Sweden
Financial System Stability Assessment
Author: International Monetary Fund. Monetary and Capital Markets Department
Publication Date:
17 Nov 2016
eISBN:
9781475554618
Language:
English
Keywords:
ISCR; CR; bank; asset; return on equity; bank assets; cross holdings; asset quality; Swedish krona; ECB supervisor; Nordea's assets
Download PDF (1.9 MB)

This paper discusses the findings of the Financial System Stability Assessment for Sweden. The Swedish financial system is large and highly interconnected, putting a premium on the accompanying policy framework. Relative to the size of the domestic economy, the financial system is among Europe’s largest. It features complex domestic and international linkages, reflecting Sweden’s role as a regional financial hub. However, the macrofinancial risks have grown since 2011, for example the rising share of highly indebted households. Stress tests also suggest that banks and nonbanks are largely resilient to solvency shocks, but concerns persist about the ability of bank models to capture unexpected losses.

Abstract

This paper discusses the findings of the Financial System Stability Assessment for Sweden. The Swedish financial system is large and highly interconnected, putting a premium on the accompanying policy framework. Relative to the size of the domestic economy, the financial system is among Europe’s largest. It features complex domestic and international linkages, reflecting Sweden’s role as a regional financial hub. However, the macrofinancial risks have grown since 2011, for example the rising share of highly indebted households. Stress tests also suggest that banks and nonbanks are largely resilient to solvency shocks, but concerns persist about the ability of bank models to capture unexpected losses.

Macrofinancial Setting

1. Sweden’s financial system is large, with assets of about 5.5 times GDP.

  • The banking sector comprises more than two thirds of the financial system, with one global systemically important bank (G-SIB) and three banks designated by FI as Other Systemically Important Institutions. The four banks account for over three-quarters of the Swedish banking market and have extensive reach across the Nordic-Baltic region (Figure 1).

  • The insurance sector is dominated by life insurance companies, many of which specialize in occupational pension insurance. Total insurance assets were about 100 percent of GDP at end-2015, of which 88 percent are in the life insurance sector. There are over 300 insurance companies in Sweden, but most are small local firms. Industry concentration is high.

  • Capital markets are well developed. Outstanding government and corporate bonds are equivalent to about 85 percent of GDP and stock market capitalization is about 140 percent of GDP. Covered bonds account for about 45 percent of the bond market, with government securities comprising an additional 25 percent. The rest is largely made up of highly rated corporates and a small but growing portion of smaller and less highly rated issuers. Foreign investors hold about one-quarter of outstanding Swedish bonds, equivalent to more than 20 percent of GDP. Securities markets services can be easily provided on a cross-border basis.

  • Four critical domestic financial market infrastructures (FMIs) operate in Sweden: RIX, the realtime gross settlement payments system, owned and operated by the Riksbank; Bankgirocentralen BGC AB, which processes a range of retail payments, cash withdrawals, and card payments; one domestic central counterparty, NASDAQ Clearing, which clears exchange-traded financial and commodity derivatives, over-the-counter (OTC) interest rate derivatives (IRD) and repos; and one domestic central securities depository (CSD) and securities settlement system (SSS) for Swedish shares and fixed income securities, Euroclear Sweden. There are no trade repositories. Several FMIs located outside Sweden are also relevant for the Swedish financial system

Figure 1.
Figure 1.

Key Structural Aspects of Swedish Banks

Citation: IMF Staff Country Reports 2016, 355; 10.5089/9781475554618.002.A001

Sources: Mortgage Bond Association; Statistics Sweden; and Riksbank. Source: Riksbank.

2. Sweden’s oversight framework relies on three institutions and the Ministry of Finance. FI is the single supervisor, with a mandate also for consumer protection and macroprudential issues. The central bank (Riksbank) is in charge of monetary policy, systemic liquidity, and payments oversight. The National Debt Office (NDO) acts as the resolution and deposit insurance authority. The Ministry of Finance (MoF) drafts financial legislation. The Riksbank is an independent authority under the Parliament, while FI and NDO are authorities under the government. FI’s and NDO’s budgets are approved by the MoF. The government proposes legislation to the Parliament, which allocates mandates and budgets to authorities such as FI. Based on authorization by the government, FI issues secondary regulation. There are plans in the parliament to review the Riksbank Act, and the MoF is analyzing options to clarify FI’s macroprudential policy mandate.

3. A Financial Stability Council (FSC) was created in 2013. It is chaired by the Ministry of Finance and comprises the Riksbank Governor, FI Director General, and the NDO Director General. It is a forum for, among other things, monitoring financial stability and discussing the need for measures to prevent financial imbalances and crisis management measures. It has no decision-making powers.

4. The authorities have followed up on the 2011 FSAP recommendations. In addition to establishing the FSC, they increased FI’s resources, and introduced a new resolution framework for credit institutions and investment firms. Appendix I reviews the main steps taken since the last FSAP.

5. Since 2011, the economy has regained speed. Following the sharp rebound in 2010, economic activity was flat for much of 2011-13. Growth has accelerated to 4.1 percent in 2015, driven by strong domestic demand supported by exports. Employment growth picked up in 2015-16, bringing the unemployment rate down to under 7 percent in 2016, from around 8 percent in 2011-14. Household disposable income growth averaged to about 3.5 percent per annum in 2012-2015. The public debt was low at 43 percent of GDP in 2015, underpinned by the prudent fiscal framework. Inflation remains below the target, but has risen from its low in 2014 (Table 2).

Table 2.

Selected Economic Indicators, 2013–19

article image
Sources: IMF’s World Economic Outlook; Sveriges Riksbank; Sweden’s Ministry of Finance; Statistics Sweden; and IMF staff calculations.

Data for 2016 is as of August 2016.

Data for 2016 is as of September 2016.

Mortgage rates for new contracts, data for 2016 is as of June 2016.

Data for 2016 is as of Q2 2016.

Data for 2016 is as of August 2016

Based on relative unit labor costs in manufacturing.

Risks and Vulnerabilities in Households’ and Corporates’ Balance Sheets

A. The Housing Market

6. Housing finance creates vulnerabilities for financial stability due to specific features of Swedish residential mortgages, high household debt, and rising asset prices (Figure 2).

  • The stock of mortgage loans granted before June 2016 does not have mandatory amortization. A typical Swedish mortgage contract has a contractual maturity of 30-50 years, with even longer maturities not unusual. In most countries, loans with no amortization requirements are considered high-risk, and are subject to more restrictive lending standards.

  • Household debt has been rising relative to income with new borrowers taking on increasingly high debts. The share of new mortgage borrowers with debt-to-income (DTI) ratios above 450 (600) percent was about 37 (17) percent in 2015, up from 21 (10) percent in 2011. The credit to GDP gap has declined but it is still positive and credit to households is growing faster than disposable income. Households are vulnerable to interest rate increases (70 percent of residential mortgages are based on floating rates).

  • House prices have risen to high levels, slowing only recently. The price-to-income ratio is 40 percent above its 20-year average, among the highest in advanced economies, raising a red flag. Research suggests that house prices are 12 percent above long-run equilibrium (IMF Working Paper 15/276). House price gains provide incentives for households not to amortize loans and take out even larger loans relative to income, aided by longer loan maturities. Mortgage interest rate deducibility and the lack of a property tax further propel house demand.

Figure 2.
Figure 2.

Nordic Countries: Assessment of Cyclical Vulnerabilities

Citation: IMF Staff Country Reports 2016, 355; 10.5089/9781475554618.002.A001

7. The pace of housing completions represents less than 1 percent of the housing stock, lagging behind rising population, especially in urban areas. Restrictions on land acquisition and planning procedures at the municipal level impair the expansion of housing supply, and rigidities associated with rental market controls do not help.

8. FI’s view is that the rising house prices and high household debt do not entail high credit risk for banks, but they do add to macroeconomic vulnerabilities. Swedish mortgages are full recourse loans, most households have high savings, and social benefits are generally sufficient to limit defaults even in less financially secure households. Moreover, loan origination includes stress tests of the adequacy of household buffers at high interest rates. Rather than distress, supervisors are concerned with the impact of households deleveraging if they come under stress, as reduced consumption would impact employment and smaller firms. FI’s stress tests on household level data for new mortgage originations are in line with this assessment.

9. High asset valuations do not necessarily lead to asset price declines, but if a fall were to happen, the corrections could be much larger and damaging, especially given the high household debt. A 20 percent house price decline could reduce GDP by 2.6 percent,1 with a greater impact if it coincided with a global downturn. It would affect banks directly through nonperforming loans (NPLs) and second round effects may arise from the impact of lower domestic demand on enterprises. Given the high interconnectedness within the Nordic-Baltic financial systems, a fall in house prices with an associated loss of confidence in the Swedish housing market collateral could trigger disturbances across the region. The high reliance on wholesale funding and covered bond concentration in bank portfolios would act as an amplifier.

10. Even though households appear resilient, it is challenging to be conclusive about how scenarios of falling asset prices and higher interest rates would play out. Stress tests may not capture all second round effects. Prices of household assets are exposed to market volatility, and there are no data on households’ wealth distribution that can be compared with the distribution of debt burden or used for stress testing purposes. Pre-crisis data suggests that households with higher debt relative to income typically have lower liquid assets.2

11. The authorities have responded to increasing household debt with macroprudential measures focusing on credit supply (Table 3). An 85 percent cap on LTV ratios, adopted in 2010, has not contained indebtedness, given rising prices and the possibility of taking uncollateralized loans above the cap. The lack of mandatory amortization has taken a long time to address given FI’s insufficient mandate. A mandatory statutory amortization requirement became effective on June 1, 2016. The impact of recent macroprudential measures is not yet clear.

Table 3.

Macroprudential Measures Adopted since 2011

article image
Sources: IMF staff based on information from the authorities.

12. The recent amortization requirement and the government’s 22-point proposal for more housing are welcome, but more is needed to address distortions in the housing market. The impact of the amortization requirement is likely to be limited as only new borrowers are affected. The mission recommends:

  • Introducing a cap on the DTI ratio to increase household resilience and prevent higher housing prices from driving up household indebtedness. Although a cap on the DTI ratio would initially affect a small portion of borrowers, the cap would become more binding as house prices rise relative to income, tending to lean against the cycle, unlike LTV limits. Reflecting their benefits, DTI caps were introduced in the United Kingdom and Ireland, which recently experienced strong house price dynamics, and the measure is under consideration in New Zealand and Norway;

  • Giving FI flexibility to calibrate LTV ratios. LTV limits directly reduce the funding available to borrowers, hence they can reduce housing demand, leading to a decrease in credit and house price growth;

  • Eliminating mortgage tax deducibility or revisiting the property tax ceilings to moderate the incentives to accumulate housing debt; and

  • Further easing the housing supply constraints associated with excessively restrictive regulations in the use of land.

B. The Nonfinancial Corporate Sector

13. Data on Swedish corporates suggest vulnerabilities that should be monitored (Table 4). Large corporates appear to have strong interest coverage ratios, but the strength declines with corporate size. About half of corporate debt is denominated in foreign currency, and banks usually do not require clients to have income in foreign currency as a condition for granting foreign currency loans. A 250 basis points (bps) shock to interest rates would bring the interest coverage ratio of medium and small corporations to about 1.5. This would be aggravated if a recession reduced corporations’ earnings. Corporate fragility is particularly important for banks with sizeable exposures to small and medium enterprises and in scenarios in which households would deleverage. Corporates represent about 30 percent of banks’ total exposures.

Table 4.

Corporate Vulnerabilities

article image
Sources: Orbis; and IMF staff calculations.Note: Top Swedish nonfinaical corporations by sales, and OMX 30 nonfinancial corporate members found in Orbis. Scenario 1 represents a 250 basis point interest expense increase. Scenario 2 reflects a Lehman event where interest rates increase by 17 percent and EBIT declines by 23 percent. EBIT = earnings before iterest and taxes.

14. The mission recommends including corporate risks in FI’s risk analysis. This could be done by implementing regular top-down stress tests of corporate resilience (along the lines of what FI currently uses for assessing household resilience) to gauge the impact of macroeconomic instability on corporates’ ability to repay debt. Enhanced supervisory focus on small and medium enterprise portfolios, in particular for banks with concentrations in such exposures, is warranted.

Resilience of the Financial System

A. Banks

15. Banks have remained profitable in the negative rate environment (Figure 3 and Table 5). Their high returns on equity are primarily driven by low credit losses, low funding costs, and improved cost efficiency. Banks’ net interest margin has been stable. The large share of low-cost wholesale funding has partly mitigated the impact from the zero floor on retail deposit rates. With about 70 percent of mortgages having floating rate contracts, banks re-price interest rates every three months.

Figure 3.
Figure 3.

Bank Profitability

Citation: IMF Staff Country Reports 2016, 355; 10.5089/9781475554618.002.A001

Table 5.

Financial Soundness Indicators for Banks

article image
Sources: Finansinspektionen; Riksbank; and IMF staff calculations.

From 2007 to 2013: based on Basel II, with a consideration to the Basel I-floor. From 2014: based on Basel III.

On a consolidated basis.

From 2010 onward, exposures to credit institutions are included.

Non-consolidated data; parent banks only; monetary financial institutions include banks and housing credit institutions.

16. Banks’ risk-weighted capital ratios are high, but risk weights are among the lowest in Europe. Large Swedish banks appear well capitalized when capital ratios expressed in terms of risk-weighted assets, such as the regulatory Tier I capital ratio (Table 5), are compared with major global banks. At the same time, Swedish banks’ ratio of risk-weighted assets to total assets is among the lowest in Europe, apparently driven by banks’ use of internal rating-based (IRB) models (applied to 96 percent of bank assets) and low default rates in recent years.3 Staff estimates of Swedish banks’ ‘leverage ratios’ are below the average of European peers (Figure 4).

Figure 4.
Figure 4.

Swedish Bank Capitalization in a Regional Perspective

Citation: IMF Staff Country Reports 2016, 355; 10.5089/9781475554618.002.A001

17. FI is working on increasing safety margins. In May 2013, the FI introduced a 15 percent floor for risk weights on Swedish residential mortgages to address IRB model risks. In 2014, the floor was raised to 25 percent as a macroprudential measure, to target risks arising from high growth rates in residential mortgage lending. In May 2016, FI announced the adoption of a supervisory approach to internal models for corporate exposures by which one of the elements is that the banks should assume that at least every fifth year is a downturn year when estimating default probabilities. Following this approach, the risk weights for exposures to corporates are expected to be at least around 30 percent for all banks.

18. Comprehensive stress tests were used to quantify the risk assessment matrix (Table 6). Top-down (TD) solvency stress tests were implemented by the FSAP team and the authorities, and results were compared with the outcome of bottom-up (BU) stress tests conducted by banks for the European Union (EU)-wide stress testing exercise by the European Banking Authority (EBA). The FSAP TD stress tests covered the largest four banks using a balance sheet approach, and the authorities ran a simpler framework based on a projection of credit losses.4 The tests considered two five-year scenarios: a baseline, following the IMF’s April 2016 World Economic Outlook; and a stress scenario, reflecting a combination of Sweden-specific shocks and spillovers from a recession in the global economy. The stress scenario entails an ‘L-shaped path’ for economic growth: a deep recession followed by slow recovery due to domestic balance sheet adjustment. The dynamics of most variables in the first three years of the IMF’s stress scenario are identical to the EBA 2016 adverse scenario for Sweden, but the interest rate and exchange rate shocks are more severe (Table 7).

Table 6.

Risk Assessment Matrix 1

article image
article image
Source: IMF staff’s assessment.

The matrix shows events that could materially alter the baseline (the scenario most likely to materialize in the IMF staff’s view). It reflects staff’s views on the source of risks and overall level of concern at the time of discussions with the authorities.

The likelihood of risks (if the baseline does not materialize) is the IMF staff’s assessment of the risks around the baseline (’low’ is a probability below 10 percent, ‘medium’ between 10 and 30 percent, and ‘high’ between 30 and 50 percent).

Table 7.

Stress Testing: IMF Stress Scenario

article image
Sources: IMF staff scenario and calculations; past data from the authorities.

19. The stress tests suggest that bank solvency would be resilient to severe economic distress (Figure 5). In the stress scenario, one bank would fall below the supervisory hurdle rate and no bank would breach the regulatory threshold. The credit losses would be driven mainly by losses on corporate exposures, with about two-thirds coming from non-Sweden exposures.

Figure 5.
Figure 5.

Stress Test Results: Bank Solvency

Citation: IMF Staff Country Reports 2016, 355; 10.5089/9781475554618.002.A001

Source: IMF staff calculations.1/ The regulatory hurdle rate is defined as the Common Equity Tier 1 capital (CET1) ratio that does not include buffers and therefore includes a minimum Basel III capital requirement, Pillar 2 own-fund requirements associated with pension risk, concentration risk and interest rate risk in the banking book and the ‘microprudential’ capital requirements for mortgages in Sweden and Norway (15%). The supervisory hurdle rate took into account all capital buffers: ‘macroprudential’ mortgage floors (10%), capital conservation buffer, countercyclical capital buffer, and common equity systemic risk surcharge (5%). The box and whisker chart shows distribution of the results into quartiles, indicating the mean as well as outliers.

20. There is considerable uncertainty around capital ratios under stress. Losses produced by the FSAP stress tests and by the authorities’ model were larger than those estimated by banks in the EBA exercise, resulting in lower projected capital ratios under stress.5 This was partially due to the inclusion of more severe interest rate and exchange rate shocks; but the authorities’ models also produced much larger credit losses than banks’ models, suggesting significant modeling differences.6 Single factor shock analysis showed that while losses due to individual materialization of severe interest, market, and concentration shocks would be manageable, increases in risk-weighted assets due to shutdown of foreign exchange swap markets would entail higher capital needs.

21. The contagion analysis suggests that the risk of insolvency spillovers among the four largest banks is limited, but the analysis is subject to caveats. The contagion analysis likely underestimated the spillover effects, since it did not capture all domestic interbank exposures and did not take into account the default risk outside the banking sector, including abroad.

22. The stress tests provide a degree of comfort about banks’ solvency, but modeling tail risks in Sweden is challenging. Available models may suffer from overreliance on recent historical experience and have difficulties capturing extreme but plausible scenarios, such as a combination of domestic housing market distress with global financial market turmoil, including the kind of non-linear effects that a housing market correction could imply.

23. The solvency exercise has helped identify data gaps and shortcomings in the stress testing framework. The authorities’ solvency stress tests use many assumptions, especially on pre-provision income and risk-weighted assets. The mission recommends modeling approaches that would replace assumptions and allow projections of the main elements of bank balance sheets and income statements to be consistent with scenarios. As a prerequisite for the scenario-based stress testing framework, the authorities need to obtain longer time series of detailed balance sheet and income statements, and more granular data on trading and net interest.

24. To safeguard against model and measurement errors in calculating capital ratios and given the significant responsibilities of Swedish banks for financial stability in the region, the mission recommends timely adoption of a leverage ratio as a backstop. FI should promote sensitivity of bank models to tail risks and investigate drivers of risk weight variability. The leverage ratio is not meant to supplant risk-based capital ratios, but has an important complementary role in adding resilience. Unlike the risk-weighted capital framework, it does not seek to estimate the relative riskiness of assets and so can mitigate a failure to assign risk weights that reflect the true underlying risk of assets. Applied alongside risk-weighted capital requirements, it can limit incentives to reduce estimates of risk weights instead of raising capital.

25. Housing finance imposes considerable maturity transformation and refinancing risks. Sweden has one of the highest loan-to-deposit ratios in Europe (about 200 percent). Customer deposits represent only about 40 percent of bank funding, as households invest extensively in mutual funds, insurance products, and securities. The long-maturity residential mortgages rely on wholesale funding, such as covered bonds with three-year average maturity.

26. Banks could withstand funding and market liquidity shocks, but there are pockets of vulnerability beyond the 30-day timeframe. The TD liquidity tests assessed banks’ resilience to short (up to 30 days) and medium term (up to one year) liquidity shocks, using the Swedish LCR,7 and Basel III Net Stable Funding Ratio (NSFR) on an aggregate basis and by currencies. Banks appear resilient within these parameters (Figure 6), but the maturity ladder exercise (beyond one month) shows large maturity mismatches for some banks, reflecting heavy reliance on wholesale funding in foreign exchange to finance long-term loans.8 Parameters consistent with a Lehman-type scenario would lead to great short- and medium-term liquidity pressures.

Figure 6.
Figure 6.

Stress Testing Results: Bank Liquidity

Citation: IMF Staff Country Reports 2016, 355; 10.5089/9781475554618.002.A001

Sources: IMF staff calculations based on data from the authorities. Note: The chart shows the distribution of the results into quartiles, indicating the mean as well as outliers.

27. The mission recommends supervisory enhancements to the liquidity monitoring framework. This could include an extended LCR for the three-month horizon in U.S. dollar and euro to further strengthen banks’ liquidity. The authorities should also investigate inconsistent results between compliance with the NSFR in foreign currencies and the outcome of the maturity ladder exercise.

B. The Insurance Sector

28. Low interest rates are challenging for the life insurance sector. The duration gap between assets and liabilities is among the highest in the EU. Companies have reacted by reducing guarantees for new business and increasing investment allocations to equities. Compared with European peers, Swedish insurance companies hold high equity exposures, making the sector vulnerable to a ‘double hit’ of a prolonged period of low interest rates and an asset price fall (Figure 7). In the nonlife sector, risks are less sizeable.

Figure 7.
Figure 7.

Insurance Companies: Duration Mismatches and Asset Allocation

Citation: IMF Staff Country Reports 2016, 355; 10.5089/9781475554618.002.A001

Sources: European Insurance and Occupational Pensions Authority; and IMF staff calculations.

29. Insurance company stress tests indicate pockets of vulnerabilities, mainly on the asset side, but the industry seems well capitalized on aggregate and able to withstand severe price shocks. The stress test included four life and three non-life companies, accounting for 78 and 53 percent, respectively, of gross written premiums in each sector. An increase in the credit spread of Swedish covered bonds of 500 basis points could lead to a 7 percent reduction in the own funds of the median company (Figure 8).

Figure 8.
Figure 8.

Insurance Solvency Stress Tests

Citation: IMF Staff Country Reports 2016, 355; 10.5089/9781475554618.002.A001

Source: IMF staff calculations based on data from the authorities.Note: Bubble size depicts the market value of expiring assets.

30. The mission encourages FI to perform regular macroprudential stress tests. Scenarios should complement the Solvency II Standard Formula and the companies’ Own Risk and Solvency Assessment, and incorporate a multi-period perspective. Should FI consider that the Standard Formula does not adequately capture the risk profile of a company, the need for a Pillar 2 capital add-on should be assessed.

C. Investment Funds

31. The size of investment funds exacerbates the risks of transmission of redemption shocks in the fund industry to the rest of the economy. Total net assets have increased by 70 percent since end-2011, driven mostly by equity and balanced funds. Open-ended investment funds are exposed to redemption risk, i.e. asset liquidation that could represent a sizable shock for other institutions. While the liquidity risks for investment funds are alleviated by the existence of the Premium Pension Authority, which represents 25 percent of total assets, the redemption risk is still relevant since units redeemed and reinvested in other investment funds would probably not be invested in the same assets sold by the funds faced with outflows.

32. Stress tests suggest that corporate bonds markets are the most exposed to redemption shocks (Figure 9). The test measured whether markets would be able to absorb severe redemption pressures. Assets sold by investment funds were compared to turnover data. Parameters were calibrated based on historical distribution of redemptions. The results do not take into account that materialization of liquidity risks in other sectors would put additional pressures, and therefore it may overestimate the markets’ absorption capacity.

Figure 9.
Figure 9.

Investment Funds: Stress Test Results

Citation: IMF Staff Country Reports 2016, 355; 10.5089/9781475554618.002.A001

Source: IMF staff calculations.Note: ‘Sold assets- pro-rata’ represent asset sold by investment funds hit by a tail event redemption shock that have to sell their assets pro-rata, that is, making sure that the structure of assets is intact (approach 1). ‘Sold assets- assumed ordering’ represent asset sold by investment funds hit by a tail event redemption shock that have to sell their assets in descending order of liquidity (approach 2).

33. The authorities should develop monitoring tools to continuously assess market sensitivity to extreme mutual fund redemptions. As funds continue to grow and possibly change their structure of investments, it is important for the authorities to have in place liquidity risk analyses to examine the effects of liquidity shocks to investment funds.

D. Interconnectedness Across Institutions and Borders

34. Sweden’s responsibilities for financial stability in the region would become more pronounced after the planned conversion of Nordea’s Nordic banking subsidiaries into branches (Box 1). The Nordic-Baltic financial systems are highly interconnected through banks’ group structures, creating potential for inward and outward spillovers. Swedish banks in Finland account for 70 percent of assets. Baltic subsidiaries do not make large contributions to the groups’ assets, but Swedish banks account for the majority of financial intermediation in the region (80, 70, and 55 percent of total assets in Estonia, Lithuania, and Latvia, respectively). Nordea’s ‘branchification’ would increase pressures on supervisory resources, potential systemic liquidity needs, and the contingent liability of the Swedish deposit insurance scheme.

35. The mission recommends closer monitoring of financial interconnectedness. Banks have considerable interconnectedness via common exposures at the firm level and in the property sector. Moreover, the covered bond market has become a driver of interconnections. Among Swedish investors, it is primarily insurance companies, other banks, and funds that purchase covered bonds. The major Swedish banking groups invest in covered bonds to have buffers of liquid assets and to act as market makers. As of end-2015, covered bond holdings were equivalent to about 85 percent of banks’ equity, and Swedish banks held 28 percent of the covered bonds outstanding in 2015 (Figure 10). Liquidity in the covered bond market could be severely stressed in a systemic event, compounding banks’ risks from exposures to the housing market. The authorities should closely monitor banks’ cross holdings of covered bonds, and consider measures to further contain the potential contagion due to correlated risks.

Figure 10.
Figure 10.

Holders of Swedish Covered Bonds

Citation: IMF Staff Country Reports 2016, 355; 10.5089/9781475554618.002.A001

Source: Riksbank.1/ Excluding subsidiaries outside Sweden.

36. Financial stability frameworks in the Nordic-Baltic region need strengthening through closer supervisory collaboration. The Nordic-Baltic Macroprudential Forum—established in 2011 to discuss financial stability risks and macroprudential policies—brings together central bank governors and heads of supervisory authorities. Although an informal body without decision-making powers, it has proved effective in allowing regional authorities to share financial stability concerns. It should continue operating with enhancements, such as (i) publishing its risk assessments and updates on macroprudential developments in the region; (ii) collecting exposure data for network stress tests among intermediaries in the region.

Oversight Framework

A. Macroprudential Framework9

37. The macroprudential framework needs an upgrade. Effective macroprudential policy requires a clear mandate, well-defined objectives, adequate powers, and strong accountability. Currently, FI’s macroprudential mandate rests on thin legal foundations, and the macroprudential toolkit is not fully developed. The lack of separation of responsibilities between FI and government has slowed down responses to emerging financial stability risks, as testified by the protracted debate on measures to address vulnerabilities in the housing market. The Riksbank has interpreted broadly its mandate to promote a safe and efficient payments system, but has no statutory role in macroprudential policy despite its expertise in financial stability analysis.

38. Given Sweden’s choice of macroprudential framework, the law should clearly allocate the financial stability powers between the government and FI. The government should provide FI with a broad mandate and set of instruments that FI can use to address systemic risks in a timely and effective manner. This includes the ability to adopt and change instruments and their calibration.

39. The FSC’s financial stability responsibilities should be strengthened by a statutory mandate and recommendation power. The RB should be given an important role in systemic risk analysis so that the FSC can fully benefit from its expertise and information on the financial system (through its payment system oversight and systemic liquidity roles). The preparatory group of the FSC should be upgraded to a Systemic Risk Committee, which would meet at a higher frequency than the FSC, with support from Riksbank and FI staff. The Systemic Risk Committee could propose to the FSC the adoption of recommendations. To enhance traction while ensuring agencies’ independence, the recommendations should preferably have a ‘comply or explain’ attribute. The FSC should issue an annual financial stability assessment, drawing on the stability reports of its members, to provide a clearer guide to financial system stakeholders.

40. The authorities should enhance their financial stability analysis toolkit. They should conduct regular surveys on the distribution of households’ financial assets; and improve the stress testing framework. FI should be allowed to increase the resources dedicated to systemic risk oversight and to cross-institutions supervisory issues.

B. Systemic Liquidity Management

41. Banks have improved their short-term liquidity after the global financial crisis. After the Lehman Brothers collapse in September 2008, Swedish covered bonds came under heavy selling pressure. The market was subsequently calmed by a series of measures by the Riksbank and NDO, and banks started to build up short term liquid buffers. The LCR requirement of 100 percent was rolled out well ahead of the Basel timetable; in addition, the Swedish banks are required to meet LCR in euro and U.S. dollar. To date, all major Swedish banks have met the requirements, with higher LCR ratios in major foreign currencies (Figure 11).

Figure 11.
Figure 11.

Major Banks’ LCR by Currency

Citation: IMF Staff Country Reports 2016, 355; 10.5089/9781475554618.002.A001

Sources: Sveriges Riksbank.

42. A sizable share of banks’ ‘high-quality liquid assets’ consists of covered bonds issued by other Swedish banks, raising concerns about concentration risk. Although the Swedish definition of LCR has a less favorable treatment of covered bond holdings than the Basel LCR, covered bonds make up to 33 percent of banks’ high-quality liquid assets, equivalent to about 28 percent of total outstanding covered bonds in 2015 (this includes subsidiaries outside Sweden). Recognizing the potential concentration risk, in December 2015, the Riksbank has amended terms and conditions governing collateral requirements for credit with the Riksbank.

43. The emerging signs of deterioration in the liquidity of the bond market require close monitoring. The turnover ratio for government bonds fell sharply at the onset of the global financial crisis, owing to a surge in demand for safe assets. After stabilizing at levels close to those that prevailed pre-crisis, the turnover ratio began declining from 2013, even before the Riksbank started its quantitative easing (Figure 12). At the same time, bond price sensitivity to turnover in ten-year government bonds points to a smaller deterioration in market liquidity.

Figure 12.
Figure 12.

Market and Structural Liquidity

Citation: IMF Staff Country Reports 2016, 355; 10.5089/9781475554618.002.A001

Note: The Riksbank’s structural liquidity measure is based on stricter assumptions than the NSFR (Riksbank’s Financial Stability Report 2010:2, pages 82–83).

44. Nordea’s planned branchification’ will raise the potential needs for Riksbank’s ELA. The Riksbank can provide unlimited Swedish krona to the banking system, but liquidity injections in foreign currencies require a foreign exchange buffer. Swedish banks had short-term external debt equivalent to about 50 percent of GDP at end-2015. Conversion of Nordea’s subsidiary banks in the region into branches would add an amount equal to a further 25 percent of GDP. The Riksbank will have the responsibility to provide ELA to all Swedish banks, including their branches outside Sweden. Despite their solid short-term liquidity buffers, the major Swedish banking groups (including regional subsidiaries) have lower ‘structural liquidity’ compared with other European banks according to the Riksbank’s measure (Figure 12).

45. The authorities should ensure appropriate foreign exchange buffers. The mission recommends that the Riksbank seeks to establish swap agreements with central banks in the Nordic countries, the Fed and the ECB.

46. The need for foreign exchange buffers could also be contained by further strengthening banks’ foreign exchange liquidity. Over a one month period, the Swedish banks have ample liquid buffers in euro and U.S. dollar, indicated by high LCRs. Yet, the maturity ladder exercise finds that the banks need to seek foreign currency if the funding stress lasts longer than one month. Monitoring an extended LCR, beyond the ELA one-month horizon, could further strengthen banks’ liquidity and reduce potential ELA needs. The ongoing work on MoUs among the Nordic central banks may help the Riksbank to access Norwegian and Danish kronor liquidity, suggesting that the extended LCR should focus on euro and U.S. dollar.

C. Banking10

47. FI has made progress since the 2011 FSAP with supervisory approaches and techniques. Structured risk assessments for the four large banking groups have been rolled out and FI’s independence has been strengthened, but gaps remain.

  • The roll-out of FI’s Supervision Strategy addresses one of the major findings of the 2011 FSAP, namely the lack of a formalized analytical framework to assess the risk profiles of supervised institutions. A number of gaps in the prudential framework have been addressed through a combination of EU legislation and regulation, and domestic regulations. However, significant gaps remain. Implementation of EU-wide requirements has improved the granularity and frequency of reporting by supervised institutions.

  • FI reports that its independence has been strengthened, and it is better able to allocate resources and set its own priorities within the overall budget ceiling set by the government The authorities have chosen not to proceed with recommendations to enhance the legal protection for supervisors, expand the remedial powers available to FI, and require fitness and probity reviews of members of senior management. They view the first two points as being adequately addressed within the existing framework, and the third as inconsistent with Swedish law, which does not recognize the concept of senior management.

48. A major challenge is that FI continues to be under-resourced relative to the size and complexity of the supervised system. Bank supervision has less than 100 staff to supervise 124 institutions, including one G-SIB. The result is limited analytical capability, too few examinations, and over-reliance on a small number of key people. The FSAP reviews in 2002 and 2011 both raised concerns about the adequacy of supervisory resources, given such a large banking sector. Since the last FSAP, FI’s banking supervision staff has increased. However, it remains under-resourced. Moreover, given new demands, including the prospective conversion of Nordea’s Nordic banking subsidiaries into branches, resources will be even more stretched. The lack of experienced bank supervisors also substantially weakens supervisory capacity.

49. FI may issue regulations only in areas specifically vested by a law or ordinance, resulting in long delays in regulations. Swedish law requires the power to issue regulations to be specifically vested in an authority. ‘Catch-all’ provisions may only be used to issue non- binding policy guidance. In several instances, such as the issuance of credit risk standards, introduction has been delayed by several years awaiting amendment of the law or ordinance to provide the specific power required. The mission recommends giving FI a general power to issue binding regulations on safety and soundness issues.

D. Insurance11

50. Solvency II has brought higher standards of regulation, but is not being applied in full, on a mandatory basis, to occupational pensions insurance. Solvency II has brought improvements in regulatory reporting and group supervision, as well as higher overall solvency coverage. However, the requirements apply in full across only some 40 percent of the life insurance market. Sweden has allowed firms to exclude occupational pensions insurance from the application of the main solvency provisions of the framework. The mission recommends that the authorities resolve as soon as possible the uncertainties over the approach to regulation to be taken following the end of the transitional period in 2019; and that whatever regime they choose for after 2019, they ensure the same level of protection to occupational pensions as to life insurance.

51. The supervisory framework should be strengthened. FI is implementing an approach to supervision based on a well-articulated risk assessment framework. The mission recommends that FI overlay the risk-based allocation of resources with minimum supervisory staffing levels, an extended range of minimum supervisory activities for the highest impact entities and increased resourcing of the teams addressing the highest risk companies.

52. The mission noted that FI’s approach to consumer protection work would also benefit from increased resources as well as a broader mandate. FI’s new and separate consumer protection function has equipped it better to identify and address risks to consumers. Its agenda includes major issues such as the future of commissions-based remuneration. A broader mandate from government would avoid the need for FI to seek mandates in respect of new areas of focus, with the delay this can entail.

E. Securities Markets12

53. FI’s Consumer Protection and Markets Areas have the primary responsibility for regulating and supervising securities markets. Their staff currently totals approximately 140, some of which are however responsible for other activities. Markets Area is subject to extraordinarily high staff turnover, reaching 28 percent between January 1 and November 30, 2015.

54. FI should review the Consumer Protection Area’s activities to assess where the current consumer protection angle should be complemented with explicit consideration of financial stability risks. While FI’s objective of promoting the financial system’s stability and efficiency also guides Consumer Protection Area’s planning, the emphasis continues to be on consumer protection. This may no longer be appropriate given the increased financial stability risks arising from certain securities markets activities, in particular fund management.

55. Greater data availability and quality and better analytical tools would improve FI’s ability to analyze risks and plan its supervisory activities. Solving the current reporting challenges is a precondition for building sufficient tools to conduct regular and ad hoc analyses to target supervisory activities and assess financial stability risks. To help address these challenges, authorities are encouraged to cooperate in data collection and analysis. To effectively supervise cross-market and cross-border trading in the current fragmented trading environment, FI should also acquire an automated market surveillance system.

56. Wider communication of inspection findings and recommendations is important to enhance the effectiveness of FI’s current supervisory program. FI is encouraged to develop additional tools to communicate the general findings and recommendations of its thematic inspections to all supervised entities and, where relevant, to the public.

57. Enhancements to cross-border supervisory cooperation and active participation in the work of the European Securities and Markets Authority are necessary complements to FI’s current active domestic supervisory program. It is important to continue to build closer cooperation with FI’s key foreign counterparts, in particular in the Nordic region and Luxembourg. FI should actively raise issues for discussion at European Securities and Markets Authority to enhance EU level convergence to avoid regulatory arbitrage. This is important given the increased ability of fund managers, funds and investment service providers to choose their domicile within the EU.

F. FMIs13

58. There is room to enhance arrangements for FMI crisis management. FMIs are well developed, and subject to minimum regulatory requirements established in relevant international standards. However, there is overlap in the supervision and oversight by FI and the Riksbank. Arrangements have been established to facilitate cooperation. There is evidence that the authorities speak with one voice where relevant, but FI and the Riksbank should clarify how they would coordinate during a (non-resolution) crisis, possibly by establishing a joint crisis management plan.

59. Risk management at Nasdaq Clearing appears to be sound, but it does not have a comprehensive and well-developed recovery plan to ensure critical services in times of crisis. It should also have the ability to comprehensively address credit losses and liquidity shortfalls. All recovery tools must be backed by formal agreements or be in the rulebook, as appropriate. FI and the Riksbank have recognized the need for further development, and Nasdaq Clearing has begun work in this area. EU legislation, expected to be introduced in late 2016, should provide additional certainty and ensure that authorities have enforcement powers.

60. The mission encourages FI and the Riksbank to continue reviewing the adequacy of resources for FMI oversight and supervision. Effective cooperation helps alleviate pressures but international policy work and increasing demands in other areas can increase pressures. Additional resources will also need to be devoted to work on FMI crisis management when EU legislation on central counterparty resolution is introduced.

61. The mission recommends that FI and the Riksbank clarify how the two authorities would coordinate in a crisis. This should involve establishing a join crisis management plan. Further work will be required when the EU legislation regarding central counterparty resolution is implemented, possibly with a wider range of authorities.

G. Financial Integrity

62. Sweden has made significant progress in addressing deficiencies in its financial integrity regime, but improvements to the anti-money laundering/combating the financing of terrorism (AML/CFT) framework could enhance its effectiveness. Since the 2006 assessment,14 the authorities have strengthened the legal framework, covered national and sectoral money laundering/terrorist financing risk assessments, and developed a national AML/CFT strategy. Nevertheless, important shortcomings remain (e.g., low risk customers are completely exempt from customer due diligence requirements, identification and verification of beneficial owners could be improved, and suspicious transaction reports from specific sectors are of low quantity and quality). The authorities should improve their understanding of money laundering/terrorist financing risks and enhance cooperation on AML/CFT among stakeholders. Addressing the customer due diligence exemption, establishing a register of beneficial ownership through legislation, and providing feedback to reporting entities on improving suspicious transaction reports will contribute to the overall effectiveness of the AML/CFT regime.

Crisis Readiness, Management, and Resolution15

63. The authorities have addressed key recommendations of the 2011 FSAP. They held a crisis simulation exercise in late 2014 while a new crisis simulation exercise is under preparation. In mid-2011, they revised the deposit insurance scheme through implementation of EU directives, including the FSAP recommendations to shorten the payout period and to redefine the payout trigger,16 and in early 2016, they vested the National Debt Office with resolution authority and established a bank resolution regime by implementing the EU’s Bank Recovery and Resolution Directive (BRRD).17

64. The overhauled framework is untested, and further investments are needed to be able to rapidly deploy recovery and resolution tools. Establishing a new framework requires additional resources, including for a new resolution handbook to operationalize the resolution toolkit, an updated early intervention manual to ensure consistency with the BRRD, and the first recovery and resolution plans for all 200 banks and investment companies. Even though there are no official metrics to assess the needs for recovery and resolution staffing, FI and NDO appear understaffed when benchmarked against other EU countries’ agencies, particularly in light of Sweden’s home-country responsibilities.

65. The introduction of the special bank resolution regime requires a smooth transition from early intervention by FI to resolution by the NDO. The new regime also underlines the importance of close cooperation between FI, NDO, and the Riksbank in assessing the solvency and viability of institutions receiving ELA from the Riksbank, particularly during resolution. Existing MoUs between the agencies should be updated. As the home country for one G-SIB and three other major cross-border banks, Sweden will need to rely on comprehensive bi- and multilateral cross-border arrangements to ensure effective recovery and resolution including of systemic Swedish bank branches. The mission recommends that a revamped Nordic-Baltic Stability Group could oversee cross-border crisis preparedness and management.

66. The mission recommends clarifying the Riksbank’s financial stability mandate and related powers to provide liquidity assistance. Riksbank’s financial stability mandate should include a clear role in the oversight of systemic risk and explicit confirmation of statutory authority to extend ELA for financial stability purposes to individual institutions (’idiosyncratic ELA’) and the system as a whole (’systemic ELA’). The mission recommends that a special clause be introduced in the Riksbank Act and that strategies be defined to provide liquidity to banks in resolution. To strengthen the Riksbank’s ability to provide ELA in foreign currencies—should the Riksbank’s foreign exchange reserves be insufficient—the mission recommends that the Riksbank seek swap agreements with central banks of jurisdictions where Swedish banks operate through branches. To further protect the Riksbank’s balance sheet against potential exposures due to ELA, the mission recommends that the authorities consider complementary ex ante, standing indemnification and guarantee arrangements for any ELA losses if incurred by the Riksbank without subjecting each ELA operation to ad hoc approval from third parties. This will be crucial for promoting the Riksbank’s financial and operational autonomy.

67. The responsibility to actively oversee national contingency planning, including national and cross-border financial crisis simulation exercises, has not been assigned; nor is there a national contingency plan. The mission recommends that this be addressed by expanding the FSC’s mandate, which will also require that the supporting preparatory group and FSC secretariat include contingency planning expertise. The FSC would remain a platform for inter-agency coordination and the agencies would continue to use their powers within their mandate.

Proposed Conversion of Nordea’s Nordic Banking Subsidiaries to Branches

In May 2016, Nordea, the largest financial services group in Northern Europe, with assets of about EUR 650 billion, received approval from Sweden’s FI to convert its Danish, Finnish, and Norwegian subsidiary banks to branches of its Swedish bank. These subsidiary banks are currently supervised by the national authorities in Denmark, Finland, Norway, and the ECB.

The majority of Nordea’s funding is wholesale, and most of its balance sheet is in currencies other than SEK. Nordea’s assets are equivalent to 140 percent of Sweden’s GDP. The Financial Stability Board has designated Nordea as one of 30 G-SIBs based on Nordea’s cross-border presence in the Nordic region and its dependency on wholesale funding. As a G-SIB, Nordea should be subject to more intensive and effective supervision, commensurate with the potential destabilization risks that G-SIBs pose to their domestic financial systems, as well as the broader international financial system.

Nordea is awaiting approval from authorities in the other three countries before it proceeds. If it does go forward, it will significantly increase Sweden’s financial stability responsibilities within the region. Under the EU Fourth Capital Requirements Directive (CRD-IV), virtually all authority for the supervision and resolution of international branches is assigned to the home competent authority. As a result, absent an agreement with the other Nordic countries and ECB to share resources, Sweden’s FI will need to assign its supervisors to take the place of the Danish, Finnish, Norwegian, and ECB supervisors, who will no longer be responsible for the supervision of the Nordea operations in their respective jurisdictions. The Riksbank will become Nordea’s primary central bank, and the Swedish government will become the primary deposit insurer and be responsible for any potential resolution. Supervisors in Denmark, Finland, Norway and at the ECB expressed concerns that while CRD-IV anticipated that some international branches may be significant in host countries, it had not anticipated branches that are among the top three or four banks and impose systemic risks on host countries.

The Riksbank, in comments on the proposed conversion, pointed out Nordea’s cross-border operations and its size relative to the Swedish economy. It recommended that the Swedish FI impose liquidity coverage ratios on Nordea in Danish and Norwegian kronor, be given additional resources to carry out its expanded responsibilities, and be given clear authority to implement macroprudential measures consistent with those in the other countries. It also suggested tightening capital requirements of the four major Swedish banks.

The NDO’s view was that the risk to the state is not greatly affected by Nordea’s plans. The NDO thought that although the reorganization would extend Sweden’s formal liability for deposit guarantees and resolution, the increase in the sovereign’s financial commitments would be limited. The NDO pointed out that the new arrangements for handling a banking crisis applied since February 2016 mean that costs of resolution are mainly borne by the bank’s owners and lenders. In NDO’s assessment, the ‘branchification’ could mean an improvement in the conditions for effective and orderly handling of a crisis.

The Swedish FI, in its decision granting Nordea Bank AB the authorization to execute the merger plans,1 concluded that Nordea’s transformation would not significantly increase risks to public interests, nor would it impair the conditions for handling a crisis, or that the Swedish State’s potential commitments would be significantly increased. FI did not see an increased risk of a serious disruption in the payment system or the manner in which the capital market functions. At the same time, FI thought that it may be relevant to apply a temporary own funds requirement to address the risks during the transition period. FI has committed to carefully monitoring Nordea Bank AB’s contingency plans for liquidity and funding, conduct more frequent and comprehensive supervision, and, if necessary, widen the scope of the bank’s reporting.

It is crucial that the relevant supervisors, central banks, and ministries of finance work on agreements that should ease some of the concerns of the host countries and ECB, while providing support to Sweden if it is to take on these additional responsibilities.

1http://www.fi.se/upload/90_English/20_Publications/20_Miscellanous/2016/beslut-bilaga-nordea-16-maj-2016-eng2.pdf

Appendix I. Follow-up on the 2011 FSAP’s Key Recommendations

article image
article image
article image
article image
article image
article image
article image

Appendix II. Stress Test Matrix

article image
article image
article image
article image
article image
article image
article image
article image
article image
article image
article image
article image
article image
article image
article image
article image
article image
3

For an analysis of the drivers of risk weights in Sweden relative to other European countries, see Turk, Rima: “Risk Weights in Europe: Heterogeneity, Harmonization, and Possible Determinants,” IMF Working Paper, forthcoming.

4

The stress tests undertaken by the team and the authorities are summarized in the Stress Tests Matrix (Appendix II).

5

A supplemental solvency stress test analysis was carried out using contingent claims analysis. The results are consistent with the balance sheet tests reported here (see the “Stress Testing” Technical Note).

6

The main difference between the IMF’s and the authorities’ test was the specification of the dependent variable, as discussed in the Technical Note.

7

The Swedish LCR regulations are based on the LCR originally proposed by the Basel Committee in 2010, which is more conservative than the revised proposal from 2013.

8

The cash flow test assessed banks’ liquidity profile in a scenario characterized by outflows of at least 20 percent of funds within three months and additional 5–10 percent within next three months.

9

See also the “Systemic Risk Oversight Framework and Management” Technical Note.

10

See also the “Banking Regulation and Supervision” Technical Note.

11

See also the “Insurance Sector Regulation and Supervision” Technical Note.

12

See also the “Regulation and Supervision of Cross-Border Securities Activities” Technical Note.

13

See also the “Supervision and Oversight of Financial Market Infrastructures” Technical Note.

14

The Financial Action Task Force is assessing the AML/CFT regime under the revised standard, and is expected to finalize its report in 2017.

15

See also the “Financial Safety Net and Crisis Management” Technical Note.

16

The deposit insurance scheme appears well-funded, at more than twice the EU target level, although the contingent liability may increase with Nordea’s ‘branchification’ (Box 1).

17

After concluding the public consultation, the NDO aims to adopt its policy on minimum requirements for own funds and eligible liabilities (MREL) in Q4 2016 when it also expects to take the first MREL decisions when adopting resolution plans for the four cross-border banks. Meanwhile, it has set MREL equal to required capital.

Same Series

Other IMF Content

Other Publishers