Private Sector Balance Sheet Stress
Programs sought to maintain banks’ liquidity and solvency, and tackle strained private sector balance sheets where these risked holding back the economic recovery. Progress was hampered by a lack of legal and institutional readiness and concerns about potential fiscal costs. Given these considerations, more rapid progress was probably not an option. That said, sustained and proactive steps can help build frameworks to avert the build-up of risks ahead of future crises. Priorities include early attention to legal frameworks and out-of-court settlement options, prudential measures to incentivize debt write-offs and restructuring, and the creation of markets or institutions to handle distressed debts. Steps to address balance sheet data gaps can also help identify vulnerabilities and transmission channels, and inform decisions on the merits of exchange rate depreciation as part of the adjustment process.
A. Program Objectives for Balance Sheet Repair
53. Private sector debt was a concern for many programs, most prominently in Europe (Figure 42, and Annex III). High debt concerns were identified by Fund staff in the corporate and household sectors in several cases (Armenia, Cyprus, Georgia, Hungary, Iceland, Ireland, Latvia, Portugal, Romania, and Ukraine, as well as Jamaica and the Maldives). Household debt-GDP ratios (Cyprus, Greece, Ireland, Iceland, Jordan, Latvia, and Portugal) and non-financial corporate debt-GDP ratios (Cyprus, Iceland, Ireland, and Portugal) exceeded those in Korea and Thailand in the mid-1990s when the latter had high private debt amid credit booms.
54. Excessive private indebtedness can have adverse consequences for domestic demand and economic activity. Where debt service ratios are high in relation to disposable income, households and corporates can see their credit worthiness undercut, and tend to reduce consumption and investment to rebuild balance sheets.42 The recovery in activity was typically slower in cases where financial and non-financial private sector balance sheets were strained, particularly where banks faced high NPLs and limited capital and households and corporates faced elevated debt in relation to GDP (Figure 43). Furthermore, where borrowing is associated with currency or maturity mismatches, it may create vulnerability to shocks.
55. Credit growth tended to be weaker in cases where balance sheets were more strained. Balance sheet problems stemmed from exposure to highly indebted non-financial corporations and households (Ireland, Portugal, and Cyprus) and public sectors (Greece, 2010; Cyprus), banks’ business models (Hungary, Iceland, and Ireland), and protracted weaknesses in economic activity (Greece, 2012). As banks sought to rebuild capital to meet regulatory requirements, fewer resources were available for new lending to businesses and households. Over one-third of programs (12 of 32) experienced a banking crisis, defined by a sharp loss of liquidity due to a deposit run, a failure to rollover wholesale (often external) borrowing, or a large erosion of capital owing to a collapse in asset quality (Figure 44).
56. Adverse feedback loops developed, in some cases, between stressed balance sheets, falling credit, and weak economic activity. Stress in non financial and financial private balance sheets stemmed from various sources, including the post-Lehman sudden stop in capital flows that exposed banking sector business models based on wholesale funding, and the collapse in real estate and other asset prices that reduced household net worth. In several countries weak economic conditions undercut household and business cash flow, reducing credit demand and increasing NPLs, in turn contributing amid the uncertainty associated with the crisis to curtailing the ability of banks to supply new credit, with tighter credit then holding back economic activity further (Cyprus, Hungary, Iceland, Ireland, and Latvia).43
57. Program experience underlined the importance of balance sheet analysis for identifying sources of vulnerability and the transmission of shocks. As noted in the 2014 Triennial Surveillance Review,44 the Fund’s use of the balance sheet approach had fallen into abeyance ahead of the global financial crisis, which may have hindered the detection of risks associated with European banks’ reliance on U.S. wholesale funding. Steps to address balance sheet data gaps could also allow a better informed identification of the scope for external devaluation in future programs, taking into account estimated balance sheet mismatches.
B. Policies to Address Stressed Balance Sheets
58. A key goal for programs was to avert or reverse adverse feedback loops. Specifically, programs focused on addressing bank liquidity and solvency weaknesses, while putting in place measures to facilitate corporate and household debt restructuring, given linkages between financial and non-financial balance sheets. In the event, the programs could not prevent concurrent deleveraging by all sectors. Programs focused initially on liquidity support, liability guarantees, and bank recapitalization with bank resolution, asset restructuring, and insolvency framework measures generally planned for the later stages (Figure 45).
Banking sector support
59. Substantial bank liquidity support helped avoid deposit runs and limit pressures on domestic and foreign exchange liquidity (Figure 46). The significant presence of foreign banks in eastern European countries provided a non-public source of back up capital and liquidity, which in some cases benefitted from international coordination in the context of the Vienna Initiative.45 Foreign banks committed to avoid sudden stops in funding domestic institutions (Bosnia and Herzegovina, Hungary, Romania, Serbia) over the life of the programs, and promised to inject more capital into their subsidiaries to maintain capital adequacy and to cooperate with the domestic regulator on stress tests and action plans. Tight prudential requirements at the outset, as well as stable exchange rates in some cases with high foreign exchange exposure on private balance sheets, helped limit the fallout (Georgia).
60. Euro Area banks benefitted from large liquidity support from the ECB and the Eurosystem in the monetary union context. There were also some specific coordination and program design challenges. Early provision of liquidity and recapitalization of weak banks is critical to avoid major banking crises and limit pressure on domestic and foreign exchange liquidity. However, in the absence of a banking union—a single supervisory-regulatory framework, resolution mechanism, and safety net—to resolve unviable banks and recapitalize systemic ones meant higher borrowing costs for both the sovereign and the private sectors, contributing to amplifying financial market volatility and curtailed bank lending and growth.46
61. Outside the Euro Area, program conditionality focused on dealing with problem banks and improving regulatory and supervisory frameworks (Georgia, Mongolia, and Pakistan). A few programs also sought to strengthen liquidity management practices (Georgia). In non-banking crisis countries, programs included financial conditionality to contain possible adverse impacts of other program components (for instance, from fiscal reforms) and the cycle, and to increase early warning capabilities, including stress tests (St. Kitts and Nevis).
62. Overall, outcomes from financial sector policies were generally mixed. Capital adequacy improved in most program cases (Figure 47). NPLs peaked in about half of the banking crises cases, but they continued to rise in some (Bosnia and Herzegovina, Greece, Hungary, and Portugal). Although credit has started to grow again in a few cases, it continued to fall in most cases during the life of the program, holding back economic recovery (Figure 48). Weak credit reflected weaknesses in bank balance sheets as well as the lingering effect of debt overhang in the corporate and household sector. Furthermore, limited progress was made in implementing structural benchmarks and structural performance criteria related to reforms of the financial system, including the legal framework, capital markets, broker dealers, leasing, and insurance (Dominican Republic, Moldova, and Seychelles), in part owing to implementation capacity limitations.
Private debt restructuring
63. Where high private debt was a concern, program conditionality focused on cleaning up non-financial private balance sheets to help restore creditor solvency and banks’ long-term viability. Conditionality was mainly aimed at establishing or amending insolvency frameworks and facilitating voluntary out-of-court debt restructurings (Figure 49). Among the 12 programs featuring high private debt concerns, 6 programs established frameworks to implement case-by-case debt workouts, while Iceland tried to directly address balance sheet stress through a standardized approach for the write-downs of mortgages and SME debt.47 Capacity constraints in some cases implied longer implementation periods for the restructuring measures.
64. Reflecting wider experience with financial sector structural reforms, measures to tackle private nonfinancial balance sheet stress advanced slowly (Figure 50). On average, about 70 percent of conditionality on private balance sheets was implemented either with delays, or not met, compared with 30 percent for structural conditionality in general. These slippages partly reflected difficulties in getting buy-in from creditors on restructuring programs, lack of political support and delays in advancing legislation to promote private debt restructuring, and difficulties in setting up out-of-court frameworks for resolving private sector debt (Cyprus and Ukraine).48
65. Other factors also hampered efforts to reduce private sector debt ratios. In particular, weak economic growth in programs characterized by tight fiscal policies made it more difficult for the private sector to deleverage (Figures 51 and 52). In addition, write downs of non-financial private debt would have exacerbated banks’ losses and increased the amount of public support needed to recapitalize the banks, running up against fiscal consolidation objectives. In some other programs, public sector restructuring and labor market reforms adversely impacted job security and debt service capacity.
66. Overall, progress in reducing debt outright was limited. Only a few countries with high-or medium-debt concerns achieved outright nominal debt reductions for the private sector (Greece, Hungary, Iceland, Ireland, Latvia, Portugal, Romania, and Ukraine), while a larger number achieved deleveraging when stronger economic activity reduced debt-GDP ratios (Figures 53 and 54).49 While household sectors in most programs saw a decline in debt-GDP ratios, in about half of programs the non-financial corporate sector saw no debt-GDP improvement. In sum, while rapid progress in debt restructuring is clearly difficult, experience suggests the best prospects are provided by action on several fronts. Priorities include: (i) early emphasis on addressing legal deficiencies and enhancing insolvency and debt enforcement frameworks, including by establishing options for out-of-court restructuring; (ii) measures to enhance prudential supervision to incentivize banks to write off or restructure debts, such as through targets for NPL reductions (as under the Cyprus program) or time limits on carrying NPLs; and (iii) establishment of markets or institutions to handle distressed debt, such as the asset management companies established in Ireland and Latvia.
Financial supervision and regulation
67. In several program countries, particularly in Europe, the crisis revealed an excessive buildup of risks in bank balance sheets, often as a result of gaps in supervisory arrangements that did not keep pace with rapid financial integration. In the run-up to the crisis, banks in Europe increased their funding from wholesale markets, including financial centers and off-shore jurisdictions. While the buildup partly owed to asymmetries in financial development and the presence of foreign entities in domestic systems, in some cases it also reflected a relaxation of bank credit standards (Iceland, Romania) and was accompanied by currency mismatches. Over time, the accumulation of foreign liabilities also led to a weakening of the domestic monetary transmission channels, challenges posed by weak subsidiaries of parent banks from crisis countries (Greek and Cypriot banks), credit market fragmentation as private and sovereign funding costs diverged, and a worsening of the financial-fiscal linkage. Eventually, improved supervision and regulation led to the banks’ gradual deleveraging from foreign creditors, and to a credit contraction.
68. Fund-supported programs were accompanied by increased scrutiny of balance sheets through asset quality reviews and periodic stress testing at the national and regional levels, with a view to preserve solvency and viability of the financial systems. Programs also incorporated micro prudential measures such as provisioning requirements and limits on foreign exchange exposure to unhedged borrowers to mitigate risks in bank balance sheets (Hungary, 2008; Romania, 2009).
69. In general, Fund-supported programs did not include the adoption of new macroprudential measures. A few countries adopted macroprudential regulatory measures ahead of Fund-supported programs (Armenia and Romania), as these measures are typically preventive in nature and adoption during a crisis risks intensifying an already sharp slowdown in credit. Where measures were included in Fund-supported programs, they were typically designed to minimize exchange rate risk—for example, by increasing provisioning and risk weighting for bank’s foreign currency loans, raising reserve requirements for foreign liabilities, and establishing tighter limits on net foreign currency open positions (Armenia, Belarus, and Hungary). In Ireland’s program, liquidity assessments included a review of loan-to-value levels, and in Kosovo’s program, the establishment of a macroprudential committee was required, to help coordinate and communicate existing macroprudential policies.
Regional Financing Arrangements and Currency Unions: Issues for Program Design
In programs that involve co-financing by RFAs, program design needs to ensure clarity regarding the roles of various institutions and to take account of the cumulative extent of conditionality in the context of the authorities’ implementation capacity. The G-20 principles on Fund-RFA cooperation provide a helpful foundation for developing more operational guidance toward ensuring clear roles for various partners. For arrangements with members of a currency union, program design needs to deal with the possibility that union-wide policies can have a critical bearing on the member’s economic situation. Where changes in currency-union-wide policies are important for program success, the Fund should provide advice through its surveillance as warranted or, when necessary (including for financing assurances), seek commitments on prospective implementation of necessary union-wide policies; alternatively, program design may need to build in greater adjustment and financing, or Fund involvement be postponed.
A. Co-financing with RFAs
70. With the role of RFAs in the global economic and financial system set to grow, a question for the Fund is whether the recent crisis programs experience provides lessons for future Fund-RFA interaction. RFAs are playing an expanded role in many regions, including Europe, Asia and the Commonwealth of Independent States, and are an important complement to the Fund in the global financial safety net. They can bring valuable understanding of regional policy challenges and foster increased ownership of adjustment programs.50
71. In the recent Euro Area programs, financing was provided through arrangements from the IMF and a new RFA established in response to the crisis, the European Financial Stability Facility (EFSF) or its successor, the European Stability Mechanism (ESM).51 The collaboration between the IMF, EC, and ECB (the so-called “Troika”) extended beyond financing to include program design. While the Fund alone determined whether the necessary conditions to complete a program review were met, the continued support of the EC and ECB was relevant for the Fund’s assessment in that it provided needed financing assurances by facilitating EFSF/ESM financing. Effectively, therefore, reviews under the IMF and RFA arrangements depended on agreement among the three institutions on the authorities’ performance and on reaching understandings with the authorities on objectives and policies.
72. The G-20 endorsed six non-binding principles for IMF-RFA collaboration in 2011 that provide a foundation upon which to build clearer operational guidance. These principles encourage ongoing and early cooperation; the recognition of areas of comparative advantage; the importance of consistent lending conditions; the need for flexibility in adjusting conditionality and the timing of reviews; the need to respect the roles, independence, and decision-making processes of each institution; and the need to respect the Fund’s preferred creditor status (Box 4). More operational guidance on IMF-RFA collaboration could help clarify the respective roles of the various institutions, such as the Fund’s responsibility for assessing the macroeconomic framework and use of related tools such as debt sustainability analysis, as well as ensure the cumulative conditionality faced by the member is well aligned with implementation capacity. Understandings the Fund currently has in place with multilateral development banks, including the World Bank, have helped strengthen collaboration with those institutions. In any such interactions, IMF involvement needs always to adhere to the Fund’s mandates, rules, policies and procedures.
B. Program Design in Currency Unions
73. Program design concerning members who belong to a currency union faces several considerations. Membership of a currency union can bring important benefits, including gains associated with closer and deeper economic integration, a credible nominal anchor, and facilities such as a regional backstop for sovereign financing. At the same time, union-wide policies such as monetary, exchange rate, and financial policies that are under the effective control of supranational authorities may be important for addressing the underlying economic imbalances in the member that the program seeks to remedy.52 In addition, program design may need to take account of member’s commitments to other union-wide policies.
74. The Euro Area program experience shows the architecture of the currency union has important implications for crisis prevention and resolution. Fiscal union and consolidated supervision and resolution are important to achieve ex-post risk sharing and to deal with the cross border nature of vulnerabilities, respectively. A broad set of liquidity tools is important to address market dysfunction, including in the form of market runs on sovereign issuers and delays or avoidance of necessary debt restructuring, as well as to allow union-level institutions directly to recapitalize banks.
75. There are several approaches for the Fund to designing programs with members in a currency union. Conditions on union-wide policies are consistent with the Fund’s Articles of Agreement. 53They are, however, difficult to implement in practice as policy changes at the union level that may be desirable from the point of view of a particular member may not be so for others in the union, particularly if spillovers from the member to others are not considered to be systemic. More generally, union-wide policies can be hard to change quickly as they can involve complex decision-making procedures and multiple countries. In these circumstances, the Fund can take union-wide policies as given and focus instead on policies effectively controlled at the national level, such as fiscal and structural policies, although such an approach could increase the required program adjustment effort and financing needs. Alternatively, the Fund could seek commitments from union-wide institutions with respect to changes in union-level policies that may be necessary for the success of the member’s program (Box 5); it should also provide such advice through surveillance where warranted. If these approaches prove unworkable, it may be necessary to postpone Fund support until staff can give the Board an assurance that the relevant problems are being adequately addressed.
G-20 Principles for Cooperation Between the IMF and RFAs
Six non-binding broad principles for cooperation were agreed, and endorsed by G-20 Leaders in November 2011. The preamble to the Principles states that collaboration with the IMF should be tailored to each RFA in a flexible manner in order to take account of region-specific circumstances and the characteristics of RFAs.
(1) An enhanced cooperation between RFAs and the IMF would be a step forward toward better crisis prevention, more effective crisis resolution and would reduce moral hazard. Cooperation between RFAs and the IMF should foster rigorous and even-handed surveillance and promote the common goals of regional and global financial and monetary stability.
(2) Cooperation should respect the roles, independence and decision-making processes of each institution, taking into account regional specificities in a flexible manner.
(3) While cooperation between RFAs and the IMF may be triggered by a crisis, ongoing collaboration should be promoted as a way to build regional capacity for crisis prevention.
(4) Cooperation should commence as early as possible and include open sharing of information and joint missions where necessary. It is clear that each institution has com parative advantages and would benefit from the expertise of the other. Specifically, RFAs have better understanding of regional circumstances and the IMF has a greater global surveillance capacity.
(5) Consistency of lending conditions should be sought to the extent possible, in order to prevent arbitrage and facility shopping, in particular as concerns policy conditions and facility pricing. However, some flexibility would be needed as regards adjustments to conditionality, if necessary, and on the timing of reviews. In addition, definitive decisions about financial assistance within a joint program should be taken by the respective institutions participating in the program.
(6) RFAs must respect the preferred creditor status of the IMF.
Aligning Union-wide Policies to The Needs of Fund-Supported Programs
Conditionality in Euro Area programs was limited to policies under the effective control of the member. Accordingly, performance criteria and structural benchmarks focused on fiscal, structural, and financial issues.54 In addition, the Fund used two approaches to get changes in policies at the union level. First, the Fund sought formal commitments as in the case of financing assurances from the euro group. Second, at the same time, the Fund sought an adjustment in policies by union wide-institutions in the context of regular surveillance engagement. For example, the Fund in the context of regular surveillance sought looser monetary policy, relaxed collateral requirements for liquidity facilities, and measures to strengthen the union’s institutions and architecture, for example by establishing a banking union and putting in place a common fiscal backstop (IMF, 2012a). In advocating for these reforms, the Fund viewed the policy changes as appropriate both for the success of Fund-supported programs in the Euro Area as well as to address underlying vulnerabilities within the Euro Area as a whole.
Over time, Euro Area policymakers made substantial changes to the policy mix and the union architecture, radically improving its crisis management ability. Important progress was made with respect to the stance of monetary policy (including quantitative easing), emergency liquidity provision, balance of payments support, the framework for fiscal governance, and banking union. These changes were critical to ameliorating undue pressures on sovereign financing, enhancing fiscal discipline, and mitigating the contingent liabilities of governments under stress. Moreover, in the case of Cyprus, the enhanced supranational firewall allowed for deeper restructuring without fear of contagion. Europe is now in a far stronger position to prevent and address future imbalances.
Eastern Caribbean Currency Union (ECCU)
In the context of stand-by arrangements with Antigua and Barbuda (2010) and with St. Kitts and Nevis (2011), programs included structural benchmarks (Antigua and Barbuda) and a prior action (St. Kitts and Nevis), both within the competence of the Eastern Caribbean Central Bank (ECCB). This conditionality was supported by the ECCB. In the case of Antigua and Barbuda, the ECCB provided formal written assurances on the structural benchmarks that were circulated to the Executive Board. In the case of St. Kitts and Nevis, the prior action took the form of an agreement between the ECCB and St. Kitts and Nevis authorities on commercial banking issues, with implementation also reported to the Executive Board.55 In both country cases, the measures were judged not to have broader implications for the currency union as a whole.