A. Onset of crisis in Emerging Markets

1. The deepest global financial crisis of the post-war era began in the advanced economies, and spread to emerging markets only with a lag. Emerging markets became engulfed in the crisis when the extent of the damage to institutions in financial centers became apparent (Blanchard, 2009): until Lehman’s bankruptcy in September 2008, emerging markets asset prices appeared to have decoupled from developments in advanced economies (Figure 1). The sudden loss of confidence and the rise in counterparty risk triggered by Lehman’s bankruptcy led to a sudden “lock-up” in financial markets.

2. The crisis was manifested in emerging markets as a sudden stop in capital inflows, with added strains from the collapse in global activity and commodity prices. Central and Eastern European (CEE) countries and the CIS region, which had benefited from large-scale capital inflows during the global upturn, became the epicenter of the emerging market crisis when advanced country banks begun cutting back exposures. In the uncertain times in the aftermath of Lehman’s bankruptcy, global industrial production declined over 20 percent and exports plummeted by over 40 percent causing a “recoupling” of emerging market countries with advanced economies. The precipitous fall of commodity prices during the second half of 2008 provided a further drag to commodity exporters.

View Full Size

Bank flows to emerging markets

External positions of BIS reporting banks

(Estimated exchange rate adjusted change)

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Sources: BIS and staff calculations.

• Download Figure
• Download figure as PowerPoint slide

3. The global crisis has greatly increased the demand for Fund resources. In response to the crisis, the Fund moved quickly to expand its lending capacity, increase the flexibility of its lending toolkit, and provide financing, with some $163 billion in commitments made to date (Box 1). Compared to past episodes, the current arrangements are notable for their size, severity of projected output contraction, and simultaneity (Figure 2).

4 This review covers 15 Fund-supported programs approved since September 2008 (Appendix I). Hungary, Iceland, Latvia, Pakistan, and Ukraine were among the first wave of countries to adopt Fund-supported programs in the immediate wake of the global crisis in late 2008, along with Georgia, which had turned to the Fund before being swept up by the crisis. In the early months of 2009, large, front-loaded programs were also approved for Armenia, Belarus, Mongolia, and Romania; and precautionary programs put in place for Costa Rica, El Salvador, Guatemala, and Serbia—the last quickly being augmented and made nonprecautionary. An arrangement for Bosnia & Herzegovina was approved in July. In addition, Mexico, Poland, and Colombia have availed themselves of the new FCL (Box 2).

B. What led countries to request Fund support?

5. In the years leading up to the crisis, emerging market economies as a whole registered an impressive growth performance. During 2003–07, their median growth rate was about 6 percent per year. The combination of high growth with stronger institutions and better policies in most cases resulted in notable drops in external and public debt ratios and the halving of government deficits to 1¼ percent of GDP by 2007.

6. However, the boom carried the seeds of vulnerabilities. Fast growth widened the median output gap—as measured in a consistent methodology adopted for the paper—to over 2½ percent of potential output by end–2007. With large capital inflows fuelling credit booms, most notably in the Baltics, many emerging market economies also entered this period with much overheated real estate markets. The wider overheating in the economy translated into appreciating currencies and widening current account deficits, with cross-country evidence indicating a statistically significant relationship between the output gap and the current account deficit. Furthermore, with an increasing proportion of this deficit financed by debt-creating (non-FDI) inflows, the external debt-to-GDP ratio of program countries declined little (or rose in some cases), in spite of high growth.

View Full Size

Overheating in emerging markets, 2007

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: WEO and staff calculations.

• Download Figure
• Download figure as PowerPoint slide

7 These vulnerabilities were widely recognized in advance by market and other analysts. Within the Fund, the confidential internal staff Vulnerability Exercise for Emerging Market Economies (VEE), for example, highlighted weaknesses in sectoral fundamentals of many emerging market countries (and especially in the European emerging economies).¹ Figure 3 shows the sectoral distribution of vulnerabilities in the September 2007 exercise, a full year ahead of the crisis. Several factors stand out:

• All the new program cases were then seen as having a medium or high external vulnerability. As external vulnerabilities are those manifested in balance of payments pressures, it seems natural that these members may approach the Fund for assistance.

• In fact, all members requesting Fund-supported programs had vulnerabilities, in addition to the external sector, in another sector (fiscal or financial), while countries which were identified as having fiscal or financial vulnerabilities but not external vulnerabilities have not approached the Fund for a program. These combination effects underscore how balance sheet weaknesses in one sector may get amplified by weaknesses in other sectors, making crisis more likely when the country is hit by shocks (as discussed in Ghosh et al., 2008).

8. An examination of various macroeconomic variables also underscores the role of external imbalances in driving countries to turn to the Fund (Figure 4).

• External imbalances are statistically significant predictors of crisis: program countries have a 6 percentage points of GDP higher current account deficit than nonprogram countries in 2007. These deficits were also much higher than in previous crises. Differences in total external debt, however, are more subdued and not significant.

• Reserves relative to near-term financing needs also played a role, with reserve coverage in 2007 significantly lower in program countries than in those without programs, and comparable to those observed in previous capital account crises.

• Credit growth is higher in program countries relative to nonprogram cases (and to previous capital account crises), but variance also tends to be high. Credit booms help explain program participation when interacted with external imbalances.

• Other initial conditions (GDP growth, inflation, the government balance, and public debt) were not significantly different between program and nonprogram countries.

C. Recent developments in program countries

9. In both the depth of contractions and revisions to the outlook, CEE and CIS economies stand out. As the realization of underlying vulnerabilities prompts members to turn to the Fund for financial support, lower growth in program countries should come as no surprise. While the projected average real growth decline for 2009 in countries with Fund-supported programs is around 5½ percent, with the comparable number for other emerging markets being about 1½ percent, this is driven mostly by CEE and CIS economies. Controlling for initial conditions— with worse external imbalances being critical—program participation is not associated with a worse growth outturn. In terms of downward revisions to growth (Figure 5), too, CEE and CIS countries are outliers—with average downward revisions between October 2008 and now of about 12 percentage points of GDP. Within the regions, there is no statistical difference in the revisions between program and nonprogram countries.

View Full Size

Growth outturns and initial vulnerabilities

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

1/ The fitted regression on real GDP includes current account deficit, fiscal deficit, external debt, public debt, and reserves (all actuals, as of 2007).

• Download Figure
• Download figure as PowerPoint slide

10. The composition of the output declines in program countries in 2009 is markedly different from that observed in past crises (Figure 6). On the demand side, current programs have been characterized by sharp declines in domestic demand, although to a considerably lesser extent than past crisis cases. But, while in the past the sharp drop in domestic demand has been partly offset by improvements in exports, the turnaround in net exports in the current cases has been exclusively driven by a dramatic import compression: exports are also declining against the backdrop of very subdued global demand, adding to the effects of the sizable drop in domestic demand. On the supply side, preliminary data for the first quarter show the bursting of the pre-crisis boom in the nontradable sector, notably services and construction, which had contributed to overheated economies, loss of competitiveness, and a buildup of external imbalances in many current programs countries.

View Full Size

Contributions to real GDP growth ^1/

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: WEO, Haver.^1/ Medians within groups.2/ Tradables comprise agriculture, manufacturing, and mining sectors; non-tradables comprise services and construction sectors.

• Download Figure
• Download figure as PowerPoint slide

11. Given the global deflationary forces, most current Fund-supported programs face a benign inflation outlook. For emerging markets as a whole, median inflation is estimated to decline five percentage points to 4½ percent in 2009. Among program cases, estimated median inflation for 2009 is a little higher, at around 7½ percent. This stands in stark contrast to the experience of past capital account crises, which were marked by a spike in inflation in the immediate aftermath of the crisis due to large currency depreciation.

12. The falls in output have prompted an important reconsideration of potential output during and after the boom years. In retrospect, the years 2005–07 involved even more “froth” of excess growth above the economies’ underlying potential than was recognized at the time. This has important implications for fiscal policy—with past apparently comfortable fiscal positions incorporating large cyclical elements masking more serious structural deficits—as well as the prospects for growth in the medium term as countries seek to exit from Fund support.

A. Overall objectives and crisis response

13. Program objectives have been tailored to country-specific circumstances, with most focusing on mitigating external and financial sector pressures (Table 1). With countries generally facing large external imbalances, particularly in Eastern Europe, key objectives revolved around smoothing current account adjustments and mitigating liquidity pressures, while preserving market confidence by addressing underlying vulnerabilities over time. Avoiding systemic banking crises, which featured prominently in past crises, or restoring bank solvency where banking crises were underway, were also key program goals. Finally, where weak structural fiscal positions called for adjustment, excessive frontloading of measures that would have exacerbated the economic downturn was avoided, with market confidence secured via medium-term fiscal consolidation plans backed by structural reforms.

Table 1.

Initial stated program objectives

Sources: SBA request documents.

Table 1.

Initial stated program objectives

	Macro-economic stability/ adjustment	Crisis response/ preparedness	Adequate financing/ reserves, confidence in currency	Financial sector stability	Fiscal sustainability/ adjustment
Armenia	•		•	•
Belarus	•		•
Bosnia			•	•	•
Costa Rica	•
El Salvador		•		•	•
Georgia			•
Guatemala	•	•
Hungary			•	•	•
Iceland			•	•	•
Latvia	•		•	•
Mongolia	•	•
Pakistan	•		•		•
Romania	•		•	•	•
Serbia	•	•		•	•
Ukraine	•			•

Sources: SBA request documents.

View Table

Table 1.

Initial stated program objectives

	Macro-economic stability/ adjustment	Crisis response/ preparedness	Adequate financing/ reserves, confidence in currency	Financial sector stability	Fiscal sustainability/ adjustment
Armenia	•		•	•
Belarus	•		•
Bosnia			•	•	•
Costa Rica	•
El Salvador		•		•	•
Georgia			•
Guatemala	•	•
Hungary			•	•	•
Iceland			•	•	•
Latvia	•		•	•
Mongolia	•	•
Pakistan	•		•		•
Romania	•		•	•	•
Serbia	•	•		•	•
Ukraine	•			•

Sources: SBA request documents.

14. As the crisis tightened financing constraints, it became essential to shift the focus of policies and Fund resources toward the budget. As programs were designed amid an unusual degree of global economic uncertainty, when initial growth assumptions proved too optimistic, program policies were adapted. The forcing event was the harsh contraction in global economic activity, which quickly translated into sharply falling revenues in most countries, while widening financing needs were made acute by the drying up of funding opportunities. With weak external demand constraining export-led recoveries (a key driver in past crises) and the credit channel operating weakly on account of disrupted credit markets, programs shifted their focus—debt sustainability considerations permitting—to fiscal policy as the main countercyclical instrument supported by a more direct use of Fund resources for budget financing (Box 3).

15. The Fund was able to mobilize very quickly financial support to countries hit by the financial turbulence of late 2008. By activating fast-track procedures under the Emergency Financing Mechanism, the Fund was able to field missions within days and approve exceptional access arrangements within 3½–6 weeks of initial indication of interest by authorities in Hungary, Iceland, Latvia, Pakistan, and Ukraine, despite the challenging task of reaching understandings on difficult policy measures in a short period of time. The mechanism was also used for Georgia’s and Armenia’s recent SBA augmentations.

B. Fund financing and burden sharing with other creditors

16. Almost all arrangements entailed exceptional access—beyond the usual limits— to Fund resources, with frontloaded disbursements. Widespread recourse to exceptional access (in 14 of 15 SBAs) reflects the severity of the global crisis and the large financing gaps. While there is much variability, access in current programs (with a median of 7 percent of GDP) has been higher than in past capital account crises (4 percent of GDP), although similar in terms of the share of gross financing needs (reflecting much larger needs this time around), and slightly lower in terms of members’ quotas (Figure 7). The degree of frontloading has been higher than previous crises, despite larger initial reserve buffers, consistent with recent research findings on the role of financing in mitigating crisis effects.

View Full Size

Access, phasing and duration in current and past capital account crises

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: WEO and staff estimates.

• Download Figure
• Download figure as PowerPoint slide

17. Financing packages have included support from other official creditors, enabling risk sharing. Of $133 billion in aggregate financing packages, the Fund has committed $75 billion, the EU $21 billion (with especially large contributions to Latvia, 42 percent of the total financing package; Hungary, 33 percent; and Romania, 25 percent), the World Bank $11 billion, and other bilateral creditors some $26 billion. In the last group, European member states with close ties to crisis countries have pledged sizable contributions (reaching 31 percent in Latvia and 80 percent in Iceland, including loans tied to the repayment of deposit insurance liabilities), although, in other cases, burden-sharing with official creditors has been much more limited. While there is significant variation—from 21 percent in Iceland to 84 percent in Serbia—on average, the Fund’s share of financing packages (56 percent) is considerably higher than in past crises (40 percent). (See Appendix I for financing details.)

18. Private sector involvement has also been sought in a number of European programs both informally and through the Bank Coordination Initiative. The latter is a platform involving host governments, national supervisory authorities, banks, and official agencies, and responds to the need for coordination in a highly integrated European financial system (Box 10). Through this initiative—adopted so far in Bosnia & Herzegovina, Hungary, Romania, and Serbia—European parent banks agree to maintain exposure and, if necessary, recapitalize subsidiaries, thus mitigating concerns about the potential effects of deleveraging.

Box 3.

Use of Fund resources for budgetary financing

The current crisis has called for flexible fiscal responses where warranted by debt sustainability considerations. In a global deleveraging environment and with increased country emphasis on (de jure) central bank independence, funding such fiscal accommodation has in some cases required directly tapping official resources.

View Full Size

Direct Budget Support in Current SBAs disbursed amounts, in percent of quota

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: Staff estimates.

• Download Figure
• Download figure as PowerPoint slide

Legal basis. The Fund’s legal framework allows using Fund resources for direct budget support. This requires, as prescribed in the Articles, that (i) the member country has an actual balance of payments (BoP) need when making a purchase (requiring either an above-the-line BoP deficit or inadequate reserves); and (ii) there are adequate safeguards by virtue of policies that address the BoP problem and ensure timely repayments to the Fund. From an economic perspective, BoP needs often coincide with fiscal needs. As seen in this and past crises (Russia (1993), Argentina (2000), and Turkey (2001) when the Fund provided direct budget support), countries generally face external financing gaps together with fiscal financing pressures from collapsing revenues (Appendix II).

Economic rationale. Fund financing enables implementation of a countercyclical policy mix that avoids excessive external adjustment. The customary role of central banks in making Fund purchases is closely related to their ability to distribute balance of payments support to the economy. However, there could be situations when liquidity does not find its way to where the need is the greatest (say, the government), because of either restrictions on the central bank’s ability to on-lend resources (as would be the case when central banks are independent or in monetary unions, currency boards, and fully dollarized economies) or because of impaired market access by sectors of the economy. In such cases, it would be appropriate to channel Fund resources directly to the government, when restoring domestic and external stability in the context of a balance of payments crisis calls for a larger fiscal deficit than could be financed from external or domestic sources.

Risks to the Fund. Use of Fund resources for budgetary purposes carries risks. First, repurchases could become subject to budgetary appropriation processes and government’s ability to either generate a budget surplus or borrow from other sources. Second, fiscal policy could become unduly reliant on Fund resources to finance what may be permanent expenditures. Third, there are risks related to the potential misuse of Fund resources, as the main focus of the current Safeguards Assessments framework remains on central banks.

These concerns could be mitigated in different ways. The Fund could, for instance, require that (i) the government builds up minimum levels of cash deposits through overborrowing; (ii) institutional arrangements be put in place to clarify the respective responsibility of the central bank and treasury to service the liability to the Fund; (iii) the members’ SDR accounts be used as a quasi-escrow account; and (iv) borrowing from the Fund and its subsequent use is reported in the accounts of the central bank. Ultimately, well-designed and fully-owned programs provide the strongest safeguard of exit from and appropriate use of Fund resources.

C. Financing versus adjustment

19. The appropriate mix of external financing and macroeconomic adjustment is a key design issue in IMF-supported programs. In general, the greater the financing, the smaller the required short-run correction in the current account and vice versa (although this is not a simple dollar for dollar relationship, since capital account flows react endogenously to the strength of policies and financing). For the current program countries, lack of timely access to official financing would have forced even more painful and disorderly demand contractions. At the same time, access to official financing should reflect, not only short-run demand management considerations, but also a medium-term perspective on external viability. This perspective is particularly relevant for the current program countries, which entered the crisis with very large current account deficits and are now experiencing sharp increases in external indebtedness that will need to be unwound through adjustment efforts.

20. Mirroring larger than expected economic contractions, current accounts are projected to adjust markedly in 2009, and by more than initially programmed. Initial program design aimed for more gradual adjustment. In the event, imports fell more than expected (and by a similar amount to previous crises), in line with domestic demand trends. This led to a sharp turnaround in current accounts, despite larger-than-expected falls in exports—in sharp contrast to previous cases, which featured continuing strong export growth. Latest forecasts show median current account adjustments for 2009 of some 6 percentage points of GDP, although adjustments in some countries (Iceland and Latvia) are projected to be much larger (Figure 8).

View Full Size

External adjustment

(t=2008 for current cases)

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: Staff estimates.

• Download Figure
• Download figure as PowerPoint slide

21. However, the degree of adjustment taking place in 2009 is still much less than in past cases and not extreme when set against pre-existing external imbalances. On average, current account adjustments in 2009 are projected mainly to unwind the deterioration observed in 2005–07, and accentuated in 2008, when an already evident slowdown in exports combined with still buoyant imports and worsening terms of trade. Thus, the adjustments between 2005–07 and 2009–10 are indeed much smaller than in previous crisis cases. Furthermore, most programs envisage current accounts balances to remain below debt-stabilizing levels in 2009, allowing countries to further accumulate external debt to smooth the adjustment over the medium term (Figure 9). Programs that envisage adjustment beyond debt-stabilizing levels (Hungary, Iceland, Latvia, and Ukraine) aim at bringing down high initial debt levels. This contrasts markedly with previous capital account crises, which displayed current account balances significantly above debt-stabilizing levels even for countries with relatively low external debt levels.

View Full Size

Current account adjustment and external sustainability

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: WEO and staff estimates.

• Download Figure
• Download figure as PowerPoint slide

22. In addition, current programs have not entailed systematically greater external adjustment than nonprogram emerging market countries. Cross-country regressions suggest that changes in current account positions observed in a large sample of emerging market countries can be largely explained by the initial external position, with a program dummy showing as statistically insignificant (Box 4).

Box 4.

Initial conditions and external adjustment

A simple statistical relationship is estimated to quantify the role of initial external positions in driving the degree of adjustment in emerging market economies. The change in the current account from 2007 to 2009 for a sample of program and nonprogram emerging market countries is regressed against the initial external debt and current account levels and a program dummy.

The results show that initial conditions explain a significant share of the cross-country variance, with higher external debt and deficits associated with larger adjustments. Importantly, current programs do not show larger adjustments than nonprogram emerging markets. Dummies for pegged exchange rates or commodity exporters are also insignificant.

Among the program countries, only Iceland and Latvia show adjustments significantly above those predicted by the model, possibly reflecting omitted variables or nonlinear relations for these highly indebted countries. Conversely, Armenia, Mongolia, Belarus, and Georgia stand out as adjusting less than their comparators.

Dependent Variable: Change in current account (from t-1 to t+1)

Dependent Variable: Change in current account (from t-1 to t+1)

	Coeff.	S.E.	t Stat	P-value	Multiple R	0.75
Intercept	-3.83	1.18	-3.25	0.00	R Square	0.57
External debt average (t-1, t)	0.06	0.02	2.74	0.01	Ad. R Square	0.54
Current account t-1	-0.43	0.09	-4.56	0.00	Standard Error	4.70
Program dummy	0.52	1.52	0.34	0.73	Observations	55

View Table

Dependent Variable: Change in current account (from t-1 to t+1)

	Coeff.	S.E.	t Stat	P-value	Multiple R	0.75
Intercept	-3.83	1.18	-3.25	0.00	R Square	0.57
External debt average (t-1, t)	0.06	0.02	2.74	0.01	Ad. R Square	0.54
Current account t-1	-0.43	0.09	-4.56	0.00	Standard Error	4.70
Program dummy	0.52	1.52	0.34	0.73	Observations	55

View Full Size

Fitted versus actual

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: WEO and staff estimates.

• Download Figure
• Download figure as PowerPoint slide

23. Nor does a lack of official financing appear to be forcing greater adjustment. Many programs envisage relatively low current account corrections despite projecting low or negative private capital flows, as use of reserves and official financing are allowed to smoothen adjustment. This pattern is particularly pronounced in the cases of Iceland, Latvia, Romania, and Ukraine, which are facing large private capital outflows or large current account deficits. These results contrast with previous crises, many of which involved net reserve accumulation in the first program year, as large current account surpluses more than offset reversals in private capital flows.

View Full Size

Use of reserves and official financing

(at t+1, percent of GDP)

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: WEO and staff estimates.1/ Defined as net official financing plus decumulation of gross international reserves. Includes 2009 general and special SDR allocation.

• Download Figure
• Download figure as PowerPoint slide

D. Program conditionality

24. IMF financing is generally tied to policy conditions to ensure predictable access to financing, buttress policy credibility, and reduce repayment risks. Conditionality includes quantitative targets (performance criteria) on key policy variables (fiscal balance, international reserves, monetary aggregates). Structural policy measures (benchmarks) complement quantitative policy targets in ensuring that program objectives are attained.

25. Current SBAs carry fewer structural conditions than previous arrangements. Conditionality in past programs did not always focus on macro-critical policies, which had the effect of deterring countries from approaching the IMF early on, amplifying crisis effects. Consistent with the spirit of recent conditionality reforms, current arrangements have fewer structural conditions than earlier nonconcessional arrangements.² At the same time, there has been a marked increase in the share of financial sector conditions, and a very sharp fall in measures outside the key areas of Fund competency. Differences across countries are, however, considerable, reflecting heterogeneous starting conditions and the need to tailor the structural reform agenda to each country’s problems.

View Full Size

Average number of structural conditions per program year

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

1/ “Application of Structural Conditionality-2008 Annual Report,” IMF. Consistent breakdown unavailable.2/ Covers all nonconcessional arrangements (32 in total) approved during 2002–07.3/ Cumulative number of conditions in current programs as of August 2009, divided by years covered by conditionality.

• Download Figure
• Download figure as PowerPoint slide

26. However structural conditions typically rise over time as crises deepen and vulnerabilities shift—so it remains to be seen if the recent parsimony holds. Although still lower overall than in recent programs, conditionality has increased with program reviews (Figure 10). For Hungary and Ukraine, the initial programs were negotiated under time pressure, with the presumption that they would be modified as the crisis evolved. Also, in the case of Ukraine, the added structural conditions (with large reliance on prior actions) were needed to provide assurances of program implementation as political fragmentation raised questions about program ownership. For Armenia and Serbia, economic conditions deteriorated more than expected at the outset, requiring program policy adjustments.

View Full Size

Cumulative number of structural conditions

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: MONA database and staff calculations.

• Download Figure
• Download figure as PowerPoint slide

E. Ownership and review progress

27. Generally, timely completion of program reviews suggests good country ownership of programs (Figure 11). As of end–August, quantitative and structural conditions have largely been met—albeit with some delays—in most cases. Among the 19 program reviews so far (including those taking place in September), only six have requested waivers for missed quantitative performance criteria, and only for a small proportion of indicators. A recent opinion survey of the Fund’s role in selected program countries also supports the perception of improved country ownership, among other findings (Box 5).

View Full Size

Program approval and review timelines

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: Program documents.1/ Review scheduled for Board consideration in September.

• Download Figure
• Download figure as PowerPoint slide

28. However, three programs suffered delays of over three months in completing reviews. In Iceland, completion of the first review has been delayed by the collapse of the government in February, which complicated achieving consensus for key fiscal and bank reforms, and by difficulties in securing external financing assurances. Similarly, in Latvia, difficult negotiations on fiscal targets against the backdrop of a dramatic economic contraction and a government changeover led to a long delay in completing the first review. In El Salvador, political transition hindered the completion of the first review, while deteriorating economic conditions have shifted the authorities’ interest toward a new SBA. More generally, political instability has affected program implementation in many cases, with fragile government coalitions or changes in government since the crisis in many cases.

A. Fiscal policy in the crisis

29. The question of the appropriate fiscal stance has been central in the recent programs. Faced with severe recession and limitations to monetary policy effectiveness, fiscal policy in program countries has, broadly speaking, sought to cushion the effect of the cycle in the short run, while ensuring sustainable long-run fiscal positions backed by structural reforms. But the tight constraints facing all emerging market countries (including curtailed financing, debt intolerance, and institutional factors such as EU-wide policy strictures) have prevented the massive increases in deficits and debt taking place in the largest industrialized economies (Figure 12). And a revised view of potential output has tempered the scope for accommodation in some cases.

View Full Size

Fiscal deficits and debt in advanced and emerging market economies

(medians, percent of GDP)

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: Staff projections

• Download Figure
• Download figure as PowerPoint slide

B. Fiscal developments leading up to the crisis

30. Ahead of the crisis, the fiscal position of many emerging market countries strengthened, but the improvement was flattered by above-trend growth (Figure 13). Debt in the program country group remained lower than in nonprogram and past crisis countries, and fiscal deficits were modest. However, most current program countries were running pro-cyclical policies, with exceptionally—and unsustainably—high revenue growth setting off even faster increases in primary spending. Thus, generally benign developments in headline fiscal deficits masked a steady deterioration in the structural fiscal position.

View Full Size

Fiscal indicators in program, nonprogram and past crisis countries

(medians)

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: Staff projections.1/ t = 2008 for current cases. Real primary revenues and expenditures are computed using the GDP deflator, and are shown as the change from the average level for t-4 to t-1.

• Download Figure
• Download figure as PowerPoint slide

C. Overall fiscal stance in program countries

31. Fiscal policy has adapted to deteriorating conditions, with most programs now showing net fiscal accommodation—i.e., rising overall or primary deficits (Figure 14). Repeated downward revisions to growth projections have resulted in increasingly negative output gaps in all program countries in 2009. In response, fiscal deficits for 2009 have been revised upwards and now range between zero and 13 percent of GDP.

View Full Size

Revisions to overall fiscal balance projections (and program targets) for 2009

(percent of GDP, ranked by size of deficit projection)

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: Staff estimates.

• Download Figure
• Download figure as PowerPoint slide

32. Real revenues are projected to fall sharply, and by more than in past crises (Figure 15). Revenues are falling faster than GDP in most program and nonprogram countries. These declines can be only partly explained by tax policy changes, with net tax reductions limited to Armenia, Belarus, and Georgia, and small net tax increases in Bosnia & Herzegovina, Iceland, Latvia, Pakistan, and Romania. Other, more significant, sources of revenue weakness include falling imports, declining asset prices, and weak tax compliance.

View Full Size

Contributions to change in primary balance, 2009

(percent of GDP)

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: Staff estimates.Note: Components may not add to total due to statistical discrepancy.

• Download Figure
• Download figure as PowerPoint slide

33. Program countries show larger real spending cuts than nonprogram countries. This reduction in spending is smaller than that seen in past crises, and follows much larger pre-crisis expansions (including in wages, pensions, and subsidies). Only Armenia and the three Central American countries are expected to expand spending in real terms in 2009.

34. Automatic stabilizers have been allowed to come into play, but not fully as authorities in most programs took measures to contain the widening of headline deficits. Figure 16 shows the increase in fiscal deficits that would be required to allow automatic stabilizers work unimpeded. Automatic stabilizers are larger on average in program countries than in nonprogram or past crisis cases (as explained in Box 6, automatic stabilizers tend to be underestimated in cases of “negative tax buoyancy,” when revenues decline faster than GDP—a key feature of this crisis). Relative to the size of the estimated automatic stabilizers, however, the degree of actual fiscal accommodation in most program countries is being limited. As a result, withdrawals of fiscal impulse can be seen in a number of cases (Hungary, Pakistan, Ukraine, Romania, and Bosnia & Herzegovina show the largest withdrawals). Nevertheless, on the whole, the average fiscal impulse in the current programs is less restrictive than in past crisis cases, although it is tighter than in current nonprogram cases.

View Full Size

Discretionary fiscal adjustment in programs

(average overall balance, percent of GDP)

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: Staff estimates based on program documents.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Automatic stabilizers and change in fiscal stance

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: Staff estimates.

• Download Figure
• Download figure as PowerPoint slide

35. Much of the resistance to full play of automatic stabilizers can be explained by initial conditions. A cross-section regression of 55 program and nonprogram emerging market countries suggests that fiscal policy is less expansionary in countries with higher initial debt levels and lower starting cyclically-adjusted primary balances. Importantly, current programs do not show larger adjustments than nonprogram countries (an SBA dummy in the regression was insignificant).

Regression of fiscal impulse on initial conditions, 2009

Source: Staff estimates.

Regression of fiscal impulse on initial conditions, 2009

Dependent Variable: Fiscal Impulse
	Coeff.	Robust S.E.	t Stat	P-value
Cycl. Adjusted Prim Bal/GDP t-1	-0.48	0.06	-7.39	0.00
Public Debt/GDP t-2	0.03	0.01	2.30	0.03
Intercept	-2.71	0.61	-4.44	0.00
Adj-R2	0.46
Observations	56

Source: Staff estimates.

View Table

Regression of fiscal impulse on initial conditions, 2009

Dependent Variable: Fiscal Impulse
	Coeff.	Robust S.E.	t Stat	P-value
Cycl. Adjusted Prim Bal/GDP t-1	-0.48	0.06	-7.39	0.00
Public Debt/GDP t-2	0.03	0.01	2.30	0.03
Intercept	-2.71	0.61	-4.44	0.00
Adj-R2	0.46
Observations	56

Source: Staff estimates.

36 There are, however, outliers reflecting important country-specific factors:

• Iceland and Latvia exhibit larger deficit increases than predicted by the model, but these are explained by much larger (nondiscretionary) revenue declines than the assumed unitary elasticity to GDP growth would imply. Therefore, these cases do not involve discretionary policy loosening. Similarly Georgia and Armenia are accommodating an unexpected large revenue loss to prevent a deeper economic downturn (as indicated below, these countries also show the largest projected medium-term fiscal adjustments, backed by structural reforms).

• In the case of Costa Rica, credibility effects owing to past prudence (not captured in the regression) has created space for pursuing countercyclical policies, including by considerably expanding social safety nets.

• The unexplained element of Pakistan’s discretionary fiscal tightening can be reconciled with relatively weak policy credibility, as well as the automatic reversal of the large fiscal deterioration suffered in the run-up to the crisis as generous energy and food subsidies have fallen with the collapse in commodity prices.

• The unexplained component of fiscal tightening in Belarus can be attributed to the authorities’ preference for a balanced budget (instead of a looser fiscal stance, combined with tighter credit policies) as the best means to maintaining confidence. Similarly, in Hungary, discretionary loosening was limited by the need to preserve policy credibility with financial markets given the high debt ratio, the existence of large automatic stabilizers on the spending side, and EU-related constraints.

• In Ukraine, a smaller than predicted fiscal expansion was motivated, inter alia, by financing constraints and the deterioration in the finances of a state gas company in the face of a sudden increase in its import prices.

View Full Size

Initial conditions and variation across program countries

(based on cross-section regression, 2009)

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: Staff estimates.

• Download Figure
• Download figure as PowerPoint slide

D. Medium-term fiscal adjustment

37. Most countries are expected to face significant fiscal challenges in the coming years. A few program countries entered the crisis with weak fiscal positions, reflecting large entitlement programs, rapid public wage growth, and narrow tax bases, which the recent crisis has amplified. Even those countries that entered the crisis from stronger fiscal positions than in past episodes will experience large increases in public debt from widening deficits in the downturn, bank restructuring costs, and adverse macro conditions (lower growth, higher interest rates, and weaker exchange rates). Even so, most program countries are projected to preserve public debt ratios comfortably below 50 percent of GDP in outer years—though such outcomes are very sensitive to assumptions about potential output, interest and exchange rates (Figure 17). Latvia, Hungary and Iceland, however, stand out as countries with significant debt sustainability problems.

View Full Size

Public debt dynamics

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: WEO and staff estimates.

• Download Figure
• Download figure as PowerPoint slide

38. Programs envisage the unwinding of near-term fiscal accommodation through medium-term fiscal adjustment plans backed by structural reforms (Figure 18). Reflecting heterogeneity in initial fiscal positions, program projections show generally larger medium-term fiscal adjustments in countries with heavier debt burdens. Many programs project fiscal adjustments somewhat beyond debt-stabilizing levels, reflecting the priority of reducing, rather than stabilizing, public debt—with the degree of “over-adjustment” also generally greater for those countries with the heaviest debt burdens. This adjustment is generally buttressed by fiscal structural reforms (see Section II.D and Box 7).

View Full Size

Primary balance adjustments and debt-stabilizing levels in program countries

ranked by peak debt level during 2009–14 (percent of GDP)

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: Staff projections.1/ Primary balance that stabilizes debt ratio using 2014 program projections for debt ratios, growth, interest and exchange rates.

• Download Figure
• Download figure as PowerPoint slide

E. Social spending

39. Commitments to sustain or expand social safety nets to protect and shield the poor have been made by authorities in all program countries (Table 2).³ Authorities planned a mix of policies to preserve and/or expand protection of poorer households, including conditional cash transfer programs, housing utility allowances, labor-intensive infrastructure projects, and unemployment insurance. Pakistan has made stronger social safety nets a key priority (cash transfer programs were augmented) backed by program conditionality to strengthen targeting. In Costa Rica, the authorities are using “fiscal space” to increase spending on education and labor-intensive infrastructure projects, and to expand conditional cash transfer programs and noncontributory pensions (totaling 1 percent of GDP).

Table 2.

Social safety nets and spending on pensions and transfers

Source: Staff calculations

Table 2.

Social safety nets and spending on pensions and transfers

Source: Staff calculations

View Table

Table 2.

Social safety nets and spending on pensions and transfers

View Expanded

Source: Staff calculations

40. Over time, however, fiscal constraints have pushed some countries to change their strategy from higher expenditure to better targeting. Bosnia & Herzegovina, Hungary and Mongolia have shifted from higher spending in favor of better targeting.

41. The need to put fiscal positions on a sound long-term footing is also driving selected wage restraint and cuts in transfers and pensions. In some cases, these measures partly unwind the rapid wage and entitlement growth of the pre-crisis period. In Latvia, wage cuts in the public sector are being undertaken after years of very rapid growth to help restore competitiveness and secure public debt sustainability. Bosnia & Herzegovina—a peg like Latvia—is taking a similar approach. Cuts in a generous system of entitlements were also necessary in Hungary to address fiscal sustainability concerns. In Ukraine, after years of rapid increases in pensions, downward adjustments are expected to take place (minimum pensions will be safeguarded) to help place the pension system on a sound footing and facilitate medium-term fiscal adjustment.

A. Approach and overview

42. During a crisis, monetary and exchange rate policies have to trade off a number of objectives: inflation, external adjustment, and financial stability. Monetary tightening may be required to stem a run on the currency or prevent exchange rate overshooting— though adjustment may be unavoidable when the currency is out of line with fundamentals. Exchange rate and interest rate exposures in balance sheets need to be taken into account in considering the combination of monetary tightening and exchange rate adjustment, with capital controls a potential further tool in particular circumstances.

43. The fact of only modest levels of interest rates, currency depreciation, and inflation in this crisis stand in stark contrast to past experience. In most cases, low, if not negative, pre-crisis real interest rates combined with rising risk premia, required nominal policy rate hikes to preserve market confidence, but these increases were relatively modest compared to past episodes. As falling food and fuel prices reduced inflation pressure, such rate hikes, were typically sufficient to anchor medium-term inflation expectations and prevent the large currency overshooting that created havoc in previous crises (Figure 19). This outcome likely owes also to a number of supportive factors: (i) timely and frontloaded financing packages and larger initial reserve buffers removing tail risk scenarios; (ii) sharp reductions in advanced country interest rates; (iii) a variety of approaches to exchange rate policy, with greater emphasis on country ownership of currency regimes; (iv) better than expected bank rollover rates; and, possibly, (v) a reduced capacity for speculative attacks on currencies among hedge funds and other investors, given their own parlous conditions. As noted below, there were some exceptions to these patterns, notably the case of Iceland, which required capital controls to deal with a free-falling currency and large deposit outflows.

View Full Size

Post-crisis real interest rates and real exchange rate changes

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: IMF, International Finance Statistics, Information Notice System; and staff estimates.1/ Change from October 2007 to July 2009 for current programs, and for past crises, from 12 months before crisis to 12 months after crisis. A decrease means depreciation.2/ Average real central bank policy rates (for current programs) and discount rates (for past crisis cases) for latest three months (t+7 to t+9 for past crises). “t” is the beginning of crisis for past crises, and October 2008 for current programs.

• Download Figure
• Download figure as PowerPoint slide

B. Exchange rate policy

44. The Fund has sought to respect the authorities’ choice of exchange rate regime, while ensuring their consistency with macro policies. In the context of an adjustment program, the regime choice needs to provide credibility to the adjustment strategy. And, even when warranted by the need to unwind large currency misalignments, changes in regime are often controversial, not least because of the intrinsic difficulty in weighing their costs and benefits, as made clear by the Latvia case (Box 8). Only the Belarus and Ukraine programs involved a change in the de jure exchange rate regime toward greater flexibility (Figure 20 shows bilateral exchange rate developments through the crisis using an illustrative—and necessarily somewhat arbitrary—characterization of countries’ de facto exchange regimes).

View Full Size

Daily spot exchange rates under programs

Percent change from June 2008 (- = depreciation), circles mark program approval dates ^1/

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: Bloomberg.^1/ Exchange rates are defined as units of national currency per euro for Bosnia & Herzegovina, Hungary, Iceland, Latvia, Romania, and Serbia, and per U.S. dollar for remaining countries.

• Download Figure
• Download figure as PowerPoint slide

45. While the overshooting that typified past crises has largely been avoided so far, recent currency stability needs to be set against potential future needs for adjustment. After the initial period of volatility, exchange rates have tended to stabilize (Figure 20): except for Hungary, Iceland, and lately Ukraine (which has experienced some depreciation since late August), no program country exchange rate fluctuated by more than 5 percent throughout May–August 2009. For the most part, this stability has not involved major direct exchange rate intervention, but rather a combination of the positive impact of confidence returning and, in some cases, moral suasion. However, very stable currency movements in some programs may hinder the objectives of unwinding preexisting currency misalignments or adjusting to large terms-of-trade shocks, and avoiding a buildup of risky foreign exchange exposures in balance sheets.

View Full Size

Nominal effective exchange rates

Median and interquartile ranges, average of t-12 to t-7 = 100, monthly ^1/

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: IMF, International Finance Statistics and Fund staff estimates.^1/ “t” represents Oct 2008 for current programs, and beginning of crisis for past crises.

• Download Figure
• Download figure as PowerPoint slide

46. In some cases, the limitation of currency depreciation reflected concerns about immediate losses from foreign exchange exposures in private balance sheets (Section V). In cases such as Hungary, Latvia, Iceland, and Ukraine, program policies gave significant weight to the potential disruptions in balance sheets from excessive currency depreciations, and in many cases, program reserve targets have given room for needed intervention in the foreign exchange market to avoid disorderly movements. In Iceland, the severity of the currency and banking crisis required a combination of large-scale foreign exchange intervention and capital controls to stabilize the krona, which depreciated by more than 30 percent against the euro in the months prior to the approval of the Fund arrangement.

47. In real terms too, currency depreciations have so far been strikingly more moderate than in past crises and closer to current nonprogram countries. Inflation has been falling rapidly in program countries, although it remains higher than in their trading partners. As a result, in program countries, the median real exchange rate depreciation for October 2007 to July 2009 was negligible at 0.3 percent (Figure 21). This compares to a median appreciation of some 5 percent for nonprogram emerging market economies and a 33 percent depreciation in past crises at a similar stage. Pegs have seen their real exchange rates appreciating on average by about 6 percent, further eroding competitiveness. The modest real depreciations to date in many program countries mirror relatively mild current account adjustments, as noted in Section II, and suggest that further real exchange rate depreciation may be needed down the road to allow the unwinding of the large pre-crisis external imbalances that induced some program countries to turn to the Fund (Section VI).

View Full Size

Real effective exchange rates

Median and interquartile ranges, average of t-12 to t-7 = 100, monthly ^1/

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: IMF, Information Notice System and staff estimates.^1/ “t” represents Oct 2008 for current programs, and the beginning of the crisis for past crises.

• Download Figure
• Download figure as PowerPoint slide

C. Monetary policy

49. Recent Fund programs use a variety of nominal anchors for price stability. For those countries that do not peg their currencies, monetary programs either establish a combination of quantitative targets on monetary aggregates (net domestic assets of the central bank or base money) or preserve existing inflation targeting frameworks. Complementary floors on the central bank’s net international reserves aim to safeguard Fund resources, while providing adequate room for foreign exchange intervention. Even in programs that use monetary aggregates as intermediate policy targets, policy rates have been adjusted in a discretionary way in the face of volatile market conditions. As with other program targets, monetary policy goals are revised periodically as conditions change: for example, some programs have recently relaxed targets on domestic credit. Policy rates were also reduced as inflation has eased and risks of disorderly currency movements have abated.

50. The increase in policy interest rates to bring real rates into positive territory and stabilize market confidence have been modest compared to past crises (Figure 22). In fact, in some cases, outright policy tightening was accompanied by injections of (domestic and foreign currency) liquidity to address shortages of liquidity in financial systems (Section V). All things considered, monetary conditions were overall tightened initially to prevent excessive currency depreciations while exchange rates found their new equilibria and contain second-round effects of currency depreciation on inflation.

View Full Size

Interest rates

Medians and interquartile ranges, monthly ^1/

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: IMF, International Finance Statistics and Fund staff estimates.^1/ “t” represents Oct 2008 for current programs, and the beginning of the crisis for past crises. EMs include Argentina, Brazil, Chile, China, Croatia, Czech Republic, Ecuador, India, Indonesia, Jordan, Korea, Lithuania, Malaysia, Mexico, Morocco, Peru, Poland, Russia, Thailand, Turkey, Venezuela, Vietnam.

• Download Figure
• Download figure as PowerPoint slide

51. In recent months, falling inflation and financial stability have created space for monetary policy easing. Against the backdrop of a global deflationary environment, the avoidance of large exchange rate depreciations in most program cases has set inflation on a downward trajectory. This is in stark contrast with past crises, where inflation picked up sharply in tandem with significant exchange rate depreciations. As a result, where financial stabilization has taken hold, prudent relaxation of program targets and policy rates has begun. However, the more rapid pace of interest rate cuts observed in nonprogram emerging market countries, suggests that there may be room for some further reduction in policy rates in program cases, although such easing would need to be consistent with individual country constraints and inflation objectives.

View Full Size

Inflation

y-o-y change in percent, medians and interquartile ranges ^1/

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: IMF, International Finance Statistics and Fund staff estimates.^1/ “t” represents Oct 2008 for current programs, and beginning of crisis for past crises.

• Download Figure
• Download figure as PowerPoint slide

A. Incidence of financial sector crisis

52. Banking systems in program countries have largely been shielded so far from the dislocations observed in some advanced countries or previous crises. In only Iceland and, to some extent, Latvia and Ukraine can it be plausibly argued that the financial sector was key in triggering or exacerbating the crisis (Box 9). By contrast, banking system meltdowns were hallmarks of most other recent capital account crisis cases. Even so, banking problems typically lag the economic cycle and pressures are likely to materialize going forward.

View Full Size

Gross fiscal cost of systemic banking crises

(percent of GDP)

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: Laeven and Valencia (2008); FAD-MCM: Public Interventions in Financial Systems.

• Download Figure
• Download figure as PowerPoint slide

53. The relative banking system resilience in most emerging market countries reflects both past institutional reforms and decisive policy responses in this crisis. Emerging markets have made substantial progress in recent years in strengthening supervision, addressing systemic risks, and reducing balance sheet vulnerabilities—as a result, for most program countries financial system initial conditions were generally benign compared to previous crisis episodes. The policy response to the current crisis internalized early on the potential for sharp reductions in cross-border flows from global deleveraging, by adopting comprehensive measures to address incipient liquidity needs. At the same time, efforts were made to avoid disruptive currency overshooting in light of large foreign currency exposures in private balance sheets.

B. Policy considerations

54. Avoiding liquidity runs and addressing underlying bank vulnerabilities were key to contain the effects of the crisis. The risk of liquidity pressures was significant given banking systems’ reliance on external financing. Such risks were most concentrated in Iceland and Latvia with bank external debt liabilities approaching 250 percent and 100 percent of GDP respectively, but were also prevalent in many other countries (Figure 23). The high incidence of foreign ownership of domestic banks was also a source of potential instability given that advanced-country parent banks themselves were under serious pressures. Program design also took into account country-specific domestic vulnerabilities, including pre-existing insolvency of key banks (Iceland, Ukraine); corporate sector distress (Georgia, Latvia, Pakistan); and deficient supervisory and legal frameworks (Belarus, Costa Rica, Guatemala), as well as directed lending concerns (Belarus and Guatemala).

View Full Size

Bank cross-border exposures

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: IFS

• Download Figure
• Download figure as PowerPoint slide

55. It was also important to avoid currency overshooting given widespread liability dollarization. The incidence of unhedged foreign currency liabilities was considered to be serious in Belarus, Georgia, Guatemala, Hungary, Iceland, Latvia, and Serbia. Moreover, in many of these cases, net external exposures were also very large.

C. Policy measures and outcomes

56. Financial policies focused on maintaining bank liquidity and solvency. Curbing liquidity pressures and financial stress (including by safeguarding external credit lines, Box 10) took precedence over structural measures aimed at strengthening balance sheets and addressing solvency concerns. Measures, some of which were incorporated in program conditionality, took different forms depending on the nature of shocks and country situations (Table 3, Box 11). Measures to facilitate household and corporate debt restructuring also formed an important element of the longer-term agenda in some countries (Box 12).

Table 3.

Summary of financial sector measures

Source: Program documents

Table 3.

Summary of financial sector measures

	Immediate stabilization							Longer-term structural
	Liquidity Support
	Introduction of central bank liquidity facility	Broadening eligible collateral	Cuts in reserve requirements	Strengthening deposit insurance	Regulatory forbearance	Deposit freeze	Intervention of distressed banks	Resolution framework	Supervision	Bank recapitalization	Debt restructuring
Armenia	•	•			•			•
Belarus		•	•		•				•	•
Bosnia			•	•
Costa Rica	•			•				•		•
El Salvador								•	•
Georgia	•		•			•			•
Guatemala	•		•		•			•	•	•
Hungary	•	•	•	•				•	•	•	•
Iceland		•	•	•	•	•	•		•	•	•
Latvia		•	•	•		•	•	•	•	•	•
Mongolia				•			•
Pakistan					•			•	•		•
Romania	•	•	•	•	•			•	•
Serbia	•	•	•	•	•		•		•
Ukraine	•	•	•	•		•	•	•	•	•	•

Source: Program documents

View Table

Table 3.

Summary of financial sector measures

	Immediate stabilization							Longer-term structural
	Liquidity Support
	Introduction of central bank liquidity facility	Broadening eligible collateral	Cuts in reserve requirements	Strengthening deposit insurance	Regulatory forbearance	Deposit freeze	Intervention of distressed banks	Resolution framework	Supervision	Bank recapitalization	Debt restructuring
Armenia	•	•			•			•
Belarus		•	•		•				•	•
Bosnia			•	•
Costa Rica	•			•				•		•
El Salvador								•	•
Georgia	•		•			•			•
Guatemala	•		•		•			•	•	•
Hungary	•	•	•	•				•	•	•	•
Iceland		•	•	•	•	•	•		•	•	•
Latvia		•	•	•		•	•	•	•	•	•
Mongolia				•			•
Pakistan					•			•	•		•
Romania	•	•	•	•	•			•	•
Serbia	•	•	•	•	•		•		•
Ukraine	•	•	•	•		•	•	•	•	•	•

Source: Program documents

Box 10.

Bank Coordination Initiative (BCI) in CEE countries

The BCI was established by the IMF, the European Commission, and the EBRD to involve the private sector in stabilizing CEE banking systems, preventing disorderly deleveraging, and reducing uncertainty. Since early 2009, coordination meetings have taken place for Bosnia, Hungary, Romania, and Serbia. Other countries (including Latvia) are expected to join this Initiative soon.

Foreign banks’ participation and originating countries, 2009

Foreign banks’ participation and originating countries, 2009

	Bosnia	Hungary	Romania	Serbia
Foreign banks’ market share (percent of assets)	95 percent	70 percent	88 percent	75 percent
Country of origination of foreign banks	Austria, Italy, Slovenia	Austria, Italy, Germany	Austria, France, Greece, Italy	Austria, Italy, Greece, France

View Table

Foreign banks’ participation and originating countries, 2009

	Bosnia	Hungary	Romania	Serbia
Foreign banks’ market share (percent of assets)	95 percent	70 percent	88 percent	75 percent
Country of origination of foreign banks	Austria, Italy, Slovenia	Austria, Italy, Germany	Austria, France, Greece, Italy	Austria, Italy, Greece, France

Under the BCI, parent banks commit to maintaining exposure in host countries and maintaining adequate capital and liquidity in subsidiaries. Such commitments do not have legal standing and are intrinsically difficult to monitor. But their nonobservance would eventually expose parent banks to reputational costs.

As part of the BCI, host countries commit to sound policies under IMF-supported programs, focusing on the stability and solvency of banking systems. To this end, bank stress tests in individual countries and at the regional level have either been completed or are underway. Policies in some cases afford greater flexibility to the overall banking system. For instance, in Romania, capital shortfalls were allowed to be met with Tier 2 instead of Tier 1 instruments; in Bosnia & Herzegovina and Romania, reserve requirements were eased; and in Serbia, central banks established liquidity facilities.

A concern raised with the BCI is that it could shift the location of deleveraging to other countries. Thus, regional coordination is essential for the ultimate success of the Initiative.

Box 11.

Financial sector measures in program countries

Liquidity support has been substantial in some cases.

Strengthening deposit insurance has included the creation of new frameworks, extension of coverage, including blanket guarantees in some cases, and the boosting of resources backing guarantee funds. In some countries, measures in this area have been limited as insurance had already been strengthened in recent years, or the authorities were concerned about adverse signaling effects.

View Full Size

Liquidity support, selected countries

(percent of GDP)

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: Staff estimates

• Download Figure
• Download figure as PowerPoint slide

Deposit freezes have been used sparingly and on a temporary basis. In some cases, the imposition of controls has effectively put limits on deposit withdrawals.

Regulatory forbearance measures included easing prudential requirements or allowing banks more time to meet them. At the same time, supervisory powers were also strengthened in certain areas in response to the crisis.

Interventions of distressed banks have involved: Iceland—three largest banks, via adoption of “new bank/old bank” approach, as well as four other financial institutions; Latvia—takeover of 85 percent of the shares and a subsequent recapitalization of the second largest bank; Mongolia—one bank was placed into conservatorship; Serbia—a small bank was put under receivership; and Ukraine—the sixth largest bank, and a number of smaller banks, were put under receivership, and the authorities are resolving two systemic banks.

Bank recapitalization is crucial to restore bank viability, but progress in this area has typically been slow reflecting a variety of factors including insufficient progress in resolving asset valuation issues, low private investor interest, and ambiguities on the role of the public sector.

Key areas under the resolution framework include setting up mechanisms to support bank intervention, restructuring and recapitalization, including legislative changes where needed.

57. Financial sector stability has been maintained for the most part, helped by liquidity measures and relative currency stability.

• Initial measures have generally avoided deposit runs and limited pressures on domestic and forex liquidity. In particular, liquidity support, complemented in some cases by administrative measures and controls, as well as early bank interventions, were successful in stemming deposit runs in Iceland, Latvia, and Ukraine. The more persistent deposit declines in Georgia and, to a lesser extent, Mongolia and Ukraine, appear to reflect broader, including (in the case of Ukraine) solvency concerns.

• Bank external credit lines have held up better that initially assumed. For the most part, parent banks appear committed to funding their subsidiaries, and in some cases central banks of the parents’ jurisdiction (e.g., Nordic central banks) have publicly committed the use of their credit lines for liquidity support of program country subsidiaries. Hence, contrary to initial concerns, foreign bank ownership has so far proved a net strength, with foreign bank flows remaining on balance much more stable than other types of private capital flows in most program countries. Still, recent figures suggest that external bank financing has showed marked declines in some cases, including Ukraine and, to a lesser extent, Georgia, Guatemala, Latvia, and Pakistan.

• As discussed in Section IV, exchange rate stability has been a major factor in ensuring financial sector stability, especially for countries with large net open forex positions in financial, corporate, and household balance sheets.

View Full Size

Change in bank external debt liabilities

(08Q4–09Q1 to 08Q2–08Q3 change, in percent of end-2007 stock)

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: IFS.

• Download Figure
• Download figure as PowerPoint slide

58. On the other hand, the asset side of banks’ balance sheets is evolving less favorably than originally envisaged in most programs. In particular, credit to the private sector continues to decline in most cases, in part reflecting weaker demand, but also tighter lending standards (Figure 24). While the evolution of non-performing loans (NPLs) has been mostly benign so far, the experience of past crises suggests that NPLs are typically recognized with long lags (Figure 25) and could be expected to peak in the next couple of years in countries experiencing deep recessions, putting renewed pressure on bank capital. Thus, rapid progress in recapitalizing banks based on forward-looking assessments of asset quality, and in restructuring household and corporate debts would prevent the recurrence of banking sector problems and pave the way for a recovery in credit flows.

View Full Size

Deposit and credit developments ^1/

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: IFS.^1/ Adjusted for exchange rate changes.

• Download Figure
• Download figure as PowerPoint slide

View Full Size

Bank balance sheet indicators

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: Staff calculations

• Download Figure
• Download figure as PowerPoint slide

A. Early signs of stabilization

60. Evidence from previous crises points to a duration of capital and exchange market pressures of around two years. In previous crisis episodes, spreads have returned to their pre-crisis levels, exchange rates have stabilized, and capital flows have resumed on average after about 1½ years of crisis inception (Mecagni et al, 2007). However, there has been greater variation in the time required to regain market access (defined as first primary issuance of international government bonds since the beginning of the crisis), although this may reflect cases where the country did not need to place bonds, despite having the capacity to do so at favorable rates (for instance, despite a V-shaped recovery, Korea’s first sovereign bond issuance took place only several years after crisis inception).

View Full Size

Crisis duration in past crisis cases ^1/

(in quarters)

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: Mecagni et al (2007).^1/ Combined index of capital flows, spreads, reserves, and nominal effective exchange rate.

• Download Figure
• Download figure as PowerPoint slide

61. Stabilization in recent program cases appears so far to be on a relatively fast track compared to this past experience. After almost a year since Lehman’s bankruptcy, currencies have largely stabilized in all program countries, barring for Ukraine, which remains subject to bouts of market volatility. And, although spreads remain well above their pre-Lehman level, they have declined significantly since April, in line with developments elsewhere. In particular, with some exceptions, median values for spreads are significantly below their pre-SBA approval levels. Capital inflows have also resumed, albeit at low levels, to most countries except for Latvia and Ukraine, partly thanks to continued net inflows of FDI, EU transfers, and portfolio flows. The recent market stabilization has also created a window of opportunity for sovereign Eurobond placements. Among program countries, Hungary successfully placed a five-year €1 billion Eurobond in July, and Romania has announced plans to issue a €0.5–1 billion Eurobond in September 2009.

View Full Size

Crisis duration indicators, 2008–09

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: Bloomberg, IFS, and staff estimates.

• Download Figure
• Download figure as PowerPoint slide

B. Recovery and risks to external sustainability

62. Following the sharp downturn in 2009, the recovery in program countries is projected to be slow, with large downside risks. Current program countries are unlikely to be able to benefit from the buoyant external demand and large exchange rate adjustments that have underpinned export-driven recoveries in past crises. As a result, the recovery in current programs is expected to be more subdued than in previous crises. Relatedly, it is as yet unclear whether potential growth can return to the rates seen in the past, which puts a premium on making rapid progress in structural reforms.

View Full Size

Trade volumes: imports by G3 countries

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: WEO and staff estimates.

• Download Figure
• Download figure as PowerPoint slide

63. Medium-term current account adjustments are projected to be limited in most cases, pointing to continued large external financing pressures. With a few exceptions, following the fluctuations in 2008– 09, current account deficits are expected to return to levels only marginally below pre-crisis, which is in stark contrast to the sharp current account turnarounds observed in past crises. But, if these projections prove accurate, the vulnerabilities that contributed to this crisis will continue to persist in the medium term.

View Full Size

Current account adjustment

percent of GDP

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: WEO and staff estimates.

• Download Figure
• Download figure as PowerPoint slide

64. However, continued current account deficits do not necessarily imply external debt sustainability problems. Given projected FDI, EU capital transfers, and other nondebt creating flows, in most cases external debt is projected to remain manageable, though there are significant downside risks to this outlook (Figure 26). Moreover, even compared to the highest debt cases of past crises, Iceland and Latvia face a very challenging debt burden, made heavier by low growth and deflation. Hungary also shows only a slowly declining trend, with the external debt ratio remaining well above 100 percent of GDP mainly due to weak long-term growth.

View Full Size

External debt dynamics

Citation: Policy Papers 2009, 067; 10.5089/9781498335485.007.A001

Source: WEO and staff estimates.

• Download Figure
• Download figure as PowerPoint slide

65. Continued large financing needs and high debt ratios pose risks to regaining full market access, both for the private and public sectors. The timing and extent to which program countries will have access to market financing has a bearing on their capacity to repay the Fund.

C. Exit from Fund support

66. The length of current Fund arrangements is comparable to that of past cases at inception, but extensions and successor arrangements are a possibility. Current programs have an average duration of about two years in line with past programs. Nevertheless, in the past, several programs have been extended and successor arrangements put in place, with the overall duration of Fund support reaching 10 years for Uruguay, nine years for Turkey, and eight years—on a precautionary basis—for Peru. Almost 20 percent of past extensions have also entailed augmentation of access—notable are the cases of Uruguay and Brazil in 2002 (with the latter arrangement turning precautionary at the time of its extension) and Serbia in 2009. Thus, if downside risks to the global economy materialize, extensions of current arrangements or successor arrangements may be required for some current program cases. In contrast, early exits from Fund support and greater use of precautionary financing could well occur if upside risks materialize. Following the program period, capacity to repay the Fund would be monitored in most current program countries under Post-Program Monitoring.