Sweden
2009 Article IV Consultation: Staff Report; Staff Statement; Public Information Notice on the Executive Board Discussion; and Statement by the Executive Director for Sweden
Author:
International Monetary Fund
Search for other papers by International Monetary Fund in
Current site
Google Scholar Close
Publication Date:
07 Aug 2009
eISBN:
9781451836035
Language:
English
Keywords:
ISCR; CR; authority; rate; Riksbank; staffs assessment; board assessment of policy issue; potential GDP; exchange rate; statistics Sweden; financial crisis; labor market; Fiscal stance; Global; Baltics
Download PDF (1.3 MB)

This 2009 Article IV Consultation highlights that Sweden has been hit hard by the global financial crisis. Two of its banks built up large exposures in the Baltics that significantly increased loan losses beyond normal recessionary levels. In response to the crisis, the authorities have taken wide-ranging measures to stabilize the financial system and support demand. Executive Directors have welcomed the authorities’ prompt and appropriate policy responses, which have allayed immediate concerns with financial sector stability, and have helped cushion domestic demand.

Abstract

This 2009 Article IV Consultation highlights that Sweden has been hit hard by the global financial crisis. Two of its banks built up large exposures in the Baltics that significantly increased loan losses beyond normal recessionary levels. In response to the crisis, the authorities have taken wide-ranging measures to stabilize the financial system and support demand. Executive Directors have welcomed the authorities’ prompt and appropriate policy responses, which have allayed immediate concerns with financial sector stability, and have helped cushion domestic demand.

I. Summary and Appraisal

Up to the Spring of 2008, the Swedish economy boomed

1. Since 2002, Sweden thrived in buoyant global conditions, reflecting strong policies and a composition of output—investment goods and consumer durables—particularly favored by the boom.

2. But the same worldwide factors which supported this record also left Sweden highly exposed to the post Fall-2007 financial crisis. Its investment goods and consumer durables exports were hurt by weakening external demand long before the major contraction in global trade late in 2008. And as the international financial wholesale markets closed and investor concerns with Baltic exposures mounted, liquidity crunched. Having surfed the earlier global wave, Sweden has been hit hard by its crash.

Policy action has steadied confidence, but the tide is still out

3. The authorities’ response was prompt and appropriate, and immediate concerns with financial sector stability have been addressed. The full range of measures typical elsewhere was implemented, incorporating the agreed European Union (EU) response, including on the fiscal side, full operation of automatic stabilizers and a discretionary budget loosening for 2009—taking the budget from a surplus of 2½ percent of GDP in 2008, to a deficit of 4 percent in 2009. Much has been done.

4. Immediate prospects for recovery are at the mercy of developments abroad. As a producer of goods disproportionately favored by the prior global boom, Sweden was one of the first into the downturn. And if demand for those goods recovers slowly relative to other components of global demand—as seems likely given large output gaps, credit constraints, and wait-and-see behavior by investors and consumers abroad—Sweden could be one of the last out. Accordingly, with GDP already declining slightly through 2008, staff project it to fall by 6 percent in 2009 with recovery firmly taking hold only from the middle of 2010. In this context, unemployment will rise significantly.

Policy for 2009–10 should continue to aim to minimize the fall

5. The ongoing integrated policy response could be strengthened in various ways, to further boost Sweden’s resilience in this testing environment, while maintaining confidence.

Large fiscal stabilizers should continue to operate, along with the stimulus underway

6. However, the case for additional discretionary fiscal activism to offset the downturn is not persuasive at this juncture. The scale of the fiscal action underway—as measured by the change in the general government headline balance—is, on staff projections, already one of the largest in the EU. Furthermore, plausible estimates of potential output yield widely varying estimates of the structural balance and the fiscal outlook, and it is unclear how much public debt will rise due to financial sector rescue operations that may prove to be necessary. In addition, the stabilizers are large, the multipliers are small, and with estimates of medium-term potential output growth being lowered globally, the long run fiscal strength, apparent at the outset of the crisis, will need to be thoroughly reassessed when the waters settle.

7. Accordingly, the scale of the shift in the structural fiscal balance projected by staff—a weakening of 2¼ percentage points of GDP in 2009, and a further ¼ percentage point in 2010—goes as far as is appropriate in providing a decisive fiscal response to a uniquely sharp fall in demand, without compromising sustainability. The composition of the discretionary component of the shift focuses on tax reductions, aiding supply side efficiencies even though their immediate demand impact is likely limited.

8. On the fiscal side, the key strengthening suggested concerns operation of the fiscal rules. The target of a surplus of 1 percent of GDP over the cycle should remain, as should the firm commitment to the nominal spending ceilings, because these support fiscal sustainability. But a new commitment to adjust the spending ceilings to offset the net revenue impact of any further discretionary tax reforms on projected budget balances would strengthen medium-term sustainability further.

Monetary policy is appropriate

9. On the monetary side, a modest undershoot of the inflation target—1 to 3 percent—is likely this year. However, risks of sustained disinflation appear low: in particular, short and long term indicators of inflation expectations point this way. Only if inflation is expected to fall significantly and on a sustained basis, should consideration be given to further action.

10. The exchange rate is probably modestly undervalued. Given the particular exposure of Sweden to the global collapse in demand for investment and durable consumer goods, a good part of the recent depreciation is likely an equilibrating adjustment. Accordingly, along with the reforms to labor market structures and income taxation in recent years, it will play a key role in containing increases in unemployment and the accompanying concerns with hysterisis.

Additional proactivity may strengthen the financial sector further

11. Financial sector fragilities reflect Baltic exposures, recession, and banks’ reliance for funding from global wholesale markets.

12. These fragilities have been assessed in thorough stress tests, indicating that banks will meet regulatory minimum requirements even in highly adverse contexts. But market concerns about the adequacy of Swedish banks’ capital remain, indicating, as internationally, that more remains to be done, notwithstanding the extraordinary stability support measures that are in place.

13. Further steps to strengthen the resilience of the banks, including by boosting capital, when needed, and developing contingency planning, are warranted as soon as possible. Such steps would boost resilience to short term shocks, reduce risks to credit supply and to taxpayers, and enhance an early exit from the extraordinary support measures.

14. Where specific banks appear to be at risk, enhanced supervision and (if needed) rights issues should be considered. If such steps prove insufficient, then public equity—injected at prices appropriate to ensure protection of taxpayer interests—and implementation of a bad-bank model are further options.

15. Alongside, continued broader contingency planning by the authorities is also appropriate, including further review of the toolkit for supervisory intervention to ensure robustness. There is also scope for further increases in international reserves beyond those already announced.

16. Beyond this, the institutional capacity of the Financial Supervisory Agency should be boosted. Given sizable international operations of large Swedish banks, further concerted efforts to strengthen cross-border crisis resolution mechanisms, coordinated with EU partners, should also be made.

17. Sweden should remain on the standard 12–month consultation cycle.

II. The Context—2004-08 1

In buoyant global conditions, Sweden thrived (Figures 1 and 2)

Figure 1.
Figure 1.

Sweden: The Long View, 1996–2008

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Sources: Haver Analytics, Konjunkturinstitutet, Statistics Sweden, and IMF staff calculations.1/ In thousands.
Figure 2.
Figure 2.

Sweden: The Recent Boom. 2001–07

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Sources: Eurostat, Haver, Statistics Sweden, and IMF staff calculations.

18. Activity rebounded strongly after 2001–03, led initially by a turn-around in exports and the associated boost to investment. Income flows from earlier foreign investments by Swedish banks and corporates provided an additional boost. In this context, domestic factor markets tightened, which, alongside positive wealth effects from the associated booms in stock and house prices, buoyed private consumption. Household debt rose sharply, and import growth caught up with, and eventually outpaced export growth.

This has reflected and supported strong policies

19. As noted by the authorities, in the context of high household savings rates, countercyclical policies helped restrain demand (Figure 3; ¶60). Automatic stabilizers are large and discretionary fiscal policy is tightly constrained by rules on the overall balance and on expenditure (Box 1). This framework has produced strong fiscal outcomes (Table 8). Inflation remained within the 1–3 percent target band bar mid-2008 when global food and oil price developments caused a temporary surge. These policies were consistent with past staff advice (Box 2).

Figure 3.
Figure 3.

Sweden--Fiscal Policies are Sound, 1993–2008

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Sources: Ministry of Finance, Eurostat, and IMF staff calculations.

Sweden’s Fiscal Rules

The main fiscal target is to achieve a general government surplus of 1 percent of GDP over the business cycle. This is supported by expenditure ceilings (rolling three-years ahead) for central government and social security outlays (excluding interest), and the balanced-budget requirement for local governments. The authorities use three indicators to assess performance against the surplus target (See Text Figure). The framework aims for flexibility over the cycle while responding to demographic trends, thus ensuring long-term sustainability of the public finances and intergenerational fairness.

uA01fig01

Compliance with the Fiscal Target

(Percent of GDP)

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Source: Spring Bill 2009.

Implementation of IMF Recommendations

Fiscal policy and framework. Consistent with IMF recommendations, the authorities have maintained adherence to their fiscal rules.

Monetary policy: Following Fund recommendations, prior to fall 2008, the Riksbank maintained a tightening bias to anchor inflation expectations—which remained above the upper bound of inflation target. Since October, the monetary policy stance has been accommodative to keep inflation expectations from falling below the lower band of inflation target.

Financial sector: IMF calls for a dedicated bank resolution framework have recently been implemented, but a clear PCA type framework is required, more rapid Deposit Insurance payout arrangements, and additional resourcing of the FSA remain outstanding.

20. Bank profitability grew rapidly between 2002–07, reflecting domestic and foreign lending, including in the Baltics. Overall capital adequacy was around 10½ percent, while Tier 1 capital adequacy was around 7 percent, with loan losses remaining low.

But strong growth has also reflected the composition of Sweden’s output (Text Table)

21. Sweden’s output bundle is dominated by investment-related and durable consumption goods—computers, industrial machine tools, electrical equipment, and chemicals—and increasingly, business services. These goods disproportionately benefited from the global boom after 2002. This accounts, in part, for rapid export growth to European emerging markets, oil exporters, and to booming Nordic economies.

Sweden: Key Export Sectors, 1990–present

article image
Sources: Eurostat, Statistics Sweden; and IMF staff calculations.

22. Alongside, Baltic entry into the EU in 2004 heralded a period of rapid growth in Swedish banks’ asset acquisition there—largely funded on buoyant international capital markets. This generated high income reflows and strong performance of the associated banks’ stock prices (Figure 4).

Figure 4.
Figure 4.

Sweden: Increasing Openness Over the Boom, 2000–08

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Sources: Statistics Sweden; and IMF staff calculations.1/ Sum of exports and imports as a percent of GDP.

Sweden: Geographical Composition of Exports, 1995–present

article image
Sources: Direction of Trade Statistics; and IMF staff calculations.

Notes: Excluding the Baltic countries (Estonia, Latvia, and Lithuania).

Competitiveness remains firm (Box 3, Figure 5)

Figure 5.
Figure 5.

Sweden: External Competitiveness Remains Firm, 1996–2008

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Sources: International Finance Statistics, Konjunkturinstitutet, Statistics Sweden, and IMF staff calculations.

Competitiveness and the Equilibrium Real Exchange Rate

Last year, staff analysis suggested considerable Krona undervaluation. (Text Table).

Previous CGER Estimates

article image
Source: CGER.

Since then, notably since the Fall of 2008, the Krona has depreciated considerably. However, staff concerns have not increased; indeed, we are rather less concerned than we were. This reflects revisions to staff projections for exports and investment returns. In particular, staff no longer assume that the strength of Sweden’s exports and strong factor income in recent years was permanent. Rather, both, in hindsight, appear to have reflected temporary surges, the first reflecting the composition of global demand during a bubble—the 2009 Spring WEO projects growth of fixed investment in coming years to significantly lag domestic demand in the US and Euro areas, in contrast to earlier in the decade—and the second reflecting unsustainable returns from bank investments in the Baltics. With durables, investment-input, and investment exports constituting well over half of total exports, a reduction in their earlier-estimated trend growth rate by 1/3 lowers total exports by some 1¼ percentage points of GDP by 2014. And the revised outlook for the Baltic states may lower investment returns by some 3 percentage points of GDP over the medium term.

Following these revisions, estimates as of May 2009 were as follows (Text Table). The reported range is large and all the measures are subject to uncertainty. However, the External Sustainability estimate may overstate the equilibrium exchange rate because the ongoing ageing of the population calls for a strengthening rather than a stabilization of the net external position in the medium-term. The appreciation implied by the RER estimate would be inconsistent with this. And the macro balance estimate is much smaller. Accordingly, alongside other indicators, notably the moderate level of inflation expectations, plummeting exports (like elsewhere), and the similar goods competitiveness as was realized during the last global downturn, the staff assessment, despite Krona depreciation since Q4 2008, is that the currency is probably not very significantly undervalued.

Current CGER Estimates

article image
Source: CGER.

By 2008, however, Sweden was highly exposed to external shocks

23. The global factors supporting the recent boom also increased vulnerability to the subsequent international recession. Swedish exports felt the impact as early as Q2 2008, with a significant acceleration in the decline as elsewhere in Q4 2008 (Figure 6). In addition, credit growth began to fall, both to households and corporates, albeit gently, with few signs, even in the April 2009 lending conditions survey, of rationing.

Figure 6.
Figure 6.

Sweden: Into the Downturn, 2007-09

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Sources: Haver, Statistics Sweden, and IMF staff calculations.1/ OMX Stockholm Price Index (1995=100); bond yield and STIBOR in percentage points.

24. Meanwhile, weaker consumer confidence—reflecting tumbling equity values, rising borrowing costs, and later, by falling employment—was reflected in greater precautionary saving through all of 2008, and consumption fell. Subsequently, investment—already bruised by falling exports and the global outlook—also fell starting Q2 2008. Output fell throughout H2 2008.

Overt banking and financial sector strains also emerged

25. Banks’ profitability fell sharply in 2008–09 despite negligible exposure to US subprime—or other structured—assets (Figures 7 and 8). Two of the largest banks, both increasingly funded on wholesale markets and exposed to the Baltics, both saw sharp increases in loan losses with their share prices and ratings marked down accordingly. Money market spreads—though lower than elsewhere—tightened sharply in Q4 2008 and have yet to return to pre-crisis levels. The stock market dropped nearly 40 percent in 2008, recovering somewhat recently, and spreads of mortgage bonds over government bonds rose to unprecedented highs. Alongside, solvency ratios of insurers fell (Figure 9).

Figure 7.
Figure 7.

Sweden: Selected Financial Indicators, 2005–09

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Sources: Thomson Financial/Datastream, Bloomberg, and Haver.
Figure 8.
Figure 8.

Performance of the Swedish Banking System, 2007–09

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Source: Thomson Financial/Data Stream, Bloomberg, Banks’ Annual Reprots, and WEO.
Figure 9.
Figure 9.

Sweden: Non Bank Financial Sector, 1990–2008

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Sources: Riksbank, Statistics Sweden, and Haver.

Summary of the Performance and Operation of Swedish Four Major Banks

(In percent; unless otherwise indicated; end period)

article image
Sources: Banks’ annual reports; and IMF staff estimates.

With Basel II transition rule.

For Nordea and Handelsbanken, excludes loans to credit institutions; for SEB, includes all credit portfolios (such as commitment and guarantees); and for Swedbank, includes credit institutions.

Monetary and Financial Sector Stability policies adjusted decisively

26. Until September 2008, the Riksbank had maintained a tightening bias but the stance has since switched to accommodating, as the severity of the deflationary impact arising from the global financial crisis was increasingly recognized (Figure 10). The policy rate has been cut by 425 basis points to ½ percent. Against this background, the freely floating Krona depreciated significantly. The authorities have also taken a wide range of measures to stabilize financial markets (Box 4).

Figure 10.
Figure 10.

Monetary policy, 2003–09

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Sources: Thomson Financial/Datastream, Bloomberg, Haver, and Riksbank.

And weaker activity was quickly reflected in budget revenue

27. Against this backdrop, a shortfall in revenue lowered the general government fiscal surplus from 3.8 percent of GDP in 2007 to 2½ percent of GDP in 2008. While expenditure rose broadly in line with a planned increase of 4 percent, revenue rose just 1.6 percent (compared with the budget-envisaged 4.1 percent). In addition to automatic stabilizers, the significant decline in asset prices notably in stock market indices following the global financial crisis took their toll on corporate income and capital income tax that had been particularly buoyant in recent years. Excluding the cyclical and one-off factors, however, given a small opening of the output gap, the structural balance is estimated to have improved slightly in 2008 to about 2½ percent of GDP. Public debt stood at 38 percent of GDP at end-2008.

And discretionary policies will weaken fiscal outturns further in 2009

28. With automatic stabilizers at full play together with fiscal measures, a large fiscal stimulus—6.6 percentage points of GDP deterioration in the headline budget balance—is expected.

Actions Taken to Stabilize the Financial Sector

Measures to increase Krona liquidity

  • Since October 2008, the repo rate was cut by 425 basis points to ½ percent.

  • The Riksbank started fully accepting covered bonds and lowered the minimum credit rating requirements for long-term securities pledged as collateral.

  • The Riksbank set up new 3, 6, and 12-month loan facilities to facilitate banks’ access to longer-term funds.

  • The Riksbank established a new temporary credit facility using commercial paper as collateral (with a maturity of up to one year) to facilitate the supply of credit for non-financial companies.

  • The National Debt Office issued treasury bills and invested the funds raised in covered bonds to boost covered (mortgage) bond market.

  • The Riksbank granted emergency liquidity assistance facilities to Kaupthing Bank Sverige AB and Carnegie Investment Bank AB (both SEK 5 billion). Later, Kaupthing Bank was liquidated, while the licensing of Carnegie Investment Bank BA was revoked.

Measures to increase foreign exchange liquidity

  • The Riksbank and U.S. Fed set up temporary reciprocal swap facilities ($30 billion). A separate swap facility was also established with the ECB.

  • New dollar term loan facilities (with the maturity of 28 and 84 days) have been offered.

  • The Riksbank began to restore the level of the foreign currency reserve funded by borrowing by the National Debt Office of equivalent SEK 100 billion.

Measures to support banks’ capital and assure market confidence

  • The government increased the deposit guarantee from SEK 250,000 to SEK 500,000, and extend the coverage to include all types of deposit in accounts (October 6, 2008).

  • In October 2008, the government approved a debt guarantee scheme for the medium-term borrowing of banks and mortgage institutions. The total amount of guarantee was set at SEK 1.5 trillion, of which a maximum of SEK 0.5 trillion would be used for covered bonds with the maturity of 3 months -5 years. An institution applying for the guarantee would pay fees and be subject to restrictions on remuneration for senior management. Two financial institutions have availed themselves of this guarantee scheme.

  • The bank recapitalization scheme is intended for banks and other credit institutions. The government’s capital takes the form of shares or hybrid capital (Tier 1 capital). Participating institutions are subject to restrictions on remuneration for senior management.

  • In October 2008, the government enacted “Government Support to Credit Institutions Act” which gives the National Debt Office power to take over a troubled bank if there is a serious systemic risk and bank capital falls below 25 percent of the regulatory requirement (¶67).

  • The government set up a stabilization fund to finance government measures to support the financial system (the sources of the funds are annual fees from banks and other credit institutions).

Staff has estimated the net expected cost of all these actions at 7.7 percent of Swedish GDP.

III. Outlook and Risks

External factors continue to weigh on near term prospects

29. Given output and exports dominated by capital goods and consumer durables, Sweden may be both an early and chronic victim of the global recession, if global demand for those goods recovers slowly relative to other components of global demand. This may be compounded by exposure to the Baltics, and the likely slow recovery of global wholesale markets on which Sweden’s banks have hitherto depended heavily.

30. Imports of Sweden’s trading partners grew on average over 8½ percent annually between 2004-07, by just 2½ percent in 2008, and are projected to fall 11 percent in 2009, recovering significantly only in 2011. Given strong US and Eurozone spillovers, this will weigh heavily on the Swedish outlook (Attachment I). Accordingly, income, unemployment, and confidence effects will depress private consumption growth through 2009—which is not expected to return to positive territory until early 2010, and only to make a fuller recovery well beyond that. This, and the outlook for exports, is expected to be reflected in a fall in investment through 2009, with recovery beginning in the second half of 2010. These developments will also be reflected in falling imports, shielding activity somewhat, though the trade balance will likely deteriorate in both 2009 and 2010.

31. Consequently, though the authorities were somewhat more optimistic, quarterly GDP will likely fall through 2009—with annual activity projected to shrink by 6 percent—before recovering sufficiently to yield a flat path in annual terms by end-2010 (Text Table; Table 2; ¶61). Recovery in the global economy post-2010 is anticipated to result in a quicker growth in exports and domestic demand beyond 2011, raising GDP growth up to 3¾-4½ percent during 2012-14. But on the external side, weakness in the Baltics is expected to permanently lower income flows relative to the recent past.

32. Headline annual CPI inflation will be considerably lower through 2009 reflecting weak demand, the impact of lower interest rates on housing costs, and global energy and food price developments, despite the impact on prices from the Krona depreciation. A gradual increase—corresponding to that in demand—is expected in 2010.

Sweden: Near-term Economic Projections; 2009–10

article image
Sources: Statistics Sweden; and IMF staff projections.

Various downside risks are associated with this outlook

  • Prolonged global financial market strains, even weaker-than-projected export markets, and worse Baltic outturns.

  • Weakening housing and property markets, along with rising unemployment, would decrease collateral values and the supply of credit-worthy borrowers.

  • Corporate financial positions are weakening. Since 2007, companies’ capacity to cover interest payments has deteriorated.

But these are mitigated by various factors

  • Global recovery could be faster than projected.

  • Stabilization efforts in Baltic countries could succeed, with stronger-than-projected activity, reducing associated uncertainties.

  • The significant fiscal and monetary relaxations, including via a significant depreciation, could cushion the projected downturn in 2009.

And the outlook for potential output is unclear

33. Recent Swedish global integration has raised growth and amplified the volatility of output. Correct identification of the trend and the cycle is key to projections of the near- and medium-term prospects (Box 5).

Prospects for Sweden’s potential output

Conventional approaches—HP filtering of real GDP or simple estimation of Solow residuals in Cobb-Douglas framework—may malfunction given the unique depth of the ongoing recession. In particular, application of the HP filter to the Cobb-Douglas production function does not capture the fall in trend capital accumulation and trend growth in total working hours that may accompany the current downturn—as they did during the early 1990s recession. Consequently, it assigns greater part of the volatility in real GDP to the cycle, and hence, potential GDP exhibits the least correlation with output over the last decade under this approach implying scope for a large recovery from the present downturn. At the opposite extreme, simple HP-filtering of the real GDP series may deliver an excessively volatile trend estimate.

To address these issues, two further exercises were conducted. First, the production function approach was modified to allow for a deterioration in trend capital accumulation and in trend growth in total working hours of magnitudes comparable to the early 1990s over the forecast horizon. These ad hoc adjustments likely provide a lower bound on trend growth in factor inputs, as reforms to labor market structures and income taxation in recent years is anticipated to limit labor market hysteresis relative to the 1990s.

Second, a trend line was fitted to the Solow residuals from a Cobb-Douglas decomposition of real Swedish GDP, and used to estimate potential output. The raw TFP estimates indicate various breaks: (i) in 1975, possibly reflecting the global oil shock; (ii) in 1995, possibly reflecting the banking crisis and associated institutional changes (e.g., floating currency, labor market reforms enhancing productivity); and (iii) in 2006, possibly an early sign of the unsustainable global boom then underway. So the linear representation of TFP trends, with turning points reflecting these 3 breaks, was constructed and the resulting series applied together with labor and capital inputs to obtain a trend GDP series (Figure 11).

Figure 11.
Figure 11.

Sweden: Potential Growth and the Cycle, 2000-14

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Sources: Eurostat, Haver, OECD, Statistics Sweden, and IMF staff calculations.

These exercises underscore the fragility of the standard—and indeed other—approaches to estimating potential output now. And even though most of them suggest medium-run potential growth in the region of 2 percent, the uncertainty surrounding all the measures suggests that this “average” is not a strong result. In the short term, both additional exercises described yield estimates of the output gap which lie between the extremes yielded by the standard methodologies. Finally, because the segmented time trend approach seems to best describe developments during the 1990s recession, that approach has been used as the baseline case.

IV. Implications for Policies and Policy Frameworks: 2009–10

34. Even with domestic policies well cast, prospects are gloomy for an early end to the recession that is underway. And any case for more stimulative countervailing domestic policies should carefully weigh risks that they may backfire and so compound the downturn.

35. In that context, the immediate priority is to ensure that the financial stability framework is ready to address downside tail risks should those materialize. Alongside, with the medium-term focus of fiscal policy remaining strongly credible, the operation of automatic stabilizers should proceed unimpeded. Furthermore, with inflation pressures abruptly diminished, support for activity from monetary action—both conventional and, if needed, unconventional—should remain aggressive. These steps will cushion the downturn, and buttress resilience to any additional shocks.

A. Financial Sector Policy and Framework

36. Challenges to financial sector stability derive from exposure to Baltic countries whose EU and Euro convergence aspirations have faltered, a severe Nordic (including domestic) economic recession, and banks’ reliance on global wholesale markets—now impaired—for their financing. All three interrelate, through market assessments of bank capital adequacy. And the framework for handling crises requires further strengthening (Boxes 6 and 7).

37. The authorities’ actions have addressed the immediate fallout from these strains (¶66). Official liquidity support has cushioned the loss of wholesale finance; the increase in bank deposit insurance limits has helped stabilize the deposit base; one domestic bank has been resolved; others have raised capital; and the authorities’ commitment to “do what is needed” has calmed both deposit and interbank markets. With the market still exposed to further possible shocks, including from abroad, as the authorities pointed out, that leaves the challenge of how and when to strengthen banks further, how to ensure that the flow of credit does not become a bottleneck to growth, and how to eventually exit from these extraordinary measures (¶67).

Recent stress tests follow best international practice

38. The most recent step addressing these challenges was publication of stress tests by the Riksbank and the FSA respectively. Both base case scenarios reflected 4½ percent GDP declines in 2009, and elevating credit losses in the Baltics, with probabilities of default and losses given default ratios based on past Swedish and international patterns. The stress scenarios both reflected expert adjustments to those ratios, consistent with assumed GDP contraction at twice the base case rate, along with impaired earnings. Under these assumptions, both sets of stress tests found that all banks observed the 4 percent Tier 1 capital requirement even in the stress scenario. The severity of the underlying assumptions used, the detailed institutional analysis, and the transparency and speed with which the results were reported all reflect best practice. But market concerns about the adequacy of Swedish banks’ resilience remain, as indicated by bank stock prices, CDS spreads, and the continued freeze in the interbank market, although financial prices have eased from their troughs of a few months ago, in the context of the stabilization of international financial markets.

This indicates, as elsewhere, that more needs to be done

39. Ultimately, banks’ business models, contingency plans, and capital ratios should all signal both institutional and market resilience, even in the absence of extraordinary stabilizing support. And significant steps towards such resilience would best be required over a relatively short time horizon. Early action—with due regard to costs—would reduce vulnerability to short term shocks, risks of curbed credit supply, contingent claims on taxpayers, and it would anticipate eventual exit from the extraordinary support measures.

40. As part of this, work should continue with all banks to ensure appropriate contingency plans. Enhanced supervisory scrutiny is the appropriate first response to institutions most at risk. Using Basel’s Pillar II approach, such banks could also be assigned a higher capital requirement, which they would be expected to meet over time including through new rights offerings and (as a bridge) via the government’s capitalization program. This approach would not be procyclical. Market concern with banks’ robustness constitutes the binding constraint on banks’ credit extension decisions now. Swift relief of that constraint, backed by public injection of equity where necessary, will boost credit prospects. The authorities are proceeding cautiously (¶67).

41. Such a strengthened preemptive strategy will require continued close cooperation of all the domestic and host countries authorities of foreign subsidiaries of Swedish banks. And to address risk that the strategy is overtaken by events, such as further bouts of investor uncertainty, contingency plans should be developed alongside in coordination with relevant regional authorities. With banks overtly stronger and contingency plans strengthened, resilience would be enhanced. Beyond this, the institutional capacity of the FSA should be boosted, including by prompt determination of strengthened arrangements for its ongoing resourcing, and it should expand its mandate to deposit-taking non-bank financial institutions.

42. The 2006 FATF assessment report suggested some weaknesses in Sweden’s AML/CFT framework. Among several measures to address such concerns, Sweden implemented the third EU Money Laundering Directive on March 15, 2009.

Is Sweden Ready to Manage Financial Instability?

Some elements of the financial stability framework could usefully be strengthened.

  • Bank resolution framework. In October 2008, the government enacted “Government Support to Credit Institutions Act” which gives the National Debt Office power to grant credit guarantees, and if there is a serious systemic risk and bank capital falls below 25 percent of the regulatory requirement, take over a troubled bank The effectiveness of this new scheme, including how and when to use intervention power, has yet to be demonstrated in practice (¶67).

  • A Prompt Corrective Action framework is needed, covering all banks and fully empowering the FSA to take the full range of supervisory corrective actions.

  • Supervisory capacity. The FSA has been constrained by high staff turnover, and it would benefit from greater funding, including to allow more thorough on-site inspections. Supervisory oversight over the leverage ratio and maturity mismatch could be strengthened.

  • Deposit insurance scheme. Under the current legal framework, deposit insurance funds can only be disbursed after a failed bank is placed into bankruptcy, which could be a lengthy process. Furthermore, the ability of the authorities to obtain relevant information about the balance sheet of a troubled bank at an early stage should be clarified.

  • Cross border coordination. There are major uncertainties in how a cross-border crisis would be resolved in a coordinated way. Memorandums of Understanding on crisis management were signed at the EU level and with three Baltic countries, but generally lack specifics and are not legally binding.

  • Emergency liquidity assistance (ELA). The design of the ELA framework is generally sound to deal with domestic liquidity problems. Nonetheless, how ELA would work for cross-border liquidity problems remain unclear.

  • Non-bank financial institution supervision. A number of non-bank financial institutions raise deposits, but they are not subject to FSA supervision nor reporting requirements.

How do the Baltic Subs Affect the Case to Support Swedish Banks?

Swedish banks’ exposures to the Baltic are extensive. SEB and Swedbank hold significant market shares there (40-80 percent in loan markets and 30-85 percent in deposit markets), and the authorities remain engaged with these activities (¶67).

Official financial cooperation between Sweden and the Baltics is considerable. In December 2006, the Riksbank signed a quadripartite Memorandum of Understanding with central banks of Estonia, Latvia, and Lithuania to set an agreement on cross-border cooperation in crisis management situations. In February 2009, with uncertainties about the external viability of Baltic countries heightened, the Riksbank entered into currency swap agreements with Latvia (the Riksbank to lend up to EUR 375 million) and Estonia (up to SEK 10 billion).

uA01fig02

SEB and Swedbank: Exposures to Baltic Countries (As Of End-2008)

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Sources: Banks’ annual reports; the authorities’ websites; and IMF staff Deposits estimates based on publicly-available information.1/ Deposits exclude non-residential deposits.

This reflects close Swedish-Baltic linkages:

  • Direct financial linkage. Swedish banks’ equity and loan claims on their Baltic subsidiaries at end 2008 represented 8 percent of Swedish GDP, while their loans to their subs amounts to 35-45 percent of bank capital. In addition, Swedish banks’ reliance on operating profits from Baltic operations is extensive (25 percent for Swedbank and nearly 10 percent for SEB). Accordingly, a deterioration in asset quality or profits in Baltic subsidies could present material risks to Swedish banks’ capital.

  • Possible contagion. Retreat from the Baltic subsidiaries could damage the franchise value of the parents, with possible contagion through the Swedish banking system, and the region.

  • Overall macroeconomic stability. A disorderly retreat from the subsidiaries could prompt broader loss of confidence in Sweden.

Given these considerations and the risks in the Baltics, a coordinated regional contingency plan is needed. It should address (i) the division of crisis management responsibility; (ii) methodology for evaluating bank asset quality and diagnosing the viability of institutions; (iii) restructuring policies; and (iv) burden sharing agreements addressing the respective liabilities of the home and host fiscal authorities. The last will need to be carefully considered—given various options that may be considered. A credible contingency plan would help stabilize confidence in the Baltic region and, by extension, in Sweden itself.

B. Fiscal Policy and Framework

The fiscal framework is strong

43. Sweden entered the downturn in robust fiscal health. Debt was low and falling, the framework of rules guiding policy has been consistently adhered to and the authorities have reiterated their commitment to it even in the context of the ongoing global shock. And not only is Sweden one of only a few countries to have completed a full long-term fiscal balance sheet exercise, but of those, it is one of very few to have strong results even under that most testing of examinations.

44. As the authorities underscore, that does not however establish the case for discretionary fiscal activism beyond that which is already planned (¶62). First, uncertainty over the level and path for potential output implies a substantial range of uncertainty around estimates of the structural balance now, and therefore the consistency of the current stance with the fiscal rule (See Box 8). Second, the fiscal stabilizers appear to be large and fiscal multipliers appear to be small, as suggested by the 2-country version of the IMF Global Integrated Monetary and Fiscal Model (See Text Table & Figure and Attachment III). Furthermore, it is unclear at this point how significant the public debt implications will be of any financial sector rescue operation, net of recoveries. And finally, the composition of any stimulus matters.

Budgetary Impact of a 1 Percent Change in GDP

(In percent of GDP)

article image
Source: OECD (2005), using 2003 weight.
uA01fig03

Multiplers in Sweden Using GIMF

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Source: IMF staff estimates.

45. Given the absolute size of automatic stabilizers and multipliers, the potential burden of the financial sector on public debt, and uncertainties about the structural balance, the scale of the authorities’ proposals for discretionary action in 2009-10 has merit. But it will strain efforts to remain compliant with the fiscal rule, and the composition of the discretionary stimulus raises some concerns.

Implications of Various Estimates of Potential Output for the Budget

The wide range of estimates for potential GDP and the output gap imply uncertainty in assessing the underlying fiscal stance—a range of nearly 3 percentage points of GDP for 2008—with associated widely varying implications about the space within the fiscal rule for discretionary stimulus. And trend output may be hit by further external shocks, with staff model-based estimates suggesting that such output volatility could reach ±1.1 percentage points of GDP a year.

uA01fig04

Sructural Fiscal Balance for Various Measures of the Output Gap

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Source: IMF Staff estimates.
uA01fig05

Fiscal Balance: Staff Scenarios

(Percent of GDP)

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

The 2009 budget appropriately anticipates sizeable fiscal support for the economy

46. With automatic stabilizers at full play together with fiscal measures, Sweden is providing large fiscal stimulus to the economy—6.6 percentage points of GDP in 2009, on staff estimates, on of the largest in the EU (See Text Figure). However, even on the authorities’ more optimistic assumptions, the fiscal swings from a surplus of 2½ percent of GDP in 2008 to a deficit of 2¾ percent of GDP in 2009.

uA01fig06

Change in Fiscal Balance 2008-09

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Source: WEO Spring 2009.

47. Much of the turnaround reflects the effects of the automatic stabilizers. With relatively high tax burden to finance the generous welfare system, public finances in Sweden are comparatively sensitive to cyclical fluctuations and labor market developments. The Spring Bill projects a sizeable widening of the output gap in 2009 (to 7.2 percent). In this context, the full operation of the automatic stabilizers accounts for more than half of the deterioration in general government finances.

Sweden: Fiscal Measures 2009–10

(In percent of GDP)

article image
Source: 2009 Budget Bill and 2009 Spring Fiscal Policy Bill.

Includes lower contributions to the unemployment insurance fund, changes in under-pricing rules, changes in interest deductibility, and widening of the CIT tax base.

48. In addition, discretionary fiscal measures are being implemented, amounting to 1.6 percent of GDP. These are mostly on the tax side, including permanent cuts in personal, social contributions and corporate income tax, amounting to 1 percent of GDP. Though these steps will yield supply-side efficiencies and the income tax cut was well targeted to lower income households, their immediate impact on aggregate demand is likely small. In particular, the estimated multiplier for such tax cut is only 0.35, compared with a 0.7 from expenditure measures. In response to further deterioration in the macroeconomic outlook, additional measures were introduced especially to boost active labor market policies.

49. The structural balance—on the authorities’ estimates of potential output—is projected to remain in surplus of at least 1 percent of GDP at all times through the projection period—consistent with the fiscal rule. But the margin for error—or to allow further adjustment if needed—is limited under the rule. On official projections, public debt rises to 46 percent of GDP in 2012 (See Text Table).

50. A larger than projected deterioration in the fiscal accounts in 2009 appears likely, reaching a deficit of slightly above 4 percent of GDP, in part reflecting staff’s more pessimistic view about the near-term growth outlook ¶61). In addition, possible official overestimation of the structural balance in 2008—reflecting staffs more pessimistic view about Sweden’s potential growth in recent years—and the end of “windfall” revenues associated with the boom in asset prices, impacting corporate and capital income, also play a role in this difference of view. This assessment lowers even further the estimated margin for discretionary policy consistent with the fiscal rule.

Sweden: Comparison of Fiscal Outlook

(Percent of GDP)

article image
Source: 2009 Spring Fiscal Policy Bill and staff projections.

As a percent of potential output.

51. The budget outlook is relatively robust to assumptions about prospects for aging costs (Figure 12). Projections of these costs—2.6 percent of GDP over the period 2007-60 based on the 2009 Aging Report by the European Commission—compare with 5.2 percent for Euro area countries). Furthermore, the public sector balance sheet approach assessing the long-run intertemporal financial position based on staff baseline scenario shows positive net worth through 2060, albeit by a small margin (Table 9). Accordingly, debate about changing pension adjustments for asset returns raises concerns for sustainability. The rule requires cuts in nominal pensions in 2010 due to low recent returns. But even a “temporary” suspension of the rule could, de facto, become permanent if it is repeated. Adjusting the formula to smooth the path will address these long-term risks, only if smoothing is symmetric and the formula remains unchanged thereafter.

Figure 12.
Figure 12.

Sweden--Despite the weaker fiscal position, fiscal policies would remain sustainable but with smaller margin to cope with higher aging costs. 1/

2000-60

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Source: Commission Aging Report 2009.Source: IMF staff calculations based on information provided by the authorities.

A change to the expenditure ceiling framework may be needed

52. In the current economic downturn, the established fiscal indicators (the structural balance, 7-year rolling average and the cumulative overall budget balance measures) will malfunction. Even if the structural fiscal balance is maintained at a surplus of 1 percent of GDP, these indicators will flash red simply because headline budget balances, boosted by automatic stabilizers will show such large deficits during the recession. For this reason, these indicators should be deemphasized and an alternative means of showing adherence to the medium-term surplus target may be needed.

53. In this context, the firm commitment to the nominal spending ceilings should be maintained. This would also avoid a structural drift in spending especially when the underlying fiscal position is difficult to measure. In addition, this allows automatic stabilizers to operate unhindered on the revenue side, while accommodating considerable latitude for this on the spending side. However, to ensure that this is consistent with the surplus target over the medium-term, any impact on net revenues reflecting further discretionary tax reforms on projected budget balances need to be offset by an equal downward adjustments in expenditure ceilings. The Fiscal Council could play a useful role in monitoring compliance with the letter and the spirit of these principles. This option remains under review by the authorities (Attachment III; ¶63).

C. Monetary Policy and Framework

Inflationary pressures have abruptly eased

54. Inflation has fallen below the lower bound of the target—2 percent ±1 percent—(Text Figure). GDP is projected to contract further and output gap to widen. Wage growth has continued to slow. Inflation expectations have fallen, but in recent months have risen back to a low range. In this context, though real interest rates measured using the headline inflation rate have remained broadly unchanged, this largely reflects commodities price falls and the impact of the interest rate cuts on housing costs in the headline measure. Real interest rates measured relative to inflation expectations and to underlying or core measures of inflation have, appropriately, fallen significantly.

uA01fig07

Inflation forecast

(Annual percentage growth)

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Source: Riksbank.

An undershoot of the 2009 inflation target is now likely

55. The staff’s baseline forecast—consistent with the authorities—is for a steady decline in inflation through most of 2009, followed by a return to a positive inflation—but still below the lower bound of target—by the middle of 2010 (¶64). This forecast is based on (i) the authorities’ expansionary fiscal policy stance; (ii) subdued commodity process (WEO projections); and (iii) the growth outlook discussed in Section III.

However, the monetary stance should be retained, for now

56. Risks of sustained disinflation appear low. Nominal wage patterns, inflation models, market analysts’ assessments, and short and long-term indicators of inflation expectations, including 5-10 year breakeven inflation rates, all point this way, with many of the latter having rebounded from troughs at end-2008. This partly reflect significant depreciation of the exchange rate since fall 2008, which has, implicitly, relaxed monetary conditions considerably.

57. Accordingly, the case for further immediate steps into Quantitative Easing (QE) is not compelling. As the authorities indicate, only if inflation expectations fall below target on a sustained basis, should consideration be given to the various QE options (Box 9; ¶65). Actions to support individual credit markets, if any of these fail, should remain under review, nevertheless.

Options for Quantitative Easing

  • Outright purchase of assets. In close coordination with the Ministry of Finance, the Riksbank could increase purchase of government bonds at its own discretion. This measure is expected to lower long-term government bond yields. The Riksbank could also assist specific frozen markets, such as mortgages, to restore normal functioning of the monetary transmission mechanism. However, risks to the Riksbank’s balance sheet as well as an exit strategy should be considered, with the government guaranteeing the Riksbank against financial losses, if needed.

  • The usage of the NDO’s treasury functions. The NDO could increase the issuance of treasury bills in exchange of frozen market securities. This option would clarify the division of labor between monetary and fiscal functions.

  • Non-sterilization of foreign asset purchases by the central bank could be considered.

The Inflation Targeting framework serves well

58. The Riksbank is among the most transparent of inflation targeters. It publishes policy committee minutes with two-week lag, a three-year inflation forecast, and corresponding interest rate path. This enhances effective communication. Recently, publication of individual votes of the Executive Board members has begun. However, in 2007, the government-sponsored special commission proposed amendments to the Riksbank Act, aimed at strengthening the Riksbank’s financial independence. Until now, no follow-up actions have been taken. And, as with inflation targeting regimes elsewhere, there may be need in the aftermath of the global financial crisis to review if and how their operations could be further strengthened in light of that experience.

59. The recent decision to raise international reserves by SEK 100 billion—2 percent of total external debt—will boost resilience (See Text Table). As the authorities suggested, although reserves are complemented by a swap arrangement with the U.S. Federal Reserve and ECB, given that domestic and external bond amortization by large banks in 2009-10 is some four times larger than international reserves, additional holdings even beyond those planned would boost capacity to respond to banking sector liquidity stresses (¶65). The need for foreign resources for this purpose arises because destabilizing depositor withdrawals from subsidiaries of Swedish banks could spur similar withdrawals from their parents and from Swedish banks in general, and curb institutions’ access to international capital markets. In this way, banking stresses could spill over into pressures on the Krona. Thus, to support the subsidiaries, the exchange rate regimes of their hosts, and address possible fallout on the Krona, additional foreign currency resources may be required.

Sweden: Official Reserves

article image
Source: Riksbank.

Subsidiaries of Swedbank and SEB.

V. The Authorities’ Views 2

60. Sustained strong policies—as reflected in low inflation, significant current account and budget surpluses translating into long-term fiscal sustainability and low public debt, low unemployment, and structural labor market and tax reforms aimed at increasing supply-side efficiencies—have prepared Sweden well to face the global crisis (¶18–22, 43, and 58).

The economic outlook has weakened significantly

61. Given global conditions and recently revised quarterly national accounts data, the Ministry of Finance in the Spring Bill in April projects output to fall some 4¼ percent in 2009 (¶23–35). Downside risks are contained by expansionary fiscal and monetary policies, the strong automatic stabilizers, Krona depreciation, and the strong and flexible structural characteristics of the economy. In this context, reforms to unemployment benefits, labor market programs, and the tax regime in recent years will reduce risks of labor market hysteresis. But options for further such measures remain under active review.

Fiscal policy—allowing scope for large stabilizers—should remain bound by rules

62. Though stimulus measures predated the intensified global financial crisis from the Fall of 2008, they provided the stimulus needed in the face of the downturn. In this context, given firm commitment to maintain strong fiscal sustainability, no need is apparent now for further broad based discretionary fiscal stimulus, and the fiscal rule—targeting a 1 percent overall surplus over the cycle and respecting the three year expenditure ceilings—remains (¶144–51).

63. The risks of the established fiscal indicators malfunctioning, even if policy remains on course for its medium-term target, are recognized (¶52). But the appropriate option to address this difficulty remains under review. In principle, staffs suggestion of adjusting the expenditure ceiling mechanism would address the difficulty, but such a change runs the clear risk of undermining a long-established and credible anchor of the fiscal framework (¶53). In the authorities’ view, there is no alternative to a continued thorough annual analysis of the budget balance, notwithstanding the uncertainties associated with measures of the structural balance.

Interest rate cuts have successfully averted threat of disinflation

64. A further decline in the headline CPI is anticipated for the remainder of 2009, but underlying inflation (excluding mortgage interest rates) is expected to hover around 2 percent, and inflation expectations continue to be anchored at the target level of 2 percent. The outlook for policy remains accommodative, with the repo rate expected to stay at ½ percent through 2009–10. The impact of monetary actions already taken remains under review (¶ 54–56).

65. Nevertheless, the preparation for possible next steps are underway, partly because term rates remain more elevated than is desired. While the repo rate could be further lowered—the floor is not necessarily at ½ a percent—options for the unconventional measures included the provision of 12 month fixed rate loans at the repo rate and the outright purchase of government bonds. Beyond that, purchase of mortgage or corporate bonds would imply undue credit risks and distort still-functioning corporate credit markets (¶57). In regard to the exchange rate, the freely floating regime will be maintained, but increases in international reserves are warranted, on a pre-announced schedule, to address financial stability risks (¶59). The Riksbank targets inflation, but has no target for the Krona.

Financial sector stabilization policy is comprehensive

66. Liquidity support, the guarantee program, and other steps consistent with the EU response, have contributed to stabilizing financial markets (Box 4; ¶37). However, normal functioning has not resumed and the situation remains fragile. The recent stress tests for major banks show that loan losses are expected to rise, but banks will still meet the minimum regulatory capital requirements even under the most stressed scenarios (¶38–41).

67. The recent bank resolution law empowers the government to take over failing banks, with recourse for the shareholders to the Examination Board of three judges strictly limited to examine the appropriateness of the conditions—the takeover itself is not subject to review by the Examination Board or any other legal authority. Looking ahead, as credit losses become clearer, banks may have to raise capital to fill the shortfalls over time. If needed, banks will be recapitalized through government capital injections of common stock. Swedish banks have made commitments to continue to support their subsidiaries with liquidity and capital. Since the outbreak of the crisis, the authorities have continuously been involved in a dialogue with the major Swedish banks. It should be noted that the measures that the government has adopted have not been ring-fenced and have thus been able to benefit entire banking groups. Looking ahead, the government’s actions will be guided by need to safeguard the legitimate interests of taxpayers (¶39–41).

Table 1.

Sweden: Selected Economic and Social Indicators

article image
Sources: Statistics Sweden; Riksbank; Ministry of Finance; Datastream; INS; and IMF staff estimates.

Includes SEK 100 billion borrowing planned by National Debt Office, at June 15, 2009 market exchange rate.

Based on relative unit labor costs in manufacturing.

Table 2.

Sweden: Medium-term Scenario, 2007–14

article image
Sources: Statistics Sweden and IMF staff projections.

Contribution to real GDP growth.

In percent of nominal GDP.

Annual average, in percent.

In percent of potential GDP.

Table 3.

Sweden: Financial System Structure, 2002–08

article image
Sources: Riksbank, Financial Supervisory Authority, and IMF staff estimates.

Including parent banks’ foreign branches.

Not including minor local companies

Market value of funds

Number of institutions is computed on unconsolidated basis.

Table 4.

Sweden: Financial Soundness Indicators: Banks, 2003–08

(End-period, in percentage)

article image
Sources: Financial Supervisory Authority, Riksbank, and IMF staff estimates.

From 2007, the calculation of capital base follows rules under Basel II.

On consolidated basis

Table 5.

Sweden: Financial Soundness Indicators: Non-Banks, 2003–08

(End-period, in percentage)

article image
Sources: Financial Supervisory Authority, Riksbank, and IMF staff estimates.

End June 2008.

Table 6.

Sweden: Balance of Payments, 2007–14

article image
Source: Statistics Sweden; and IMF staff projections.

Includes investments in financial derivatives.

Positive number indicates an accumulation of foreign assets.

Table 7.

Sweden: International Investment Position, 2007–14

article image
Sources: Statistics Sweden; and IMF staff projections.

Includes investments in financial derivatives.

Table 8.

Sweden: General Government Financial Accounts, 2007–14

article image
Sources: 2008 Fiscal Policy Bill (Fall 2008) and IMF staff estimates.

Excludes debt issued to raise international reserves.

Table 9.

Sweden: Public Sector Balance Sheet

article image
Source: Swedisch authorities; and Fund staff calculations.

Stream of discounted projected future primary fiscal balances under current policies and with the indicated aging costs. Discount rate is the average cost of government funding.

Considers only financial assets and liabilities (i.e. excludes capital stock). This measure is a liquidity indicator whereas the comprehensive net worth is a solvency indicator.

Attachment I. Global spillovers to Sweden, and its external vulnerability.1

1. Sweden’s exposure to the global business cycle—as measured by both, trade and financial linkages—has increased substantially over the last two decades, with a marked acceleration in this process during the recent boom.

2. On the trade front, this is evident from the increase in export and import volumes, both in absolute terms, and in relation to the size of the economy. Activity related to exports contributed, on average, over 35 percent of economic growth since 1990, a substantial increase compared to the 1980s.2 Sweden’s export basket is dominated by goods and services characterized by cyclically sensitive demand.

Sweden: Key Export Sectors, 1990–present

article image
Sources: Eurostat, Statistics Sweden; and IMF staff calculations.

Sweden: Growth in Volume of Trade, 1985–present

article image
Source: Statistics Sweden; and IMF staff calculations and projections.

Constant prices.

Contribution to real GDP growth.

Annual percentage change.

3. Over the last decade, the rapid development of global and EU capital markets and products, and the recovery and growth of its own banking sector since the mid-1990s, has resulted in large financial flows into and out of Sweden. Banks and larger NFCs have increased borrowing from international capital markets, and in turn, this has translated into a substantial increase in overseas assets held by Swedish institutions. For example, investment of financial capital into the Baltic region has increased systematically over the last decade, and particularly since 2005. The resulting income flows have bolstered the current account—and national income—non-trivially during the recent boom.3

Growing financial linkages & income / wealth dependence

article image
Source: Statistics Sweden; and IMF staff calculations and projections.

Claims on, or obligations to non-residents

4. The time pattern in the ebb and flow of financial capital appears sensitive to key global cyclical variables, most prominently, economic growth in the U.S. and the Eurozone.

uapp01fig01

Sweden: Foreign assets and financial inflows, 1998-present

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Sources: OECD, Riksbank, and IMF staff calculations.1/ Annual percentage change

Spillover Channels

5. The primary objective of this note is to quantify the trade and output impact—visually apparent in these GDP dynamics—on Sweden of cyclical fluctuations in its major trading partners. After discussing the various spillover channels, we formally investigate their significance via alternate VAR models.4

6. These trends suggest that global financial and economic cycles may be expected to exert significant impact on Swedish GDP. There are at least 3 channels through which such an impact may arise:

  • Trade Channel: through the direct impact on Swedish exports due to shocks to domestic—hence import—demand, and to GDP more generally, in Sweden’s key trading partners. This includes the Euroarea, the Nordic countries, the U.S., and since 1995, emerging Europe, particularly, the Baltic countries (Text Table).

  • Investment Channel: directly, through cost of capital in international financial markets where major Swedish firms borrow (e.g., cost of equity as proxied by global stock indexes such as the NASDAQ composite), and key input costs (e.g., oil prices). But indirectly also, through the impact on investment of movement in exports.

  • Consumption Channel: through the income channel—depressed exports and investment leading to reductions in wage growth, or worse, to layoffs—and through the wealth channel—equity and bond wealth, which in Sweden are affected by global and regional equity cycles and debt spreads.

Sweden: Direction of Exports, 1990 - present

article image
Source: Direction of trade statistics; and IMF staff calculations.

Nordic countries include Denmark, Finland, and Norway; Baltic countries include Estonia, Latvia, and Lithuania.

uapp01fig02

Synchronization of business cycles

(Annual real GDP growth, in percent)

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Sources: OECD; and IMF staff calculations.

7. A visual examination of Swedish, US, and Euroarea real GDP growth indicates a high degree of synchronization of their business cycles. While Swedish growth appears to comove with that of the Euroarea, both of the latter appear to follow growth dynamics of the U.S. economy (Text Figure).

GDP

8. The informal description above renders plausible, as a starting hypothesis, that Swedish GDP is statistically associated with GDP in the US, the Eurozone, Nordic-Baltic trading partners, global equity indexes—proxied throughout by the NASDAQ composite—(positively), and with oil prices, domestic credit market conditions, and unemployment (negatively).

9. Results are reported for a specification including in reverse order, Swedish GDP, domestic unemployment and basis spread, Baltic GDP, Nordic GDP, Eurozone GDP, and U.S. GDP (Text Figures). For a number of alternative specifications and variable orderings, impulse responses of Swedish GDP are consistently of the expected sign, and quantitatively large and persistent, with respect to innovations in U.S. and Eurozone GDP, and domestic credit spreads. For example, a one percentage point increase in U.S. and Euroarea GDP growth is associated with as much as 3/5 and ¾ percentage point rise in Swedish GDP over four quarters. For others (e.g., domestic unemployment), the impulse response was sensitive to the specification being in-levels as opposed to in-rates-of-change, and for some (e.g., oil prices, Nordic-Baltic GDP), impulse responses were consistently quantitatively negligible.5

However, in a manner most likely reflective of the small sample size that begins in the 1990s (1995Q1), the impulse responses are statistically insignificant for almost all specifications, including those covariates for which they are quantitatively large (Text Figure).

uapp01fig03

Impulse Responses of Swedish GDP Growth to External Covariates

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Source: IMF staff calculations.Note: Impulse response (in percentage points) of real GDP growth (quarterly) to a percentage point increase in covariate.
uapp01fig04

Impulse Response of Swedish GDP to Selected Domestic Covariates

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Source: IMF staff calculations.Note: Impulse response (in percentage points) of real GDP growth (quarterly) to a percentage point increase in covariate.

10. The negative results for Nordic and Baltic GDP may be partly explained by the fact that the covariates are designed to be the component of these economies’ GDP that is “not driven by” U.S. GDP, and it may well be the case that purely idiosyncratic economic shocks to these small, open economies may not produce a tangible cyclical response.

Trade

11. GNP (or GDP) growth in trading partners may affect exports through both consumption demand—as captured; for e.g., by retail sales—and investment demand. Credit market conditions that affect production costs of exporters is an additional factor affecting exports.

12. Impulse response analysis suggests strong and persistent positive association between the level of Swedish exports and level of retail sales in the U.S., the Eurozone, and the Nordic countries. 6 While the impulse responses of Swedish exports to innovations in domestic fixed investment in the Euro zone and the US are quantitatively large, they are statistically insignificant. The strong response of Swedish exports to Nordic retail sales—uncorrected for the impact thereon, of the U.S. consumption cycle—taken together with the absence of a corresponding response of Swedish GDP to idiosyncratic volatility in Nordic GDP indicates that the trade effect may reflect global interlinkages rather than a Nordic-Swedish linkage per se.

uapp01fig05

Impulse responses of Swedish exports to retail sales in selected trading partners 1/

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Source: IMF staff calculations.1/ Percentage change in real GDP in response to percentage change in covariate.
uapp01fig06

Impulse Responses of Swedish Exports to Domestic Fixed Investment in Trading Partners

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Source: IMF staff calculations.Note: Impulse response (in percentage points) of real GDP growth (quarterly) to a percentage point increase in covariate.

Domestic demand

13. Formal statistical analysis revealed an absence of robustness of estimated impulse responses of both, consumption and investment, to innovations in the covariates. However, impulse responses of Swedish consumption and investment indicate no significant direct impact of any of the external covariates.

Attachment II. Have the right lessons been learned from the banking crisis of the 1990S? 1

1. Sweden’s approach to its banking crisis of the early 1990s—generally perceived as “quick action, bank nationalization, bad banks, and transparency—is widely viewed as a success. Many suggest that this provides useful lessons for global policy now. However, this review suggests that the lessons may be richer and more subtle than that.

A. The Roots of The Early 1990S Banking Crisis

2. Financial market deregulation spurred a lending boom; this was compounded by pro-cyclicality in macroeconomic policy; and the fixed exchange regime aggravated financial system vulnerabilities.

  • Financial liberalization through the 1980s lifted credit constraints, resulting in a major boost to bank lending (Drees and Pazarbaşioĝlu, 1998, and Attachment II Figure 1). Banks focused on market share rather than profitability, alongside weaknesses in risk management, governance, and internal controls (Ingves, 2002).

  • Bank supervision lacked authority and capacity, and supervisory frameworks were unprepared for deregulated markets (Ingves, 2002).

  • Deregulation led to a real estate bubble in the late 1980s, which further fueled credit expansion through increased collateral value (Lundgren, 2009).

  • Monetary policy options were limited under the fixed exchange rate regime, and the expansionary macroeconomic policy stance led to higher inflation by early 1990. Meanwhile, the real exchange rate followed a steady upward trend.

Appendix II Figure 1.
Appendix II Figure 1.

Sweden: Macroeconomic Circumstances in the Early 1990s Crisis Periods, 1980–95

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Sources: IMF WEO and International Financial Statistics and Haver.

3. In 1990, the economic boom ended, and strains in the financial system began to emerge. Global interest rates rose in the wake of German reunification, domestic recession began, real estate prices fell sharply, and speculative attacks began on the Krona. In response, monetary policy was sharply tightened, compounding activity weakness. In the Fall of 1990, financial companies faced financial difficulties, and a year later, two large banks, Nordbanken (state-owned) and Första Sparbanken, announced difficulties in meeting prudential capital requirements. The government injected capital to the former and provided a loan guarantee to the latter. By mid-1992, volatility in foreign exchange and money markets had heightened significantly, and increased numbers of banks and mortgage companies faced financial difficulties. International capital markets were all but shut down for Swedish banks and firms (Appendix II Box 1).

Sweden: Major Events of the Swedish Banking Crisis in 1990-1996

article image

B. “Swedish Model”

Banking crisis resolution policies

4. By mid-1992, investor confidence had plummeted. To head off the risk of a fullblown banking and BOP crisis, in September 1992, the authorities announced a blanket unlimited guarantee to all creditors and depositors, except share capital and perpetual debentures, covering the entire banking system. This was followed, on December 18, 1992, by a new bank resolution framework bill, “Measures for Strengthening the Financial System,” a watershed move toward resolving the banking crisis.

The key resolution measures taken:

  • The Bank Support Authority (BSA). The government created a new agency—separate from the Ministry of Finance (MOF), the Riksbank, and Financial Supervisory Authority (FSA)—in order to avoid capacity constraints and conflict of interest issues in these other organizations, and in recognition of the unique nature of the task of bank resolution. Parliament gave the BSA an open-ended budget for rescue operations. Furthermore, politicians left most of the implementation to technocrats.

  • The upfront recognition of expected losses. Two strategies were initially considered: one involving deferring losses and gradually writing off the loss making assets using bank profits; the other rapid evaluation of the financial conditions of each bank over a horizon of three-to five years, to estimate the capital shortage in banks—a de facto stress test. Sweden took the latter approach to restore confidence quickly.

  • Bank classification. Based on the stress test results, the BSA grouped the banks into three categories, Category A (strong banks with regulatory capital adequacy ratio above eight percent), Category B (banks whose regulatory capital adequacy ratio could temporary fall below eight percent), and Category C (nonviable banks to be resolved).

  • Bad bank strategy. The government took control of a Category C bank (Gota Bank) and splits their assets between good and bad assets. The evaluation of assets was assisted by the Valuation Board (a body of expert auditors set up by the BSA) and carried out purposefully to err on the side of undervaluation—but this valuation of these assets by the BSA set floor prices in the market, contributing to maintaining market liquidity (Ingves and Lind 2008, and Jonung, 2009). The bad assets were transferred to an asset management company, established for each of the troubled banks. The bad-bank approach was considered to be most cost efficient mainly because it allowed specialization in the associated tasks.

5. In sum, the essence of the Swedish model was its comprehensive and systematic nature. Sweden legislated a special bank resolution bill, which gave the authorities power to evaluate the financial condition of troubled banks in a forward looking manner, to force bank shareholders to recapitalize banks; and if necessary to take control, including nationalization. Parliament also authorized the BSA to provide an unlimited guarantee to the banking system, and political influence in the individual bank workouts was successfully held at arms length.

End of the banking crisis

6. As soon as early 1993, confidence in the financial system began recovering. The financial position of the banks stabilized, with profitability and assert quality improved and loan losses decreased. At the beginning of 1993, Swedbank and SEB entered into discussions with BSA about the possible rescue plan, but subsequently, their financial difficulties were solved without public assistance as their shareholders injected additional capital. Through the resolution period, overt disruption in the financial system was avoided, though credit volumes dropped sharply.

7. Four major banks availed themselves of the government rescue package. Most of the government’s assistance was concentrated in two of them, Nordbanken and Gota Bank, and their associated asset management companies (Securum and Retriva, respectively).

Sweden: Government funds used in rescue operations

(In millions of kronor)

article image
Source: Ministry of Finance, 1995, “Ending the Bank Support.”

8. In total, the government’s commitment amounted to SKr 88 billion (5 percent of GDP),9 and the government initially anticipated that the resolution operations would take 10 to 15 years (Attachment II Table 2). However, not all government guarantees committed were paid out, and total budgetary costs of rescue measures were smaller at SKr 65 billion (4 percent of GDP). In addition, much of these costs were later recovered through dividend payments, privatization receipts, and sales of bad assets at prices higher than anticipated.

9. In July 1996, the government announced the end of the banking crisis. It annulled the 1992 bank resolution bill, removed the blanket guarantee, and dissolved the BSA (transformed to the Deposit Guarantee Board).

C. Discussion

10. The usual summary of the essence of the Swedish model greatly oversimplifies a much richer picture. Furthermore, key issues in assessing the implications of that model for global policymakers now—was it the small size of the underlying financial sector credit losses, the macroeconomic content, or the financial sector stabilization policies which underlay “success?”—have to be carefully considered.

Two years elapsed before decisive action.

11. The initial signs of financial system strain emerged in Autumn 1990, but until the end of 1992 (the passage of the bank resolution bill), the government relied on ad-hoc individual bank support measures. Nonetheless, this time lag was shorter than elsewhere (e.g. Japan in the 1990s) and, in retrospect, intervention came quickly enough to contain crisis, as reflected in the absence of disruption in the banking system disruption and modest fiscal costs. In addition, to bridge the gap between diagnosis and resolution, the government also introduced a blanket guarantee, which sustained market confidence in the interim.

Nationalization was not the centerpiece, while the bad bank approach proved effective.

12. Only one bank, Gota Bank, was nationalized. The other troubled bank, Nordbanken (Category B), was already partially owned by the government prior to the onset of the crisis.

13. The bad bank strategy worked well for the efficient recovery of bad assets. The relatively homogeneous nature of those assets (mainly bank loans related to commercial property development and residential mortgages) apparently helped rapid valuation. The amount of bad assets to be transferred to a bad bank was carefully calibrated to ensure a level playing field between troubled (Category C) banks and other healthier banks. An immediate sale of bad assets was considered to be more costly given the risks of fire sale at very low prices and lack of securitization markets in Sweden at that time (Thomas 1998).

Transparency served well and regulatory forbearance was avoided.

14. The transparent upfront recognition of loan losses, based on the purposefully cautious assessment of expert auditors, helped allay markets concerns and restore confidence in the banking system promptly.

The modest scale of underlying credit losses may have allowed a bold and transparent strategy to work.

15. The Swedish crisis was smaller than many others. Nonperforming loans peaked at 13 percent of total loans in Sweden, compared to the average of 25 percent for a sample of 42 crisis episodes (Laevan and Valencia, 2008). Gross fiscal costs totaled just 4 percent of GDP in Sweden, and after dividend payments and privatization receipts, net costs were much smaller.10 By contrast, other countries bore much higher fiscal costs. The larger scale elsewhere may have been key in impeding decisive and transparent intervention.

The macroeconomic environment may have underpinned successful bank resolution.

16. Fear of a full blown twin crisis—balance of payments and banking—in mid 1992 prompted comprehensive macroeconomic adjustment, with measures going well beyond the banking sector. Following the abandonment of the fixed exchange regime, the Krona depreciated more than 30 percent in 1993. Monetary policy was eased, and interest rates fell, supporting aggregate demand. On the fiscal front, the authorities consolidated, including: tighter rules on transfers to households, increases in income taxes, social security fees, and employee payroll taxes, all aimed at achieving fiscal balance.

17. With these policy shifts, Sweden also reaped the benefits of an ongoing global economic recovery around 1992-93. The exchange rate devaluation improved external competitiveness, and with strong global demand, exports quickly rebounded in early 1993, and GDP growth resumed from the middle of 1993 (Appendix II Figure 2). In contrast, despite substantial measures taken in the banking sector, bank lending did not recover until the middle of 1995. This suggests that economic recovery helped banking resolution, rather than vice versa.

Appendix II Figure 2.
Appendix II Figure 2.

Sweden: Economic Indicators Around Crisis Period January 1990 - December 1996 1/

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Sources: IMF International Financial Statistics; Datastream; Haver; and OECD Stat.1/ The vertical bar indicates December 1992 when the authorities enacted a bank resolution bill.

D. Lessons from the earlier experience for global—and Swedish—policymakers now

18. Determination of the appropriate lapse of time before decisive fundamental corrective action is taken should balance two considerations—difficulties should not fester overlong, but adequate time for diagnosis and preparation is required. It is notable that the lapse of time between Fall 2007, when the current global crisis broke out, and now, is shorter than the lapse of time before decisive action in Sweden in the 1990s. From that perspective, global policymakers have moved promptly in the current context.

19. Blanket guarantees can play a useful role in bridging the lapse of time between diagnosis and action, when they are credible. Credibility requires both strong fiscal backing, political will, and an exit strategy—in particular to ensure that they do not accommodate undue delay in corrective action. Though the EU has now abjured their explicit use on competition grounds, implicit guarantees are widely perceived and have achieved similar stabilizing effects. This lesson from Sweden appears, to that extent, to have been learned. And while many decisive corrective steps have been taken globally, it remains to be seen if the premium on avoiding undue delay in corrective action is fully reflected in policy now.

20. Similarly, liquidity provision to troubled banking systems can be key to preserving stability, but may also thereby inadvertently impede consensus building on measures to correct the underlying problems. Sweden in the 1990s was able to circumvent this problem because the macroeconomic stress signals were so transparent that political will for strong action was roused promptly—and action to restructure banks was tied to those macroeconomic coat-tails. Globally, now, similar factors may appear to be at work with macroeconomic stress indicators prompting calls for substantial change in banking practices. But, as above, it again remains to be seen if, like Sweden in the early 1990s, those calls for change can be effectively channeled into action to remedy the underlying problems in the financial sector.

21. Context also appears to be critical. Though in the heat of the moment, the size of underlying credit losses is typically poorly known (indeed, that is often a key root of crisis), a crisis with small underlying credit losses will clearly be more manageable. In that sense, the Swedish experience does not afford clear lessons for all other cases: where the underlying losses are much greater, acting decisively and transparently would be a greater challenge.

22. As particular examples of the role of contextual issues, consider which assets should be transferred into bad banks, and at what price. The former requires detailed “bottom up” information about loan portfolios, which is not always available quickly, especially if banking sector problems are widespread across many asset classes and institutions, rather than, as in the earlier Swedish case, concentrated in domestic commercial and residential estate. And all authorities facing financial sector emergencies face conundra with pricing of assets—be it plummeting real estate prices in 1990s Sweden, or structured assets for which markets have disappeared now. But whereas the Swedish 1990s case suggests the benefits of purposeful low-pricing of bad assets, to identify capital shortages robustly and drive efforts to fill those gaps, it is striking that in the present global crisis, neither the Swedish nor the global authorities have taken that approach. At least as concerns their stress testing, central estimates of losses are reported in each scenario, rather than the losses one or two standard deviations below the central estimates for each of the stress scenarios that are considered. This approach may be holding back efforts to drive capital raising efforts now.

23. And the apparently dominant role of macroeconomic imbalances in the genesis and resolution of the 1990s Swedish banking crisis may also qualify broader applicability of the remedial measures for the financial sector used in Sweden then. In essence, macroeconomic policies (including depreciation) rather than bank resolution actions may have been largely responsible for the earlier Swedish success, and the bank restructuring elements may not be so applicable in other contexts. On the other hand, if the global economy recovers soon, the context for resolving global financial sector strains will be greatly eased. Perhaps an implication from Sweden in the 1990s is that macroeconomic problems have to be addressed before those targeting the financial sector can be fully effective, whether or not the financial sector was at the root of the macroeconomic difficulties.

24. The 1990s Swedish experience also yields negative lessons. The abolition of the special bank resolution framework after the crisis ended left Sweden unprepared for strains which emerged following the Fall of 2007. Furthermore, the difficulties in the regulatory framework, which had contributed in part to the genesis of the problem, remained even though they were identified by an official review which followed the crisis.

25. This raises a fundamental question for global policymakers—to understand why lessons which are learned during a crisis are then forgotten. The 1990s Swedish case seems to underscore clearly that as crisis eases, political focus moves elsewhere. This seems to imply that actions to correct long-run weaknesses in the financial stability framework should be taken immediately, even while efforts are focused on containing the immediate crisis. In the case of Sweden now, this consideration points to urgent need to reform the resourcing arrangements for the supervisor.

26. And even so, in light of current Swedish banking strains, the final implication for global policymakers suggested by the 1990s Swedish case is that “never again” is a hope too far. Banking and financial crises will recur.

Attachment III. Should the Fiscal Rule Remain?1

1. Since its introduction in 2000, the fiscal rule has played a key role underpinning Sweden’s fiscal framework. The main fiscal target is to achieve a general government surplus of 1 percent of GDP over the business cycle. Achievement of this objective is supported by expenditure ceilings (three years ahead) for central government outlays (excluding interest payments) and social security spending and by a requirement that local governments maintain balanced budgets. The framework aims at strengthening public finances to provide a buffer to cope with economic fluctuations and the impact of demographic changes to ensure long-term fiscal sustainability and intergenerational fairness.

2. To avoid problems in precisely dating business cycles (especially ex-ante), the Swedish authorities have adopted three indicators to assess actual budget performance relative to the fiscal target. These indicators are (i) the average overall budget balance since 2000, the year in which the target was first applied; (ii) a seven-year rolling average of the overall budget balance centered on the current year; and (iii) the structural overall budget balance, the budget balance adjusted for cyclical and other one-off and temporary effects.

3. The firm implementation of the framework has contributed to Sweden’s strong fiscal performance in recent years. Excluding 2002 and 2003, when the economy was affected by a global economic slowdown, the overall budget balance has consistently been in surplus. As a result, public debt declined from 65 percent of GDP in 1999 to 38 percent of GDP in 2008. The strong fiscal performance has been achieved primarily through expenditure reduction rather than higher taxes. Lower social transfers, reflecting a tightening of criteria for sickness and disability payments, were a key contributing factor. In addition, lower interest payments have contributed. The nominal expenditure ceilings, introduced in 1997, also have played a role in restraining spending growth; since being introduced, they have been consistently met.

uapp03fig01

Fiscal Performance

(Percent of GDP)

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Source: Ministry of Finance.
uapp03fig02

Expenditure and Ceilings

(Billions of krona)

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Model Simulations

4. To evaluate the performance of the fiscal rule, simulations are conducted using a two-country “New Keynesian” model with rational expectations.2 The model is estimated for 1993Q1-2008Q4 with Bayesian estimates of the parameters. In the simulations, the model is subject to various dynamic stochastic shocks. Each set of model simulations consists of 1500 simulations that last for 200 quarters.

uapp03fig03

Change in Debt Over a Cycle 1/

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Source: IMF staff estimates.1/ Expressed in gap terms. Deviation from the level consistent with meeting the fiscal surplus target.

5. The first set of simulations suggest that when the automatic stabilizers are allowed to operate fully without any discretionary policy, on average, the 1 percent of GDP fiscal surplus target is met over the business cycle. 95 percent of the time the cumulative debt at the end of each cycle would be within ±4 percentage points of GDP of the debt level consistent with attaining the target precisely over each cycle. However, this does not necessarily ensure long-term sustainability of the fiscal position. The small deviation in debt levels over each cycle could cumulate resulting in a substantial deviation in debt to GDP away from what might be sustainable.

6. Because of the problems in estimating the intensity and duration of each business cycle ex- ante and the need to assess long-term sustainability, a second set of simulations test the performance of the 7-year rolling average indicator. Simulations suggest this indicator is appropriate, especially since the model suggests that the average length of business cycles in Sweden has been close to 7 years. In these simulations, under the full operation of the automatic stabilizers only, 95 percent of the time the indicator is within ±0.5 percent of GDP with 0 average, while the cumulative debt at the end of each simulation (50 years ahead) would be within ±8 percentage points of GDP of the debt level consistent with attaining the target precisely over the simulation period. This is well within plausible tolerance limits.

7. However, Sweden is a small open economy, and its sensitivity to external shocks is enhanced by its relatively large banking sector with significant cross-boarder exposure. Accordingly, a case might be made for building a safety margin into any indicator used as a guide for achieving the fiscal target. In a third set of simulations, a small discretionary fiscal adjustment (½ percent of GDP) is introduced in periods when the economy is above trend and expected to remain so over the succeeding three years. Accordingly, in these simulations, on average the level of debt is 8 percentage points lower at the end of the simulations, eliminating the concern of a debt drift.

The Fiscal Rule and the Current Economic Downturn

8. However, the intensity of the current economic downturn lies well outside the range of average historical experience. For example, historically the 95 percent confidence interval for Sweden’s output gap is ±3½ percent. Staff projects the output gap during the current downturn will reach and remain large at about 9 percent for the next few years. Clearly, the current situation can be considered to be a “tail event”.

uapp03fig04

Distribution of Output Gaps

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Source: IMF staff estimates.

9. In these circumstances, even if the only fiscal action is to allow automatic stabilizers to operate, both the cumulative average surplus and the 7-year rolling average indicators are expected to deviate substantially from the target over the next 5 years. In addition, uncertainties regarding estimates of the output gap at present and going forward make it difficult to estimate the structural budget balance with any kind of reasonable degree of confidence. And finally, trying to meet the fiscal rule could be highly procyclical.

uapp03fig05

Compliance with the Fiscal Target

(Percent of GDP)

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Source: IMF staff estimates.
uapp03fig06

Sructural Fiscal Balance for Various Output Gap

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Source: IMF staff estimates.

10. This situation raises two questions:

  • What is the appropriate fiscal policy?

  • What is the appropriate indicator to monitor fiscal performance?

Appropriate fiscal policy

11. Reflecting the high tax burden in Sweden and a generous social safety net, the automatic stabilizers—the budgetary impact of a given change in output—are relatively large. The most recent OECD estimates (OECD 2005) of the impact of a 1 percent change in output on the budget balance as a percent of GDP (semi-elasticity) is 0.55 for Sweden, compared to an OECD average of 0.44 and an Euro area average of 0.48. In addition, Flodén (2009) estimates the budget elasticity including active labor market measures, the recent cuts in income tax, and regular unemployment benefits shows that automatic stabilizers have remained broadly unchanged at 0.62. Further he finds no clear evidence that the elasticity in Sweden has declined during the last decade (see Text Figures).

uapp03fig07

Automatic Stabilizers: Semi Elasticities

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Source: OECD (2005).
uapp03fig08

Sweden: Budget Elasticities

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Source: Flodén (2009).

12. At the same time, the impact on the economy of discretionary fiscal measures in Sweden tends to be small. Estimates of the fiscal multipliers for Sweden using the 2-country version of IMF’s Global Integrated Monetary and Fiscal model3 suggest that a temporary 1 percent of GDP increase in government investment would boost output by 0.7 percent (compared with 1.2 percent for the United States, for example). Key contributing factors affecting this result are the large share of imports given Sweden is a small open economy, a small share of liquidity constrained households, and large automatic stabilizers. Revenue multipliers are even smaller. A 1 percent of GDP permanent reduction in labor income taxes would only boost output by 0.35 percent in the first year. In contrast, being small and open implies Sweden benefits from measures introduced in the rest of the world. A temporary 1 percent increase in public investment in the rest of the world boosts Sweden’s output by 0.6.

uapp03fig09

Multiplers in Sweden Using

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Source: IMF staff estimates.

13. With stabilizers large and multipliers small, the case for discretionary fiscal action is attenuated. Furthermore, long-term sustainability issues and current risks associated with contingent liabilities in the banking sector counsel further caution.

  • The outlook for productivity growth and potential output is unclear. With a substantial range of uncertainty around estimates of the structural balance, it is difficult to correctly assess the consistency of the current stance with the fiscal rule.

  • Political difficulty in adjusting pensions during downturns in accordance with Sweden’s defined-contribution plan financed on a pay-as-you-go basis raises a concern about the sustainability of the pension fund.

  • The government at present faces considerable contingent liabilities owing to both explicit and implicit guarantees provided to the banking system arising in particular from the banks’ exposures in the Baltics. The heightened risk calls for keeping the fiscal powder dry.

14. Finally, once the uncertainty surrounding Sweden’s growth potential and the cyclical position is reduced, the surplus target will need to be reviewed to ensure long-term sustainability can be maintained.

Appropriate monitoring

15. The key question in the short term is how to try to ensure that there is not a significant deterioration in the underlying fiscal position during this period when it is difficult to measure that position and when using the rolling average and the cumulative budget deficit measures to assess policy could misstate the underlying position. One way to prevent a drift in the fiscal balance would be to retain the expenditure ceiling. However, any discretionary cuts in taxes would have to be offset by equal downward adjustments in the expenditure ceilings.

16. Given the current circumstances, a further extension could be to use the expenditure ceiling as a “guideline” and not as a strict limit. Strictly adhering to the ceiling in the event the recession is worse and more protracted than envisaged could impede the working of the automatic stabilizers on the expenditure side. Also, since the ceiling covers only central government spending, there could be an incentive to meet it by cutting transfers to local governments, compounding procyclical fiscal policy as these entities are required to meet balanced budgets. If this possible “softening” of the ceiling was adopted, any important deviation from the ceiling would need to be accounted for in a transparent fashion. On the other hand, if there is a high risk that a breach of the ceiling could seriously undermine the credibility of the fiscal framework, there may be a case for strictly observing the ceilings. The Fiscal Council could play a useful role in publicly monitoring compliance with the letter and the spirit of the fiscal framework.

Attachment IV: FISCAL Sustainability

Baseline scenario

In the baseline scenario, the underlying fiscal position is projected to improve slightly over the medium-term to about ¼ percent of GDP reflecting the impact of recent reforms to reduce welfare expenditures. After the cyclical downturns expected for 2009-2011, the public debt-to-GDP ratio declines from a high 53.4 percent of GDP in 2012 to 49.4 percent of GDP in 2014. With a primary surplus of about 1½ percent of GDP and a declining public debt ratio in 2014, Sweden’s fiscal position is expected to remain strong.

In the alternative scenarios, a one time shock to contingent liabilities and lower-than-expected output growth constitute key risks to the baseline scenario. Given the relatively low level of public debt (38 percent of GDP in 2008) in the context of strong fiscal performance in recent years, the impact of higher interest rate is projected to be small. In addition, the portion of the government’s liabilities denominated in foreign currencies is small at 20 percent share, which would imply a modest impact from a sudden depreciation in the real exchange rate. A confluence of shocks—lower growth, higher real interest rates and weaker primary balance—would pose some threats to the debt reduction plan but the impact would remain modest at about 4 percent of GDP.

External risks

  • The impact of a permanent ½ standard-deviation shock to the interest rate—a 0.7 percentage point increase from the baseline—would increase public debt by 2 percentage points to 51 percent of GDP in 2014.

  • A one-time 30 percent depreciation in the REER (in 2010) would immediately increase public debt but the impact is not large, by about 4 percentage points to 53.2 percent of GDP. Debt would subsequently fall at a pace envisaged in the baseline to 53.2 percent of GDP in 2014.

  • A one-time 10 percent of GDP shock to contingent liabilities (in 2010) would increase public debt to 55.2 percent of GDP. Assuming the baseline case of speed of debt reduction thereafter, public debt would lower to 58.7 percent of GDP in 2014 (about 10 percent of GDP higher than the baseline).

Domestic risks

  • A ½ standard deviation shock to primary balance, which lowers primary balance by nearly 1 percent of GDP each year during the forecast period, would translate into a more gradual debt reduction going forward. Debt in 2014 would be higher by 4½ percentage points of GDP (to 54 percent of GDP) in 2014 relative to the baseline.

  • A ½ standard deviation shock to growth would bring average projected annual growth down by 0.8 percent. The impact on public debt is substantial, rising it to 57 percent of GDP in 2014 due to adverse debt dynamics.

  • One-quarter standard deviation shocks to growth, real interest rate, and primary balance (i.e., compared to the baseline, lowering output growth by 0.4 percent a year, raising real interest rate by 0.35 percent a year, and reducing the primary balance by ½ percent of GDP a year) would bring public debt to 53.4 percent of GDP in 2014—an increase of 4 percentage points of GDP from the baseline.

Public sector balance sheet

Despite the weaker underlying fiscal position over the medium-term reflecting fiscal measures, Sweden’s current fiscal policies remain sustainable over the long-run. Sweden, like most industrialized countries, faces significant challenges associated with population aging that have significant budgetary implications. However, the projected increase in spending with aging (2.6 percent of GDP) is relatively small compared to other European countries (5.2 percent in the euro area). Using a public sector balance sheet approach to calculate the long-run intertemporal financial position based on the staff baseline scenario suggests that the net worth constraint continues to be met by 2060, albeit with a smaller margin to absorb upside risks to the aging costs.

Figure 13.
Figure 13.

Sweden: Public Debt Sustainability: Bound Tests 1/

(Public debt in percent of GDP)

Citation: IMF Staff Country Reports 2009, 247; 10.5089/9781451836035.002.A001

Sources: International Monetary Fund, country desk data, and IMF staff estimates.1/ Shaded areas represent actual data. Individual shocks are permanent one-half standard deviation shocks. Figures in the boxes represent average projections for the respective variables in the baseline and scenario being presented. Ten-year historical average for the variable is also shown.2/ Permanent 1/2 standard deviation shocks applied to real interest rate, growth rate, and primary balance.3/ One-time real depreciation of 30 percent and 10 percent of GDP shock to contingent liabilities occur in 2010, with real depreciation defined as nominal depreciation (measured by percentage fall in dollar value of local currency) minus domestic inflation (based on GDP deflator).

References

  • Drees, B., and C. Pazarbaşioğlu, 1998, “The Nordic Banking Crises, Pitfalls in Financial Liberalization?” IMF Occasional Paper, 161 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Dziobek, C., and C. Pazarbaşioğlu, 1997, “Lessons from Systemic banking Restructuring: A Survey of 24 Countries,” IMF Working Paper, WP/97/161 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Ingves, S., 2002, “The Nordic Banking Crisis from an International Perspective,” Speech at Seminar on Financial Crises, Kredittilsynet, The Banking, Insurance, and Securities Commission of Norway, Oslo, September 11, 2002.

    • Search Google Scholar
    • Export Citation

  • Ingves, S. and G. Lind 1997, “Loan Loss Recoveries and Debt Resolution Agencies: The Swedish Experience,” in Banking Soundness and Monetary Policy: Issues and Experiences in the Global Economy, edited by C. Enoch and J. H. Green, (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Jonung L, 2009, “The Swedish model for resolving the banking crisis of 1991-93: Seven Reasons Why It Was Successful,” European Economy Economic Papers, No. 360.

    • Search Google Scholar
    • Export Citation

  • Laeven L., and F. Valencia, 2008, “Systemic Banking Crises: A New Database,” IMF Working Paper, WP/08/224, (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation

  • Lundgred, B., 2009, “Testimony of Bo Lundgren, Director General, Swedish National Debt Office, Congressional Oversight Panel,” March 19, 2009.

    • Search Google Scholar
    • Export Citation

  • OECD, 2002, “Experience with the Resolution of Weak Financial Institutions in the OECD Area,” in Chapter IV, Financial Market Trend, No. 82.

    • Search Google Scholar
    • Export Citation

  • Rose, T., 1998, “Managing Bank Crises in Other Countries,” U.S. Federal Deposit Insurance Corporation Symposium, April 29 and 30, 1998.

    • Search Google Scholar
    • Export Citation

  • Berg, A., Karam, P., and D. Laxton, 2006, “Practical Model-Based Monetary Policy Analysis—How-To Guide,” IMF Working Paper No. 06/81

    • Search Google Scholar
    • Export Citation

  • Flodén, M., “Automatic Stabilizers in Sweden 1998-2009”, 2009, SSE/EFI Working Paper Series in Economics and Finance No. 719

  • Girouard, N. and C. André, 2005, “Measuring Cyclically-Adjusted Budget Balances for OECD Countries”, OECD Working Papers No. 434

  • Honjo, K., 2007, “The Golden Rule and the Economic Cycles”, IMF Working Paper 07/199

1/

Paragraph references in sections II-IV refer to the paragraphs reporting the authorities’ views on the relevant topic.

2

Paragraph references in this section refer to the paragraphs reporting staff’s views on the relevant topic.

1

Prepared by Jay Surti.

2

On an average contribution basis, net exports has supplied 0.8 percentage points to annual GDP growth since 1990—36 percent of the total. The comparable numbers through the 1980s were 0.1 percentage points—or 2 percent of the total.

3

The turn-around in net factor income from abroad has—in a longer-term context—also benefited by the secular depreciation in the public sector’s external liabilities.

4

While a large number of potential domestic and external covariates will be identified, the actual specifications will be necessarily parsimonious as the sample size is limited to 1995Q2–2008Q4. Data needs are met via time series information on Swedish domestic demand, trade, unemployment, and GDP (Statistics Sweden); GDP for trading partners in the Baltic region (Eurostat), the Nordic region, the Eurozone, and the U.S. (OECD); retail sales and gross fixed investment in the Nordic countries, the Eurozone, and the U.S. (Haver); Swedish basis spread (Haver); and the NASDAQ composite index (Bloomberg).

5

Specifications wherein the covariates were represented in terms of rates of change did not present problems of data non-stationarity. For the in-levels specifications, vector error correction models were chosen instead of VARs where necessary. GDP for non-U.S. trading partners is taken to mean the component of the data series statistically orthogonal to the U.S. GDP series. Finally, exchange rate effects are eliminated by transforming all observations into U.S. dollar equivalent series using 2000Q1 exchange rates, and level effects by using an index for the in-levels specifications using 2000Q1 as the base quarter.

6

In order to deal with differing orders of integration and across-series cointegration, the retail sales relation was assessed using a vector error correction framework. As eviews does not report standard errors for estimated impulse responses, these are not reported in the chart.

1

Prepared by Kotaro Ishi.

9

GDP is an average for 1993-96.

10

The costs are not calculated on a net present value basis.

1

Prepared by Keiko Honjo.

2

See Berg and Laxton (2006) and Honjo (2007) for a detailed description of the model.

3

See Kumhof, M. and D. Laxton (2007) for more detailed description of the model.

Same Series

Other IMF Content

Other Publishers

  • Collapse
  • Expand
Cover IMF Staff Country Reports
Sweden: 2009 Article IV Consultation: Staff Report; Staff Statement; Public Information Notice on the Executive Board Discussion; and Statement by the Executive Director for Sweden
Author:
International Monetary Fund
Volume/Issue:
Volume 2009: Issue 247
Publisher:
International Monetary Fund
ISBN:
9781451836035
ISSN:
1934-7685
Pages:
89
DOI:
https://doi.org/10.5089/9781451836035.002
Figure 1.

Sweden: The Long View, 1996–2008
Figure 2.

Sweden: The Recent Boom. 2001–07
Figure 3.

Sweden--Fiscal Policies are Sound, 1993–2008

Compliance with the Fiscal Target

(Percent of GDP)
Figure 4.

Sweden: Increasing Openness Over the Boom, 2000–08
Figure 5.

Sweden: External Competitiveness Remains Firm, 1996–2008
Figure 6.

Sweden: Into the Downturn, 2007-09
Figure 7.

Sweden: Selected Financial Indicators, 2005–09
Figure 8.

Performance of the Swedish Banking System, 2007–09
Figure 9.

Sweden: Non Bank Financial Sector, 1990–2008
Figure 10.

Monetary policy, 2003–09
Figure 11.

Sweden: Potential Growth and the Cycle, 2000-14

SEB and Swedbank: Exposures to Baltic Countries (As Of End-2008)

Multiplers in Sweden Using GIMF

Sructural Fiscal Balance for Various Measures of the Output Gap

Fiscal Balance: Staff Scenarios

(Percent of GDP)

Change in Fiscal Balance 2008-09
Figure 12.

Sweden--Despite the weaker fiscal position, fiscal policies would remain sustainable but with smaller margin to cope with higher aging costs. 1/

2000-60

Inflation forecast

(Annual percentage growth)

Sweden: Foreign assets and financial inflows, 1998-present

Synchronization of business cycles

(Annual real GDP growth, in percent)

Impulse Responses of Swedish GDP Growth to External Covariates

Impulse Response of Swedish GDP to Selected Domestic Covariates

Impulse responses of Swedish exports to retail sales in selected trading partners 1/

Impulse Responses of Swedish Exports to Domestic Fixed Investment in Trading Partners
Appendix II Figure 1.

Sweden: Macroeconomic Circumstances in the Early 1990s Crisis Periods, 1980–95
Appendix II Figure 2.

Sweden: Economic Indicators Around Crisis Period January 1990 - December 1996 1/

Fiscal Performance

(Percent of GDP)

Expenditure and Ceilings

(Billions of krona)

Change in Debt Over a Cycle 1/

Distribution of Output Gaps

Compliance with the Fiscal Target

(Percent of GDP)

Sructural Fiscal Balance for Various Output Gap

Automatic Stabilizers: Semi Elasticities

Sweden: Budget Elasticities

Multiplers in Sweden Using
Figure 13.

Sweden: Public Debt Sustainability: Bound Tests 1/

(Public debt in percent of GDP)