Results of in-depth reviews under Regulation (EU) No 1176/2011 on the prevention and correction of macroeconomic imbalances

Sweden continues to experience macroeconomic imbalances, which require monitoring and policy action. In particular, developments regarding household indebtedness, coupled with inefficiencies in the housing market, continue to warrant attention. Although the large current account surplus does not raise risks similar to large deficits, and is partly linked to the need for deleveraging, the Commission will follow the developments of the current account in Sweden in the context of the European Semester.

More specifically, household debt has increased again after a period of stabilisation, as the main contributing factors – low interest rates on mortgages, debt-bias in taxation, slow mortgage amortisation and limited housing supply – remain in place. Various indicators of credit supply and demand conditions do not indicate imminent deleveraging pressures. After stabilising in recent years, house prices increased again in 2013, driven by favourable demand conditions and supply inefficiencies. Moreover, rental market inefficiencies, cumbersome planning procedures and little competition in construction, have also contributed to the house price dynamics. More stable house prices are needed to limit private indebtedness and reduce macroeconomic risks, once debt service costs increase. The recent macroprudential measures are instrumental, but likely not enough, to reduce macroeconomic risks. In particular, strong tax incentives for debt financing are perceived as key drivers of house prices. Higher residential investment would also improve the savings‑investment balance. As regards the non-financial corporate sector indebtedness the analysis finds that it does not give rise to macroeconomic risks. Moreover, recent reforms in company taxation are likely to further reduce the level of corporate debt by limiting tax minimisation by multinationals.

Excerpt of country-specific findings on Sweden, COM(2014) 150 final, 5.3.2014

Executive Summary and Conclusions 7

1.       Introduction  11

2.       Macroeconomic Developments 13

3.       Imbalances and Risks 19

3.1.     Competitiveness and external position  19

3.1.1.  Competitiveness and the savings‑investment balance  19

3.1.2.  Net International Investment Position  30

3.2.     Private indebtedness: non-financial corporate sector and household indebtedness 33

3.2.1.  Evolution of indebtedness of the non-financial corporate sector 33

3.2.2.  Factors behind Swedish corporate indebtedness and reforms undertaken  35

3.2.3.  Evolution  of household indebtedness 36

3.2.4.  Measures taken so far and current debate on household indebtedness 40

3.2.5.  Private indebtedness: assessment of imbalance  42

3.3.     The housing market 44

3.3.1.  House price developments 44

3.3.2.  Underlying market fundamentals 46

3.3.3.  Supply side challenges: New constructions 49

3.3.4.  Supply side challenges: The rental market and utilisation of the existing housing stock  52

3.3.5.  Outlook and conclusions 54

4.       Specific Topics 57

4.1.     Banking sector overview   57

4.2.     Credit risk  58

4.3.     Funding risk  60

4.4.     Conclusions 62

5.       Policy Challenges 63

References 67

LIST OF Tables

2.1. Key economic, financial and social indicators 17

4.1.     Financial soundness indicators 58

LIST OF Graphs

2.1.     Contributions to growth  13

2.2.     Export market share decomposition  13

2.3.     Decompostion of REER developments 14

2.4.     Labour market 15

2.5.     Poverty  15

2.6.     Consolidated debt decomposition  16

3.1.     Decomposition of the Swedish current account 19

3.2.     Net borrowing/lending by sector 20

3.3.     Investment rates 20

3.4.     Investment in dwellings 20

3.5.     Non-cyclical current account 21

3.6.     Decomposition of the rate of change of ULCs 21

3.7.     Swedish export market shares: goods and services 22

3.8.     Growth rate of Swedish export market shares 22

3.9.     Swedish market shares and the nominal exchange rate  22

3.10.   Revealed comparative advantage in services 27

3.11.   Revealed comparative advantage in goods 27

3.12.   Total Swedish exports by destination as a percentage of Swedish GDP (2011) 27

3.13.   Sweden imports by geographical origin, as a percentage of trading partner GDP (2011) 28

3.14.   Sweden's current account balance (excluding investment income) by partner region  28

3.15.   Decomposition of the Swedish NIIP  30

3.16.   Decomposition of Rate of Change of the Swedish NIIP  30

3.17.   NIIP by Sector 30

3.18.   Valuation effects 31

3.19.   Valuation effects per financial instrument 32

3.20.   NIIP and FDI at market and book values 32

3.21.   Decomposition of y-o-y changes in corporate debt-to-GDP ratios 33

3.22.   Non-financial corporations debt as % of GDP (consolidated and super-consolidated view) 33

3.23.   Leverage of NFCs 34

3.24.   Leverage of non-financial corporations 34

3.25.   Evolution of credit flows 34

3.26.   Corporate default rate  35

3.27.   Evolution of housing prices 36

3.28.   Households' leverage  36

3.29.   Decomposition of y-o-y changes in debt-to-GDP ratios, households 36

3.30.   Household leverage discounting the price evolution  37

3.31.   Loan-to-value ratio of new mortgage loans 37

3.32.   Discretionary income for Swedish households 38

3.33.   Stress test on interest rate  40

3.34.   Stress test on house price  40

3.35.   Loan-to-value ratio in the mortgage stock, per cent 41

3.36.   Deleveraging pressures 42

3.37.   Real house prices in the Nordic region, Index 2010=100  44

3.38.   Nominal House Price Index in Sweden and the EU   44

3.39.   Price to rent and price to income  45

3.40.   Price to income ratio; index 2000=100  45

3.41.   Price to income ratio; index 2000=100  46

3.42.   Overvaluation gap with respect to main supply and demand fundamentals in Sweden  46

3.43.   Distribution of population by tenure status, 2014, % of total population  47

3.44.   Residential Investment and Building Permits in Sweden  49

3.45.   Residential investment in % of GDP  49

3.46.   Rent Control in the EU27 (quantitative indicators for rent setting) 53

3.47.   Rent to income ratios, selected countries - Rebased 2000=100  53

3.48.   Rent of newly built apartments vs existing stock  53

4.1.     Lending to private sector 57

4.2.     Return on equity (%) 58

4.3.     Non-performing loans 58

4.4.     Tier I capital regulatory ratio  59

4.5.     Tier I capital leverage ratio  59

4.6.     Loans to deposits 61

4.7.     Funding structure of major banks 61

4.8.     Average funding maturities 61

LIST OF Boxes

3.1.     Balance-of-payment revisions reduce the Swedish current account surplus. 24

3.2.     Over-indebtedness 39

LIST OF Maps

No table of contents entries found.

In May 2013, the Commission concluded that Sweden was experiencing macroeconomic imbalances, in particular as regards developments related to private sector debt and the housing market. In the Alert Mechanism Report (AMR) published on 13 November 2013, the Commission found it useful, also taking into account the identification of an imbalance in May, to examine further the persistence of imbalances or their unwinding. To this end this In-Depth Review (IDR) provides an economic analysis of the Swedish economy in line with the scope of the surveillance under the Macroeconomic Imbalance Procedure (MIP). The main observations and findings from this analysis are:

· The indebtedness of the private sector, and in particular household indebtedness, continues to constitute a macroeconomic risk. Against the backdrop of rising house prices, household debt, in particular in the form of mortgages, remains high. At present, credit growth for mortgage loans is less expansive but still outpaces GDP growth. Corporate debt is still high but decreasing in a low growth and investment context. Recent taxation reforms aim at minimising the tax-planning component of corporate debt and further measures are to be expected. Concerns about the private sector indebtedness are to some extent mitigated by very moderate default figures for corporations as well as households but the evolution of credit flows needs to be monitored, especially once economic growth rebounds.

· Although house price are not growing extensively since 2010, inefficiencies in the Swedish housing market continue to imply potential risks for macroeconomic stability, in particular in case of potential house price corrections. These inefficiencies, linked to the structural under-supply of dwellings and the inefficient use of the existing housing stock, have built up over the course of recent decades and have a macroeconomic impact mainly via increased household indebtedness. Despite some recent measures which point in the right direction, the rental market remains subject to far-reaching regulation through the utility-value based system, while high transaction taxes reduce homeowners' incentives for selling houses. The planning and zoning processes for new construction remain lengthy and opaque; the limited competition among construction and property development companies further creates upward-bias in house prices. Generous deductibility of mortgage interest payments from the income tax debt together with low recurrent property taxes keep fuelling house price developments. Housing issues are most pertinent in the Stockholm and Gothenburg regions. .

· Current account surpluses, though large, continue on a declining trend as domestic savings moderate. Increasing residential investment from its comparatively low levels could help improve the savings investment balance. The decline in the current account balance is partly explained by a structural decrease in the goods trade surplus, while service exports, the income balance and a comparatively small energy deficit have had a counterbalancing effect. From a savings-investment viewpoint, the large Swedish external surplus is the reflection of the high net savings of the private sector, the fiscally prudent position of general government and the low level of residential investment. The Swedish tax system provides incentives for building up debt (for instance by allowing tax deductibility) as well as savings (for instance via the pension reform); this is one reason why Swedish households built up high indebtedness in parallel to high savings. As households' precautionary savings decrease, baby-boomers retire, corporate investment rebounds and the government delivers on its planned fiscal stimulus, domestic net savings are expected to further moderate. Despite continuously losing export market shares, Sweden remains one of the most competitive economies in the EU, supported by a generally favourable business environment and a strong performance in R&D and innovation. Results in basic education have, however, deteriorated. The net international investment position is positive when foreign direct investment is estimated at market value and raises no sustainability concerns.

· Credit and funding risks in the Swedish banking sector remain stable, also due to measures undertaken by Swedish authorities. However, developments in the financial sector and the effectiveness of the applied measures call for continued monitoring. On the asset side the key risk of Swedish banks is their increasing loan portfolio to households, mainly through mortgage loans mirroring the problem of high household indebtedness. On the liability side, banks are exposed to funding risks as they finance their operations predominantly on the international wholesale markets, creating maturity and currency mismatches in their balance sheets.

The IDR also discusses the policy challenges stemming from these developments and what could be possible avenues for the way forward. A number of elements can be considered:

· In order to further promote a sound lending culture and to contain household debt, a comprehensive reform could be considered that gradually reduces and/or limits the scope of the tax deductibility of interest payments while at the same time strengthens recurrent property taxation. Such a reform, including the need to address taxation and indebtedness, could already now be outlined broadly, so as to allow quick first steps and a carefully sequenced implementation once the political situation allows this. In addition, it could also be useful to push for a further strengthened amortisation culture among borrowers.

· As regards corporate debt, further ways to reduce the debt-bias in company taxation could be considered. It also seems pertinent to make sure that a sound financing framework for companies is put in place, with a view to investment needs once the low growth and investment context has been left behind.

· Focusing on the housing market, the Swedish government has taken several steps pointing to the right direction. Nevertheless, further additional steps could be taken. As concerns the rental market, these could include deregulating rent setting in the sense that individual tenants and landlords would be able to agree on rent levels and hence increase the freedom of contract. A first step could be to allow rents of newly produced rental units to fully reflect tenants' willingness to pay or that rental prices can be set more freely in case of privately owned rental units. Eventually, rents of the remaining stock could be adapted. As regards construction, it would seem that a more strategic, overarching approach to solve the inefficiencies weighing on the housing market would be useful. Further streamlining of the planning and zoning processes could be considered. In particular, these processes could be rendered more efficient by more standardised building requirements across the country and by decreasing the extensive zoning and planning requirements. More transparent land allotment procedures at municipal level could be another priority as well as facilitating the access to tenders for smaller or foreign developers to increase competition. Identifying clear mechanisms to incentivise developers to construct on land which has already been subject to detailed planning could also be considered. In a similar vein, the incentives for municipalities to support new constructions could also be re-assessed.

The suggested measures are interlinked and reinforce each other: for instance, changes on the housing market would likely affect household debt development. The timeframe of their impact could also differ substantially: while for instance any limitation on the tax deduction would be likely to have short term effects, a simplified and faster building process would likely impact the housing supply in the medium term.

It is crucial to avoid abrupt policy changes in these areas due to their pivotal macroeconomic impact: gradual implementation, well-considered timing, wide political and public support and continuous evaluation of the impact of the measures would be necessary. As discussed in this In-Depth Review, these imbalances have been built up over a longer time horizon, thus their unwinding cannot take place overnight. Nevertheless, a forthcoming stronger economic growth period and the ongoing wide debate in Sweden on these issues could pave the way for further sound policy actions in these areas. 

On 13 November 2013, the European Commission presented its third Alert Mechanism Report (AMR), prepared in accordance with Article 3 of Regulation (EU) No. 1176/2011 on the prevention and correction of macroeconomic imbalances. The AMR serves as an initial screening device helping to identify Member States that warrant further in depth analysis to determine whether macroeconomic imbalances exist or risk emerging. According to Article 5 of Regulation No. 1176/2011, these country-specific “in-depth reviews” (IDR) should examine the nature, origin and severity of macroeconomic developments in the Member State concerned, which constitute, or could lead to, imbalances. On the basis of this analysis, the Commission will establish whether it considers that an imbalance exists in the sense of the legislation and what type of follow-up it will recommend to the Council.

This is the third IDR for Sweden. The previous IDR was published on 10 April 2013 on the basis of which the Commission concluded that Sweden was experiencing macroeconomic imbalances, in particular as regards developments related to private sector debt and the housing market. Overall, in the AMR the Commission found it useful, also taking into account the identification of an imbalance in May, to examine further the persistence of imbalances or their unwinding. To this end, this IDR undertakes an economic analysis of the Swedish economy in line with the scope of the surveillance under the Macroeconomic Imbalance Procedure (MIP).

Against this background, Section 2 presents an overview of the general macroeconomic developments, Section 3 takes a more detailed look at main imbalances and related risks. Section 4 gives a closer analysis of the stability of the financial sector, while Section 5 discusses policy considerations.

A domestic-demand-driven recovery

Sweden had come out of the 2008/09 sharp recession with a strong rebound in 2010 and solid growth in 2011. However, GDP growth has been meagre in 2012 and 2013 in the wake of the euro-area sovereign debt crisis and slow economic growth among Sweden's trading partners. GDP growth is expected to pick up gradually from 0.9% in 2013 to 2.5% in 2014 and 3.3% in 2015.

Unlike previous recoveries, which were typically export-driven, the main push for future growth is expected to come from domestic demand. Household consumption is projected to be the main source of growth, on the back of expected improvements in the labour market, a low interest rate environment and income tax cuts announced in the 2014 Budget Bill. The household saving rate, which reached record high levels in 2012 and 2013 against a background of high unemployment rate and uncertainties about the euro-area outlook, is expected to decrease once uncertainty subsides and precautionary savings are reduced in the mid-term.

Fixed capital formation is set to resume gradually. Linked to the expected recovery in exports, industrial capacity utilisation is expected to rise, paving the way to investment outlays in 2014 and 2015.

Deteriorating exports market share and rebalancing current account surplus

Despite being one of the most competitive economies of the EU, as further discussed in section 3.1.1, Sweden has been losing export market shares (EMS) since 2008. The losses in EMS come from the goods sector whereas the services sector shows some gains since 2010. Structural factors linked to the traditional product mix of Swedish exports drive the losses in goods EMS. Paper products are increasingly losing ground as carriers of information and saw mill products are facing constraints due to the weak outlook of the construction sector. Year-on-year swings in EMS growth are also related to fluctuations in the krona exchange rate, which affect the value of Swedish exports when measured in euros. For instance, the large drop in EMS witnessed in 2009 correlates with the NEER depreciation depicted in graph 2.3.

Swedish cost developments against those of its trading partners do not indicate major challenges in terms of competitiveness. The real effective exchange rate has appreciated since 2010 due to the strong krona. While unit labour costs have been growing in line with those of Sweden's trading partners, domestic prices have been growing more slowly.

Sweden's current account surplus has been above the indicative scoreboard threshold since 2003. It adjusted from 9.3% to 6% of GDP between 2007 and 2012 and is expected to adjust further once households' precautionary savings decrease and investment rebounds. Notwithstanding two decades of positive current accounts, the Swedish NIIP, while sustainable, remains negative.

The external position and trade performance of Sweden will be further analysed in 3.1.

Low inflationary pressures

After having fluctuated around 0.9% in 2012, HICP inflation dropped in 2013 to 0.4% on account of lower energy costs and contracting prices in the non-energy industrial goods sector. Given the assumed subdued outlook for energy and the continued impact of the Swedish krona appreciation, inflationary pressures are expected to remain low in 2014. Wage developments are expected to contribute to the low-inflation environment as the 2013 pay settlements, which run for three years, indicate a moderate rate of wage growth by historical standards. The current slack in the labour market also points to moderate growth in employee compensation. Following low inflationary pressure in the economy and sluggish economic growth, the Riksbank cut the repo rate by 0.25 percentage points to 0.75% in December 2013.

Recovering labour market and stable social indicators

Despite a recently weak economy, employment has continued to grow in 2012 and 2013 and is expected to continue growing in 2014 and 2015 on account of a growing labour force and reforms in the unemployment and illness insurance schemes.

The unemployment rate which has never fully recovered after the 2008/2009 crisis peaked at 8.3% in March 2013 and seems to have since stabilised around 8%. Unemployment is expected to decrease slowly in 2014 and more significantly in 2015 in connection with stronger growth dynamics.

Although now below the EU average, the youth unemployment rate is high, 24.1% in November 2013, and climbed from 15% in 2001. Several factors tend to alleviate concerns about the youth unemployment rate: a large share of students among the unemployed, high turnover and short unemployment spells, a temporary expansion in the labour force and a generational effect. However, even if students are disregarded, the remaining youth unemployment rate is rather high (11.9% in 2012) compared to Sweden's peers in terms of labour market performance (from 2.5% in the Netherlands to 8.1% in Finland). Most of these non-student, unemployed persons are low skilled people without completed upper-secondary education and young people with non-EU immigrant background. These (over-lapping) groups appear to be the most vulnerable. ([1]) 

The at-risk-of-poverty or social exclusion rate (AROPE) has slightly increased since 2010 but remains far below the euro area average. Furthermore, absolute poverty remains stable at a very low level, 1.3% of total population (7.5% in the euro area).

Domestic debt is mainly a private debt concern

Consolidated private sector debt has been rising from 121% of GDP in 1995 to 212% in 2012, a level far above the alert threshold of 133% of GDP. Following three years of stabilisation at a high level, Swedish private debt as a percentage of GDP is expected to increase in 2013 since its absolute growth has not decreased as quickly as GDP decelerated (preliminary data based on first-three quarters suggest 235% of GDP for 2012 –To be updated).  The correction in 2009-2011 was on account of decreasing indebtedness of non-financial corporations which generate about two thirds of total private debt. On the other hand, household debt, which accounts for the remaining third, has not reduced and has remained stable at around 81% of GDP in 2009-2011 before increasing to 83.6% in 2012.  To complete the picture, financial corporations also contribute significantly to the country's debt with 98.8% of GDP in 2012.

On the other hand, public debt stands at a relatively low level of around 40% of GDP. The decline in the government debt-to-GDP ratio over the previous years was reversed in 2013, on account of a one-off additional borrowing of SEK 100 billion to strengthen the currency reserves of the Riksbank. Taking into account the 2014 budget and its focus on improving household disposable income through tax cuts, the deficit is expected to reach 1.5% of GDP in 2014 and debt to tick up before falling back in 2015. Private indebtedness will be further analysed in section 3.2.

Housing market still under pressure

Although the Swedish housing market has been stable in the recent past, it remains a potential source of instability, despite some recent measures.  In the medium term, raising interest rates and tightening of credit conditions would most likely negatively impact on the demand for mortgages. However, the strong tax incentives for housing coupled with supply imbalances on the market will likely further push house prices upwards. House prices need to stay on stabilisation path to avoid driving private indebtedness further, to avoid overheating and abrupt volatility that could renew concerns on imbalances and of macroeconomic risks.

Recent government measures are agreed to be pointing to the right direction, albeit their impact is perceived to be small in the short term. Further measures would be needed to address the growing inefficiencies on this market.

3.1.      Competitiveness and external position

3.1.1.   Competitiveness and the savings‑investment balance

Large external surpluses slowly decreasing as the savings-investment balance moderates

Swedish current account surpluses, though large, continue on a declining trend. Sweden started posting large current account surpluses in the mid-1990s following the deep economic and financial crisis which had broken out in the beginning of the 1990s and which caused the collapse of the fixed exchange rate system, a sharp depreciation of the krona and the beginning of an export-led recovery process. In the three years to 2012, the Swedish current account surplus marginally breached the 6% threshold of the MIP scoreboard. Surpluses have been on a declining trend since 2007 when they reached 9.3% of GDP (Graph 3.1), and are expected to continue declining to less than 6% of GDP by 2015, according to the Commission 2014 winter forecast([2]).

The decline in the current account balance is partly explained by a structural decrease in the goods trade surplus, while service exports, the income balance and a comparatively small energy deficit have had a counterbalancing effect. Net Swedish exports of goods have declined substantially since the mid-2000s in connection with (i) the increased importance of the services sector in the Swedish economy, (ii) heightened competition from emerging, goods export-oriented economies and (iii) the on-going trend towards servicification([3]). This trend was accompanied by increasing net service exports which, however, have not been sufficient to prevent the overall trade balance from decreasing as a share of GDP. It should also be noted that the Swedish trade deficit in energy products is remarkably small for a non-oil producing country, which is the consequence of its extensive use of renewable hydropower sources and nuclear energy. At -1.7% of GDP in 2012, the Swedish energy trade balance is well above the EU average of ‑4.7%, contributing to explain the continued high external surpluses.

While the overall trade balance has been on a slow decline, the foreign income balance has been improving over the past few decades, turning positive in 2003. This evolution is linked to the continued strengthening of Sweden's net international investment position, as discussed in the next subsection. Current transfers have delivered a steadily negative impact, reflecting Sweden's foreign aid and net positive contributions to international organisations such as the EU, as well as workers' remittances([4]). As is usually the case in developed countries, the capital account plays a residual, negative role.

Domestic savings remain high, but are edging downwards. From a savings-investment viewpoint, the large Swedish external surplus is the reflection of the high net savings of the private sector, the fiscally prudent position of general government and the low level of residential investment. The relatively high saving rate of Swedish corporations observed from 2002 to 2010 was partly due to the comparatively high net savings of the financial sector which have contributed to the high capitalisation levels of the Swedish banking sector (see Section 4). Corporate savings have since begun moderating (Graph 3.2). As global demand picks up, it is likely that previously delayed investment decisions begin to materialise, thereby contributing further to this trend. Swedish households increased their precautionary savings in connection with the 2008 international financial crisis. However, as the crisis subsides, households are likewise expected to moderate their saving behaviour. It should be noted that high corporate and household savings are also a consequence of pension and other reforms introduced in the 1990s, which included a move towards a defined contributions pensions scheme. As the population ages and the baby‑boomers retire, an increasing segment will begin dis-saving, thereby contributing to further decrease the external surplus. Sound budgetary frameworks and fiscally prudent positions have meant that public finances have been close to balance or in surplus during most of the past 15 years. However, the 2013 Budget Bill entailed a small fiscal stimulus, which was reinforced in the 2014 Budget Bill. This partly accounts for the moderate government deficits depicted in Graph 3.2 during the 2012-2015 period.

While overall investment levels compare positively with EU peers, residential investment is strikingly low. After starting from comparatively low levels in the 1990s, Swedish investment rates([5]) increased rapidly, reaching 20% of GDP in 2008. The economic and financial crisis meant a marked decline in investment in many EU countries which, however, was comparatively moderate in the case of Sweden. Over the five years from 2008 to 2012, Sweden invested more than a reference group of EU countries experiencing large current account surpluses (Germany, the Netherlands and Denmark). According to the latest available annual data, the 2012 Swedish investment rate was higher than the one for both the euro area and the group of surplus countries (Graph 3.3).

Swedish investment in dwellings, however, has been persistently low, both when compared with the investment rates of the euro area and of surplus countries (Graph 3.4). The sub‑optimal residential investment performance of Sweden is closely linked to restrictions and inefficiencies in the functioning of the Swedish housing market, as detailed in Section 3.3. Alleviating these constraints would help to balance the Swedish savings‑investment position.

Price developments are contributing to moderate the trade balance surplus. As seen in Graph 2.3, the REER has appreciated significantly since 2010 due to the strengthening of the Swedish krona. This development is partly the reflection of a safe haven effect, linked to the aggravation of euro area tensions in 2010-2011, although fundamentals continue to suggest the possibility of further appreciation([6]). The sustained strength of the krona is expected to continue to weigh down on the trade balance over the Commission's forecast horizon. In fact, the lagged pass-through of the increases in the REER means that the actual current account outturns have likely overestimated the strength of the underlying "non-cyclical" current account in recent years (Graph 3.5). It should also be noted that the effects of a negative output gap in the Swedish economy largely offset the effects of a negative cyclical position in the majority of its trading partner. This is depicted in Graph 3.5, where the income‑adjusted current account is seen to have remained close to the actual current account since 2010. Finally, as regards ULCs, their growth has been more dynamic in Sweden in the past two years than in the euro area, marking a reversion from the pronounced productivity‑driven decrease in ULCs witnessed in 2010 (Graph 3.6).

Statistical revisions have reduced the current account surplus for 2010-13. In September 2013, the Swedish statistical office announced it was revising some items of its balance of payments data, which effectively corrected downwards the Swedish current account surplus. This statistical revision is explained in detail in Box 3.1 and is the main factor behind the decrease in the current account indicator of the MIP scoreboard from 6.6% of GDP in the 2012 edition of the AMR, to 6.2% of GDP in the 2013 edition.

Declining export market shares notwithstanding the large external surplus

Swedish export market shares (EMS) have sustained significant losses since 2007, with goods exports continuing to lose ground to international competitors. In the five years to 2012, Swedish EMS shrunk by 19%, breaching the 6% decrease threshold of the MIP scoreboard indicator. As seen in Graph 2.2 and further evidenced in Graph 3.7, goods exports have long been the driver of the secular decrease in EMS, while service exports have been significantly more dynamic. The strength of services, however, has not been sufficient to offset the EMS losses in goods, not only because EMS gains in services have been comparatively moderate, but also because goods exports still represented approximately 70% of all Swedish exports in 2012.

The evolution of the EMS is affected by a structural decline, and short-term fluctuations linked to exchange rate movements and the cyclical position of the EU. Increasing competition from emerging, export-oriented economies has meant a structural decline in the EMS of many advanced economies, including Sweden's. Notwithstanding the fact that Swedish exports have grown on average 6% per year in the past 16 years, they have been outpaced by world export growth of 8% (Graph 3.8). The weak economic performance of important EU markets in the wake of the international financial crisis has also contributed to curb Swedish export growth. Finally, exchange rate movements also affect market share measurement (Graph 3.9). For instance, a depreciation of the krona has the short-run effect of reducing the value of Swedish exports when measured in a common currency and an opposite long-run effect of supporting market shares via increased price competitiveness.

The co-existence of a large current account surplus with declining market shares can be understood as a consequence of several factors, including a moderate level of imports, the importance of the income balance and the relative growth rate of goods exports. As discussed in European Commission (2012), "as for other large-surplus countries, subdued imports [in Sweden] appear to have been more important than exports in explaining the stronger current account surplus". In fact, the growth rate of imports in Sweden and other EU countries experiencing large current account surpluses was comparatively low in the boom period before 2008. This was due to a domestic demand that was less expansive than in many other EU countries, something which is partly the result of restrictions in the housing market. As previously mentioned, a comparatively high degree of energy sufficiency has also long helped to curb imports. These factors have contributed to support the current account, but have no effect on EMS. Likewise, the income balance has boosted Swedish surpluses, but is of no concern from the EMS viewpoint. When taking together the income balance and the (negative) current transfers balance, they are seen to have added 1.2% of GDP to the current account on average from 2007 to 2012. Finally, Swedish goods exports increased by 9.1% from 2007 to 2012 in euro terms, while world goods exports increased by 40% over the same period, implying large EMS losses which are of a higher magnitude than the effects that the sluggish Swedish growth rate of goods exports has had on the current account.

(Continued on the next page)

Box (continued)

A highly competitive business environment, but educational performance is sliding

International assessments rank Sweden consistently among the world's most competitive economies. In the 2013-14 edition of the Global Competitiveness Report([7]), Sweden ranked 6th out of 148 economies, and third in the EU, equalling or beating the average of its reference group of "innovation-driven economies" in all of the 12 competitiveness "pillars" considered in the report. Particularly strong performances are noted in the fields of macroeconomic environment, innovation, institutions, financial market development and technological readiness. The most problematic factors pointed out by survey respondents were restrictive labour regulations and tax rates. Likewise, in the 2014 edition of the World Bank's Ease of Doing Business report([8]), Sweden ranked 14th out of 189 countries and 4th in the EU. The only comparatively weak points noted were related to starting a business and taxation. Swedish institutions enjoy, furthermore, one of world's lowest corruption perceptions according to Transparency International's index([9]), where Sweden ranks third out of 177 countries, and second among EU countries. However, in terms of regulatory environment, Sweden's performance is less remarkable when compared with rich-world economies. According to the latest 2008 data, Sweden ranked 15th out of 35 OECD countries in the integrated product market regulation indicator. According to this indicator, the level of regulatory restrictions in Sweden was only slightly lower than the OECD average.

Sweden's performance in R&D and innovation is particularly strong. At 3.4% in 2011, Sweden displays the second highest expenditure in R&D as a percentage of GDP in the EU, according to the auxiliary MIP scoreboard indicator. Sweden ranks first among EU countries in the Commission's composite indicator of innovation output, showing a strong performance in three of the four indicator components (namely, in patents, employment in knowledge-intensive activities and employment in fast-growing firms of innovative sectors)([10]). The share of high-tech products in exports, a factor which is also considered in the composite indicator, has been broadly stable at 14% since 2001. This figure is similar to that of the largest EU economies (Germany, France and the UK) and higher than that of the other EU Nordics (Denmark, Finland)([11]). Sweden is, likewise, the highest ranking member state in the Innovation Union scoreboard([12]), heading a group of four EU countries deemed as innovation leaders.

However, results in basic education have deteriorated. Sweden performed lower than the OECD average in all three educational categories (mathematics, reading and science) considered in the latest (2012) evaluation of 15- and 16-year olds of the Programme for International Student Assessment. This represents a sharp drop when compared with the first study (2000), where Swedish pupils outperformed most other countries. This negative evolution has happened despite Sweden's comparatively high investment in education. According to the European Commission's Education and Training Monitor([13]) "general government expenditure on education as a share of GDP was 6.8% in 2011 compared to the EU average of 5.3%, (…) one of the highest levels of public expenditure per student in the world."

Export profile specialised in services and selected industrial products

Sweden has a revealed comparative advantage([14]) in services other than transportation and tourism. It has generally no revealed comparative advantage in goods exports, with the exception of machinery/electrical products, metals and, notably, wood and wood products (Charts 3.10 and 3.11). Productivity in services is particularly strong. Commission analysis([15]) shows that Swedish multifactor productivity in market services was already high in 1995 and that it continued to grow at a robust average rate of 2.9% per year until 2007. Examples of particularly dynamic sectors include business services, computer and information services and R&D. Goods exports include traditional products, such as pulp, paper, saw mill products, iron and steel, as well as more innovation-driven products such as general and special-purpose machinery, car parts, pharmaceuticals and communication appliances. The very important pulp and paper industry is, however, under increased pressure as paper declines as a communication medium.

Intra-EU trade represents an important share of Swedish exports. In 2012, 57% of Swedish goods exports were destined to EU markets. There are several EU countries among the most important trade partners, although Norway, the main export partner, and the USA also represent very import extra-EU markets for Sweden (Graph 3.12).

Benign external dynamics, with limited spill-over effects

With the exception of constraints linked to the housing market, no policies seem to be in place that could significantly depress internal demand and unduly boost external surpluses. As regards the government sector, the 2014 Budget Bill envisages a moderate fiscal stimulus which, among other measures, will improve households' disposable income through tax cuts. According to the Commission 2014 winter forecast([16]), the new budgetary measures imply that the government will run a small structural deficit in 2014 and 2015. As regards monetary policy, the Riksbank cut its reference rate by 25bp in December 2013, to 0.75%, and the krona exchange rate has long been allowed to float freely. As new prudential instruments for stabilising the housing market are made available to the authorities, it is expected that interest rate policies can be aligned even more closely with domestic demand dynamics. Finally, as regards salaries, wage-setting is carried out by the social partners and ULCs have grown above euro area rates in the past two years. However, the constraints in the housing market discussed in Section 3.3 can contribute to lower residential investment, thereby increasing net savings and, concomitantly, the external surplus. As mentioned in European Commission (2012), "the driver of differences in investment between surplus and deficit countries thus was construction; (…) construction investment in the euro area periphery (and most new Member States) exceeded that of surplus countries by several percentage points of GDP. Sweden, Germany, Belgium, the Netherlands and Denmark displayed the lowest construction investment rates in the EU. This is relevant for understanding the relative growth of tradable and non-tradable sectors and external positions of countries."

The spill-over effects to other EU countries from a boost to Swedish demand would likely be limited and geographically concentrated. In 2012, the size of the Swedish economy was equivalent to 4% of euro area GDP. As a comparison, Germany represented 28% and the Netherlands 6%. The relatively small size of the Swedish economy means that the potential spill-over effects from an increase in internal demand are likewise small. Furthermore, given the geographical profile of Swedish imports and their importance in partner economies, the benefits from a boost to internal demand are likely to be concentrated in the Baltic area and Nordic countries (see Graph 3.13). Finally, it should be noted that Sweden derives its surplus from its trade relations with non-euro area countries. As depicted in Graph 3.14, Sweden has maintained a trade deficit with the euro area since 2007.

Overall external dynamics and the sizeable current account surplus appear to be largely benign. As previously seen, lower expected domestic savings, price effects from the appreciation of the krona and on-going competition from emerging economies imply that Sweden's external surplus is expected to decrease over the coming years. Coupled with the recent statistical revisions discussed in Box 3.1, this means that the current account scoreboard indicator of the MIP is expected to fall below the 6% threshold in the medium term. The relatively small size of the Swedish economy, the geographical profile of its imports, the fact that it maintains an external deficit with the euro area and the freely-floating regime of the krona highlight the limited potential of the Swedish economy for contributing to the external rebalancing processes that are underway in several euro area member states. Sweden can, however, benefit from an improved savings-investment balance by tackling underinvestment in residential construction. Ensuring high standards in basic education would help support high competitiveness levels in the long run.

3.1.2.   Net International Investment Position

A negative but sustainable international investment position

Sweden displays a moderately negative net international investment position (NIIP). It remained broadly stable at around -10% of GDP from 2008 to 2012 and is estimated to have improved in 2013 on the back of a sizeable current account surplus and favourable valuation changes (Graphs 3.15 and 3.16).

In the past decade, the composition of net external assets was dominated by investment in equity securities. Other positive foreign net asset positions are the result of Sweden's longstanding role as a net foreign direct investor, the strengthening of central bank reserves and the increase in the value of the loans granted abroad in recent years (counted as "other investment"). Since 2010, debt securities are the only instrument displaying a negative net position, which is largely due to the sizeable bond issuance by Swedish banks placed abroad.

In fact, the banking sector is the institutional sector driving the negative NIIP (Graph 3.17). Debt securities issued by Swedish banks accounted for nearly ¾ of total Swedish debt securities held abroad by year-end 2012([17]) and, according to Special Data Dissemination Standard (SDDS) data from the World Bank, by mid-2013 the external debt of Swedish banks made up 59% of total gross Swedish external debt. The possible implications of the funding strategies of Swedish banks for financial stability are discussed in Section 4.

The external position has been slowly improving and raises no sustainability concerns. At -10% of GDP in 2012, the Swedish NIIP is well within the MIP scoreboard threshold of -35%. Persistent current account surpluses, which are projected to continue in the medium term, have contributed to slowly improve the NIIP, although to a lesser extent than could be expected, as will be discussed. According to Commission projections, in the absence of large valuation changes, Sweden is expected to close its negative NIIP position over a 2-year horizon for a range of reasonable macroeconomic scenarios. The currency composition of the NIIP is also generally favourable, with the majority of liabilities being denominated in Swedish krona.

The income balance has improved in tandem with the NIIP and turned positive already in 2003, offering further cushioning to the current account. Although the NIIP is moderately negative, this has not prevented Sweden from deriving a net positive income flow from its external position. As will be discussed, this result is related to an underestimation of the true market value of the NIIP. Also, Sweden's net assets are mainly composed of equity investments, which typically have a higher yield than the debt securities which dominate Sweden's net liabilities (Graph 3.15).

Valuation losses have slowed the improvement in the external position

Sweden's international investment position falls significantly short of what could be expected from a two-decade long accumulation of current account surpluses. The Swedish current account has been in the black since 1994 and has posted large surpluses during most of this period. As shown in Graph 3.18, the continuous accumulation of surpluses should have automatically improved the NIIP to more than 50% of GDP by year-end 2012. The difference between this theoretical value and the actual value can be attributed to valuation effects associated with the underlying external assets and liabilities.

In fact, erratic valuation changes have become a major driver of the NIIP, as can be observed in Graph 3.16. This is a common feature of advanced EU economies. In Sweden's case, it is linked to the large increase in gross external assets and liabilities which began in the 1990s in the wake of financial liberalisation reforms. The large size of the external financial stocks and the asymmetric composition of assets and liabilities render the NIIP more sensitive to exchange rate and asset price movements. In fact, the vast majority of Swedish external assets are denominated in foreign currency while most liabilities are denominated in Swedish krona. Likewise, Swedish liabilities are largely exposed to the performance of the domestic market and to domestic interest rates([18]), while assets are significantly exposed to price movements in international markets.

The appreciation of the krona, the relative performance of the Swedish stock market and a decrease in domestic interest rates can help explain the adverse valuation effects. As depicted in Graph 3.4, valuation effects have been particularly relevant since 2002. From that year to 2012, the krona appreciated approximately 20%, contributing to decrease the value of Swedish foreign assets when measured in krona. Also during the same period, the Swedish stock market outperformed the wider EU market([19]) which would tend to increase the value of Swedish equity liabilities (embodied in the Swedish stock market) more than the value of Swedish equity assets held in the EU. Finally, Swedish interest rates decreased markedly from 2002 to 2012, hiking the prices of Swedish bond liabilities.

As expected, the volatility in valuation effects stem mostly from portfolio investment in equity and debt securities (Graph 3.19). This is both a consequence of the relatively large share that these assets represent in total gross external assets and liabilities as well as consequence of the fact that these financial instruments are marked to market and thus subject to volatile price changes. The gross amounts of other investments and foreign direct investment (FDI) are also sizeable. However, the former are largely made up of loans and deposits, which are not typically subject to large valuation changes, while FDI data from Eurostat is mostly reckoned at book value. Central bank reserves and financial derivatives represent comparatively small exposures and have had a residual impact on valuation changes.

The accounting of FDI at book value is an important factor explaining the underperformance of the Swedish NIIP. When measured at market value, the net impact of FDI valuation changes is much more favourable. Eurostat publishes the FDI component of the NIIP valued mostly at book value because FDI investments are often not listed in the stock market. The Riksbank, however, calculates estimates of the market value of FDI by adjusting its book value on the basis of the trend price-to‑earnings ratios prevalent in listed companies([20]). As can be seen in Graph 3.20, when estimated at market value, net FDI assets are much higher, increasing the NIIP from -10% to 8% of GDP in 2012.

Overall, the Swedish external position presents no concerns from a macroeconomic stability viewpoint. The moderately negative Swedish NIIP is sustainable and continues to move towards a positive balance. When FDI is estimated at market value, the NIIP is seen to have been in positive territory already since 2006.

3.2.      Private indebtedness: non-financial corporate sector and household indebtedness

This section analyses the recent evolution of non-financial corporate sector and household indebtedness to assess the sustainability of the current private debt level as well as the level of deleveraging pressure. It also probes into the measures taken so far to curb the trend in private debt and the current debate in Sweden on these issues.

3.2.1.   Evolution of indebtedness of the non-financial corporate sector

Consolidated corporate debt ([21]) amounted to 128% of GDP in 2012, far above the euro area average (80.9% of GDP). This level is lower than its peak in 2009 by 15.5 percentage points. The correction is likely to have continued in 2013 as the contribution of credit flows to the debt-to-GDP ratio has turned negative since Q4-2012. This evolution of corporate credit flows is mainly due to the fact that economic growth has been weak in 2012 and 2013 which has made some companies wait with new investments.

A large share of corporate debt takes the form of cross-border intra-group loans which entail lower risks than other types of borrowing. Cross-border intra-company loans increased markedly over the past decade before decreasing since 2009 (from 53% of GDP in 2009 to 41% in 2012). As explained in the previous reviews, the wide use of intra-group borrowing from abroad is motivated by efficient tax minimisation and not by a need to cover up for insufficient profits. From a sustainability point of view, intra-group loans entail lower risks than ordinary loans from banks or from other corporations. Multinational companies typically cover net debts in one country by net assets in affiliates in another country.

In this light, the debt-to-GDP ratio draws an exaggerated picture of indebtedness of Swedish corporations. If cross-border intra-company lending were deducted from the consolidated data, Swedish corporate debt would be below 90% of GDP in 2012. Also, the development of this measure over time would seem less worrying since it has been stable between 1995 and 2006 and accumulation of debt was limited to the period 2006-2009. Furthermore, a part of this accumulation has already been corrected, partly due to less generous rules on tax deductibility as further detailed below.  Given the drawbacks of the debt-to-GDP indicator in the Swedish context, it is useful to complement the analysis with other indicators of corporate sector leverage.

Balance-sheet indicators suggest a rather healthy situation of non-financial corporations. The ratio of gross debt-to-assets (51% in 2013) in Sweden is slightly below the euro area average (54%). The gross debt-to-equity ratio (55%) is even further below the euro area average (69%).

Moreover, debt to deflated financial assets has stabilised since 2009, indicating that no further accumulation has occurred for corporate debt. When regarding growing assets as a mitigating circumstance for mounting debt, some caution is needed as valuation effects can lead to an overestimation of the firms' ability to incur liabilities. These valuation effects are unstable and tend to evaporate in bust periods whereas the liabilities will not be deflated. If these effects are discounted for, the leverage ratio (called also "notional leverage"([22]) grew faster for Sweden between 2006 and 2008 but appears to have stabilised since 2009 (Graph 3.24).

Since 2009, corporate credit flows have been below 2% or even negative which moderated the evolution of corporate debt-to-GDP ratio (Graph 3.25). Furthermore, since Q1 and Q2 2013, the evolution of corporate debt to GDP has turned negative (Graph 3.21) which would indicate that not only there is no more build-up of corporate debt above the level of deflated assets since 2009 but that some deleveraging is taking place since Q1 2013. As explained below, this deleveraging seems driven by new rules on interest deductibility leading to a decrease in intra-group loans from abroad motivated by tax planning reasons.

3.2.2.   Factors behind Swedish corporate indebtedness and reforms undertaken

As developed in the 2013 review, the higher level of corporate debt in Sweden can be explained by several country-specific factors, in particular, a long period of easy access to credit and a strong presence of multinational companies coupled with the set-up of company taxation.

The strong expansion of corporate debt relative to GDP in 2006-2009 has been facilitated by the low interest rate environment and low perceived risk of lending to corporations. Following the economic slowdown in 2001-03 related to the bursting of the IT bubble, interest rates remained at very low levels for several years. This spurred high demand for credit by Swedish companies. Low interest rates coincided with a rather lenient lending attitude by Swedish banks towards both corporations and households. In view of a long period of a low rate of company defaults, the banks may have underestimated the risks associated with lending to corporations. This led to a strong expansion of credit to corporations in 2007-2008 (annual growth of credit flows above 14%). Still today, there are concerns that banks may assign in their internal risk assessment models too low risk weights to loans to the corporate sector. This is further developed in chapter 4.

Another feature distinguishing Sweden is the wide use of borrowing from foreign affiliates of the same corporate group, a phenomenon which is motivated by the set-up of company taxation in Sweden. The generous tax deductibility of interest payments has given an incentive to Swedish companies to finance their investments with debt rather than equity and invited an aggressive tax planning by multinationals operating in Sweden ([23]).

The government has taken some measures to limit the most far-reaching tax minimisation practices by multinational corporations. In 2009, tax provisions were introduced which reduced the tax deductibility of interest payments for intra-group loans related to the acquisition of shares from an affiliate. Following a series of public inquiries on various related issues, new legislation came into force in January 2013 which extends the restrictions on interest deductibility to all types of loans, regardless of their purpose (PWC, 2012). However, as was the case before, interest payments are tax deductible if the creditor is taxed on the interest at a rate of at least 10% should the interest income be the only income for the creditor (10% tax test), or the debt has been undertaken for "sound commercial reasons".   These measures, combined with the reduction of the corporate income tax rate from 26.3 to 22% at the beginning of 2013, are aiming at eliminating a large share of corporate debt which is solely driven by tax avoidance motives, in particular the intra-group loans from abroad.

According to the Swedish Tax Authority, multinationals have started adjusting to the new rules already in 2012, taking a more conservative approach to tax planning. In this light, intra-group loans have decreased already by SEK 80 billion from 2011 to 2012 (Graph 3.22) which represent a 5% decrease and are expected to decrease further in 2013. However, the concept of “sound commercial reason” remains ambiguous and its interpretation might lead to court cases.

The use of the deductibility of interest payments on corporates’ external loans as a tax planning tool is currently investigated. About 10-15% of external loans are estimated to be solely driven by tax planning considerations. The corporate tax environment incentivises corporations to incur both internal and external debt for tax planning purposes, and promotes a debt versus equity bias in the financing of corporations. Moreover, the Tax Authority estimates that out of SEK 90 billion revenues from corporate taxes, SEK 4-5 billion are foregone in tax revenues due to aggressive tax planning. In order to analyse the issues relating to corporate taxation further, a corporate tax committee has been set up by the government. This committee's mission is to propose tax incentives to stimulate the availability of risk capital, to propose increased tax incentives for research and development and to make proposals for a more comprehensive corporate tax system, if possible replacing the current interest deduction limitations and the debt-bias against equity. Its report is due in June 2014.

To conclude, corporate debt relative to GDP is still high but decreasing in a low growth and investment context. Moreover, balance sheet indicators suggest a healthy financial situation for non-financial corporations. Recent taxation reforms are aiming at decreasing the tax-planning component of corporate debt and further investigations to render the corporate tax environment more neutral are on-going. The evolution of credit flows needs to be monitored, especially once economic growth rebounds.

3.2.3.   Evolution  of household indebtedness

Household debt stands at more than 83% of GDP, or 158% of gross disposable income in Sweden in 2012. The lion's share of household debt is made up of mortgages, the build-up of which has gone hand in hand with rising housing prices over the last 15 years (Graphs 3.27 and 3.28). At present, credit growth for mortgage loans is less expansive but still amounted to 3.3% in 2012 and still outpaces GDP growth (Graph 3.29) driving the household debt-to-GDP ratio further upward. The jump in households' debt-to-GDP ratio observed in 2009 is explained by the sharp recession that occurred in Sweden that year (-5% real GDP growth) which made real growth an additional contributor to the debt to GDP built up (Graph 3.29).

 

Despite the upward trend observed in household debt-to-GDP ratios, the debt-to-total assets or leverage indicator remained flat indicating that both debt and assets have increased in the same proportion. However, an assessment of debt sustainability should also take into account valuation effects, particularly on the asset side, given that it is much more volatile than debt instruments, especially in a downturn when indebtedness (stock) adjustments tend to last longer on average. Against this background, filtering for valuation effects in both financial and non-financial assets yields the concept of notional leverage (i.e. debt to notional or deflated assets), which represents a better indication of the ability of households to incur liabilities ([24]). As can be seen in Graph 3.30, households' debt has increased faster than households' notional assets. The existing gap between the actual and the notional leverage ratio, which mainly accumulated during the 2002-2008 period, allows defining an upper-band of potential deleveraging pressures which would amount to 20 pp. in 2011.

The sharp increase in mortgage lending over the last 20 years has gone hand in hand with several trends which can render households less financially resilient ([25]):

· Between 2002 and 2010, the loan-to-value ratio of new mortgages has increased from 59% to 71% ([26]). This trend was curbed after the introduction of a 85% loan-to-value cap in 2010.

· In parallel, the actual repayment periods as well as the share of unamortised loans have been growing.

· Variable interest rate loans ([27]) have become predominant making households more sensitive to interest rate changes ([28]). Mortgage rates used to be reset every five years to the current market rate. However, since the 1990s, fixed interest terms have gradually been shortened.

· Possible income loss due to unemployment or sickness has become more significant. Since the 1990s the unemployment and sickness insurance schemes have changed and an increasing number of people are reaching the unemployment benefits payment ceiling, which has not been increased since 2002.

The stress tests of the Financial Supervisory Authority (FSA) performed in the 2013 mortgage market survey ([29]) highlight the macroeconomic risk on consumption and therefore on banks' lending to non-financial corporations while it sees only limited risks that banks will experience major loan losses on mortgages. The FSA analysed several contagion routes.

Firstly, based on FSA calculations of households' discretionary incomes ([30]), 9% of the households in the 2012 mortgage sample have a deficit (i.e., a negative discretionary income) and one out of five has less than SEK 3,000 to live on. These results are higher than the data reported by the banks. The discrepancy in the calculation of discretionary income by banks and the FSA will be further investigated by the FSA.

Secondly, a loss of income or an increase in interest rates, considered to be currently at historically low levels, would mean that households' budget for consumption of items other than housing costs would decrease. The FSA estimates that an increase of the interest rate by five percentage points would increase the number of households in deficit by 5% and up to 9% if these households are required to simultaneously amortise the part of their loan with the highest loan-to-value (LTV) ratio.

Finally, a fall in housing prices could lead households to seek to restore equilibrium in their balance sheets by increased savings which would also have a negative impact on consumption. A stress test shows that if prices were to fall by 15%, around 11% of the households would face a negative equity.

These effects on consumption would affect non-financial corporations and in the longer-run could affect the level of losses in banks' lending to non-financial corporations.

The combination of high indebtedness, high loan-to-value ratios and more variable interest rates means that households are more vulnerable in terms of interest rate risk and income risk. Following changes to the Swedish welfare system, the impact of such risks is nowadays less mitigated through social safety nets and would probably affect consumption.

3.2.4.   Measures taken so far and current debate on household indebtedness

Over the last few years, several measures have been taken by the FSA, the Swedish Bankers' Association and the Government to curb the upward trend in households' debt and improve households' financial resilience as well as banks' stability. Moreover, as described in Chapter 4, the government has clarified the allocation of macroprudential supervision responsibilities by allocating them to the FSA.

The FSA has adopted an incremental approach to set the conditions for a gradual household debt deleveraging process with the introduction of a mortgage cap, the increase of risk weight floors on mortgage loans, and an agreement with the Swedish Bankers' Association to promote an amortisation culture:

· The 85% LTV cap on mortgages implemented since 2010 has proven effective in breaking the upward trend in LTV increase since 2002. This cap decreases the risk for households of being exposed to negative equity in the event of a fall in house price. As shown in Graph 3.31, the LTV ratio on new loans reached a peak in 2010 and decreased below 70% since. Moreover, the share of mortgages having a LTV above 85% has decreased (Graph 3.35). This cap seems to have a normative effect as 12.5% of the new loans have an LTV of exactly 85% (27% for the age-group below 26 years old), i.e. without choosing the still possible option to take an additional and, thus, unsecured loan.

· Early 2013, the risk weight floors applied to mortgage loans were introduced at 15% ensuring that banks hold more equity given the credit risks present in their mortgage activities. This is further analysed in chapter 4.

· With the support of the Swedish Bankers' Association, the FSA is promoting an amortisation culture. Since 2010, the Swedish Bankers' Association recommends that loans with a LTV above 75% are amortised to this level. Furthermore, in October 2013, the FSA issued a recommendation that banks should provide customers with individually tailored amortisation plans ([31]). According to the recommendation, a discussion between the mortgage company and the borrower about amortisation should take place before the final amortisation plan is proposed. Also, this plan should be in the best long-term interest of the customer. The objective is to allow households to make a decision about amortisation by having the banks clearly showing how amortisation affects a household's finances in the long-run. However, this recommendation is so far not legally binding as the Consumer Credit Act needs to be amended by the Government for the FSA to be able to implement a regulation in the Banking and Financing Business Act and it does not make amortisation as such compulsory. The Swedish Bankers' Association has endorsed this recommendation and is working with banks to develop a template for the individual amortisation plan. It is foreseen to enter in use mid-2014. In a recent memorandum ([32]), the FSA takes note that since 2010, repayment periods for top and unsecured loans for new borrowers have decreased but due to very long repayment periods on bottom loans the aggregate debt ratio is still held up. In addition, figures on net amortisation (i.e. gross amortisation minus new loans) suggest an overall increasing indebtedness.

In November 2013, the Government has proposed a new regulation on fairer rules for repayment of mortgages ([33]). Due to its construction the current model could be extremely expensive for a customer who wants to repay his fixed rate mortgage. The new regulation has two main components and is intended to be implemented by 1 July 2014:

· The Government proposes a new model for the calculation of interest rate compensation for the banks, corresponding to the actual costs for the bank. According to the proposal, the interest rate compensation would be calculated on the basis of the mortgage bonds market instead of the government bonds market. A lower interest rate compensation would facilitate the possibility for the consumer to switch banks and thus increase the competition in the market.

· The proposal limits the creditor's right to ask for the repayment of a mortgage in a situation where the value of the security for the loan has decreased as a result of a general downturn in the housing market. From a financial stability perspective, this measure restricts the banks' option to transfer the risk of a general market fall to the household sector. This measure might incentivise the banks to further increase the soundness and sustainability of their credit issuing practices.

Despite all these measures, household debt to GDP is still rising which keeps the debate on further measures to be implemented quite intense. While the FSA and the Government seem to be willing to follow an incremental approach allowing for time to assess each measure before implementing new ones, the Riksbank expresses more concerns about the situation and has adopted a cautious approach regarding its monetary policy arguing that it takes time for the macroprudential tools to be fully operational. Moreover, the Riksbank is in favour of the risk weight floor on mortgage loans to be increased gradually to 35% and points out that countercyclical capital buffers would probably have been activated if such a regulation had been in place nowadays ([34]). The FSA ([35]) is reviewing the discretionary income calculations made by banks during the credit assessment to prevent unsound credit issuing practices and has announced a further increase to 25% of the risk weight floor on mortgage loans. This announcement is further described in chapter 4.

Finally, all stakeholders agree that, even if the measures implemented so far are improving households' financial resilience and banks' stability, they are not solving the fundamental structural problems on the housing market that also influence household debt. This question is further discussed in 3.3.

3.2.5.   Private indebtedness: assessment of imbalance

Private indebtedness in Sweden is high. However, as shown on Graph 3.36 ([36]) Sweden benefits currently from limited deleveraging pressures. Overall loan demand and supply pressures used in Graph 3.36 are composite indicators. On the loan demand side the willingness of households and non-financial corporations to take more debt is proxied by the consumer confidence and economic sentiment indicators; unemployment rate and house price evolution are also included in the set of variables. The loan supply indicator includes variables assessing financial soundness (e.g. change in overall non-performing loans, Tier 1 capital ratio and banks' return on equity) ([37]), and indicators addressing the link between sovereigns and the banking sector (e.g. the sovereign CDS spreads, banks' exposure to high risk foreign claims).

Under this favourable loan supply and demand environment, deleveraging could occur through moderate yet positive credit flows. This is currently happening for non-financial corporate debt where credit flows, growth and inflation are contributing to some deleveraging (Graph 3.21).

This trend is not yet observed for household debt where the contribution of credit flows to debt is more important than growth and inflation effects (Graph 3.29). Moreover, the improvements to disposable income included in the 2014 general government budget might translate into further housing investments and debt build-up. This points to the fact that despite the measures taken so far, the conditions for high household credit flows are still in place: low interest rates, tax incentives through the mortgage interest payment deductibility and a low supply of new housing.

Even though the probability of credit losses on Swedish mortgages seems low at the current juncture, macroeconomic risks exist. The combined structural evolutions over the last 20 years of higher household indebtedness, larger proportions of variable mortgage rates, higher loan-to-value ratios, low amortisation and changes to the unemployment and sickness insurance make households less financially resilient than before the 1990s crisis. When the expenses of households increase, for instance via higher interest burden due to increased mortgage rates, or their income decrease, for instance in case of unemployment, this most likely will lead to reduced consumption implying lower demand for goods and services. This would have a negative impact on non-financial corporations, which can result in credit losses in other parts of the banks' loan portfolio.

To conclude, while no imbalance is identified regarding corporate indebtedness, the evolution of corporate debt needs to be monitored, in particular once companies will see increasing investment opportunities. In contrast, the analysis demonstrated that Sweden still experiences a macroeconomic imbalance concerning households' indebtedness. To address it, further policy measures are necessary notably in the field of housing supply and taxation, to remove the debt-bias shifting households' preferences towards home ownership.

3.3.      The housing market

3.3.1.   House price developments

During the last two decades, the Swedish housing market has been characterised by a steady and sharp price growth. According to OECD housing data, since 1994 Q1 (beginning of the recovery after the burst of the property bubble in the early 90s in Sweden), nominal house prices increased by more than 205%, while real house prices increased by 128% in Sweden until 2013 Q3 (compared to the European average of 80% and 25%, respectively)([38]).

House prices in Sweden weathered well the crisis that started at the end of 2008. Real house prices corrected in 2009 but sharply rebounded in 2010 reaching new record levels. Between 2008 Q4 and 2013 Q3, real house prices increased by 7.3% while the European average declined by 7.7% for the same period. Sweden has not experienced a large correction in house prices such as the one that took place in, for example, in Denmark: house price developments have been relatively smooth without abrupt volatility.

Relatively stable house prices positively affected household confidence and consumption thereby softening the impact of the recession on the economy. Since 2010, house prices remained relatively stable: nominal house prices increased by 4%, while real prices decreased by -0.5% between 2010 Q1 and 2013 Q3. Accordingly, the Commission's Alert Mechanism Report scoreboard indicator has not shown year-on-year growth rates of deflated house prices above the 6% threshold since 2010([39]).

Last years' IDRs showed that though some indicators pointed at an overvaluation (for instance affordability and dividend ratios), others suggested a development in line with underlying fundamentals. They also concluded that the Swedish housing market is characterised by several inefficiencies which could impact the underlying fundamentals driving house price developments. 

House prices are fuelling private indebtedness and can have an important impact on consumption, economic performance and financial stability. Therefore, a detailed analysis of the main challenges is necessary to avoid the building up imbalances that could impede future growth. It seems important to closely watch house price developments as, especially when in combination with favourable credit conditions, they could lead to risks of overheating.

Several different indicators are used to assess whether house prices are in line with their fundamentals or whether there is a significant misalignment. Traditional indicators, such as affordability (price-to-income) and dividend (price-to-rent) ratios suggest that house price levels are above their long-term average. In theory, these ratios tend to revert to their long-term average. Therefore, the positive gap between the latter and the actual value provides an indication of a potential overvaluation at the magnitude of 15-30%.

Conclusions on potential overvaluation based on these indicators have to be considered with caution due to their simplifying assumptions ([40]). Most importantly, long term trend lines based on historical averages cannot capture any possible structural shifts (for instance, a change in the cost of home ownership) or altered fundamentals (such as lower mortgage rates or lower taxes) which might justify higher prices than the historical norms. The price to rental ratio can also overestimate valuation gaps in Sweden, as rental prices are highly regulated and on average set below market prices. On the other hand, the graph underlines the main trends delineated above: house prices are not growing excessively since 2010, but rather switched to a stabilising path.

When looking at the price to income ratio([41]),  following a correction after 2007 and a bouncing back in 2010, the ratio is remained at a higher level than the euro area or OECD average. House prices compared to the disposable income are on a moderate downward trend since 2010, not showing the same sharp correction as other countries facing a house price bubble. The ratio reached its nadir around 1996 in Sweden following the burst of the property bubble in the early 90s. It then recovered and moved along the euro area average until the onset of the crisis. Then it spiked moderately compared to the euro area average, because of the collapse of the housing market in other EU Member States and because of the abolition of taxation of wealth, changes to property taxation and withdrawal of direct subsidies in 2005 that had their full impact on house prices by 2007.

An econometric model developed by the Commission services ([42]) seems to confirm the above finding. The model suggests that house price levels in Sweden have been correcting back to their main supply and demand fundamentals since the peak of 2007 and follow a correction cycle in line with past cycles. As depicted in the graph below, Swedish house prices went through a long and intensive boost period between 1996 and 2007 compared to their main supply and demand fundamentals. Since 2007, a downward correction is taking place with prices converging back to their fundamental levels.

In summary, when compared to their fundamentals, Swedish house prices have been stabilising and correcting since 2010 at a moderate pace, easing immediate concerns of an overheated house market. The fundamental indicators do not suggest growing imbalances; rather we can witness a stabilisation period where house prices slowly converge to their fundamental values. However, in light of the high level of private indebtedness, it seems important to continue the stabilisation path; otherwise, house prices could pose a risk for the macroeconomic stability of the country.

3.3.2.   Underlying market fundamentals

House price developments need to be analysed together with the underlying fundamentals and with other potential inefficiencies on the housing market, as these factors provide necessary insights on future house prices developments. Distorted fundamentals or serious structural imbalances on the market could revert the current stabilisation path and drive prices above their fundamental values, thereby renewing concerns of growing imbalances on this market. Furthermore, distorted fundamentals would bias upwards the fundamental value for house prices and overvalued fundamentals could underestimate the misalignment of house prices.

Key facts about the Swedish housing market

For a population of 9.5 million, Sweden counts roughly 4.5 million dwellings, divided between approximately 2.5 million apartments in multi-dwelling-buildings and 2 million single-family-buildings.  The apartment segment consists of roughly 1.6 million rental apartments (35% of the total housing), 0.95 million tenant-owned([43]) ones and a mere 566 owner-occupied condominiums (Statistics Sweden, 2013a).

The main trends regarding the structure of the housing market, as explained in last years' IDRs, have continued in 2013. Tenant-owned apartments are continuously gaining in popularity causing an increased level of new constructions and the conversion of rental apartments. The high growth in the number of tenant-owned apartments over the last decade has also been coupled with the highest price increase for dwellings of this type of tenure.

The rental market([44]) has a pivotal importance on the Swedish housing market. Approximately 40% of the Swedish population is living in rented dwellings, which is slightly higher than the EU average. The proportion of rental units has been steadily decreasing in the last decade, in particular as regards private rental units.

Price variations on the housing market are significant across the country. The main urban centres, in particular the Stockholm region, witnessed higher price increases than any other region. In Stockholm city, prices can be several times higher than in other areas of the country([45]). While a higher income level can explain part of the differences, other Stockholm-specific features seem to drive up the prices in this area (high demand due to demographic and economic factors and restricted supply).

Main drivers of the house price increase

House prices have been primarily fuelled by favourable fundamentals of the economy. The following section will assess these drivers and whether particular risks can be identified in the short term.

A steadily rising population and rising population per dwelling in some major cities([46]) increase demand for residential property. House price increases were particularly strong in the urban areas of Sweden, mainly in Stockholm and Gothenburg([47]). Population has grown the most in Stockholm, accounting for roughly half of the population growth of the entire country (Swedish National Board of Housing, Building and Planning (2012))([48]).

The economic fundamentals of the country have been particularly strong, withstanding most of the negative impact of the crisis. Sound economic policy led to stable and increasing household confidence, while low inflation, low interest rates and favourable tax measures implemented by the Swedish government kept increasing the disposable income of households (see Section 3.2 above). The high share of variable mortgage interest rates translates Riksbank's rate cuts to lower monthly mortgage payments([49]). The economic fundamentals are expected to further ameliorate in the short term (in line with the European Economic Forecast - Winter 2014) indicating a continued increase of disposable incomes.

Credit conditions proved to have been highly favourable in recent years. The availability of amortisation-free mortgages allows low monthly payments, which were further decreased by the record low mortgage interest rates (as a result of the Riksbank repo rate cuts)([50]). Recent initiatives of the Swedish authorities started to tighten these trends considering the growing risk on financial stability. The macroprudential measures introduced by Swedish Financial Supervisory Authority seem to have a normative impact on the market as regards LTV ratios (Swedish Financial Supervisory Authority, 2013c). However, these macroprudential measures are expected to only marginally push up the total interest costs for mortgage borrowers, therefore their impact on demand will likely stay limited (National Institute of Economic Research, 2013).

Taxation conditions create incentives for households to purchase their own houses. The Swedish tax system offers a generous deductibility of mortgage interest payments from the income tax. At the same time, following the reform of property taxation, the recurrent property taxes have been substantially reduced and have in practice ceased to counterweight the effect of the tax relief on mortgages. The possibility for home owners to make tax deduction from their income tax for house renovation has been introduced in 2008 providing additional incentives for home ownership vs. renting (Swedish Property Federation et al, 2013). As a result, Sweden has one of the highest tax incentives in the EU for home ownership.

In addition to the above demand-side factors, several other supply factors contributed to the steady growth of house prices fuelling potential imbalances in this sector.

Lack of sufficient housing supply, in particular in the main urban areas, is translated into particular sharp price increases. Investments in the housing sector have been significantly lower than in other European countries, causing increasing house shortages mainly in urban areas. The under-supply of housing can result in significant welfare loss. Swedbank (2013) estimates that the absence of growth in housing could cost SEK 21 billion p.a. over the next 20 years. A study by Jahnson and Lundberg (2013) estimated that prices on the Swedish housing market today are roughly one third higher than if constructions had developed in the same way as in Finland over the past 15 years.  On the other hand, since housing investment in Sweden was low for a long period, the risk of a substantial downturn of house prices is limited (National Institute of Economic Research, 2013).

Distorted incentives can prevent efficient use of the existing housing stock thereby further bolstering house price increases. Two main reasons can be identified for such a lock-in effect. Firstly, due to highly regulated prices, rental prices and market prices could deviate substantially. Such a deviation causes excessive demand on the rental market, fuelling house price increases in substitute segments (such as tenant-owed apartments). Secondly, existing housing taxation does not support mobility. Following the 2008 tax reforms, property taxes have been decreased sharply([51]), while transaction taxes have been changed resulting in substantially higher taxes paid by the seller (22% of the capital gains). As a result, homeowners' incentives for selling houses have been reduced.

The lack of sufficient housing implies social and economic constraints on growth. Housing shortage seems to have become a widespread phenomenon across Sweden: 43% of municipalities report a shortage of housing and 85% of them report a shortage of rented apartments (Swedish National Board of Housing, Building and Planning, 2013c). Lack of sufficient supply is particularly pertinent in the Stockholm and Gothenburg regions, which experience a large population inflow resulting in an excessive waiting list for rental apartments([52]). Insufficient housing supply in particular affects vulnerable groups (immigrants, students, new entrants to the labour market) that do not get access to adequate housing thereby hampering labour mobility.

In light of the information above, the supply side of the housing market merits a more in-depth analysis of the reasons for low construction activity and of the imbalances on the rental market. Accordingly, the analysis is divided in two main sections: (1) potential constraints on new investments; and (2) the utilisation of existing housing stock focusing on the rental market. 

3.3.3.   Supply side challenges: New constructions

Despite continuous and almost uninterrupted growth of house prices and steadily growing demand for housing, construction activity in Sweden remains subdued and persistently below other EU countries. As depicted in the graph below, since its peak in 2007 (when policy changes explained in Section 3.3.1 triggered a peak in construction activity), residential investments are on a decreasing trend.

From 2002 to 2013, housing investment amounted to 3.2% of GDP on average in Sweden, while the corresponding figure for the EU28 stood at 5.4% of GDP. Increasing demand in the housing sector is not matched by an increased level of new construction activity. As a result, Sweden has built up a considerable investment gap in the housing sector([53]).

The high proportion of a reported housing shortage shows the insufficiency of the existing stock. On the one hand, the population per dwelling has been decreasing continuously in the last 10 years in Sweden. Thus, the existing stock of housing might be sufficient to satisfy the housing needs of the population. In addition, the relative underinvestment in this decade might only compensate for the relative overinvestments in the decade before([54]). On the other hand, although population per dwelling has been decreasing in the last decade in Sweden overall, it goes up in the main urban areas. More importantly, the population per dwelling ratio does not explain other fundamental factors contributing to increased demand such as increasing incomes and decreasing cost of home-ownership.

The main inefficiencies holding back stronger growth of new constructions are explained below.

The particularly long and complex planning and zoning process increases building time and uncertainty. From the idea until the newly constructed building can be occupied, some 20 different steps have to be taken (County Administrative Board (2013)). In addition, appeals can be launched against the construction plan in several stages of this process. All these administrative barriers could render the process extremely long, up to even 8-10 years, compared to, for instance, the average 3 years in Germany (Hüfner and Lundsgaard (2007), Ministry of Social Affairs (2013), Müge Adalet McGowan (2013)).

Local municipalities have the planning monopoly in Sweden. They impose different standards and requirements resulting in a fragmented market with different requirements across the country.

Local municipalities have a core role in construction, since most of them own a large share of the land eligible for construction (for instance, Stockholm owns 70% of the land), while at the same time they are overseeing the planning process (County Administrative Board, 2013). However, municipalities do not have (financial) incentives to support construction activities. Tax revenues increase, if at all, only moderately and in the mid-term, while increased residential construction will entail public infrastructure investments to be financed by the municipalities.

The remaining land for construction is mainly owned by large construction companies. Since the value of the land is increasing at a higher pace than the value of new housing, large construction companies can "wait-and-see" by initiating only limited new constructions and rather capitalising on rising land prices (Swedish Competition Authority, 2013).

In addition, the sale of land for construction is not always conducted in a transparent manner. Possible investment projects are often discussed bilaterally between a few established stakeholders excluding other potential developers. Accordingly, the Swedish Competition Authority suggests that municipal land sales may need to be regulated, municipal land should be sold mainly through tenders, and that the sale/transfer of land rights of the land that is about to be developed, should occur earlier in the planning process (Swedish Competition Authority, 2013).

Construction costs have kept rising and now count among the highest in an international perspective, driven by large increases in land costs.  Construction costs are now much higher than in other peer countries, such as Finland, impeding constructions (Hüfner and Lundsgaard, 2007), in particular regarding rental dwellings where prices do not reflect market prices due to the regulations. Although the rental prices for new constructions can be set high and more in line with their real market value for 15 years, a recent study of the Swedish National Board of Housing, Building and Planning (2013b) argues that the relatively low rents for attractive homes still help to keep the housing construction depressed. Property developers plan on a longer timeframe than 15 years in accordance with the amortisation period of housing units. Therefore 15 years of more realistic rents do still not give sufficient incentives to build new rental units.

The structure of the construction market is characterised by weak competition as four major companies dominate the market. Entry barriers on the market are high deriving among other factors from the extensive ownership of land, exclusive contacts with local authorities, and complex and diverging building requirements across the country. As a result, the market tends to disfavour small companies and discourages competition (Swedish Competition Authority, 2013).

Construction subsidies in Sweden are exclusively used for the construction of dwellings for the elderly. However, due to the high constraints on certain segments of the market, in particular for student housing and affordable housing, additional support might be necessary.

The Swedish government has taken several steps to strengthen the market signals in this sector, to let supply respond more effectively to demand and to tackle the above referred constraints on the housing market in order to speed up the construction of new houses.

Most importantly, the government has taken several efforts to streamline and shorten the planning and zoning requirements, for instance, by initiating several amendments to the Planning and Building Law (Plan- och bygglagen). To speed up the assessment of appeals, the 2013 Budget provided additional resources for processing appeals at the county administrative boards. As a result, the planning process became faster (reduced on average to 6-7 years according to the industry experts), and will be further simplified and shortened according to the most recent government proposals (Ministry of Social Affairs, 2014).

A legislative proposal relating to the responsibility of municipalities for the provision of housing has entered into force since 1 January 2014, clarifying that in addition to simply evaluating the housing needs within its own boundaries, each municipality also has to take the regional perspective and, together with neighbouring municipalities, ensure that housing needs are being met ("Bostadsförsörjningslag").

In 2013, the Swedish government revised the construction standards, known as Eurocodes. These standards had to be purchased from the standardisation body SIS at a cost that was perceived to be significant for small companies, i.e. can potentially restrict competition. From 1 July 2013, these standards are accessible for free.

In order to build more student housing and dwellings specifically dedicated to young adults, more generous rules relating to temporary dwellings have been introduced, lowering requirements in terms of the standard of the dwellings. The proposal on temporary dwellings for students will come into force on 1 July 2014. Since this special segment of housing is mainly built for social reasons, and typically not provided by the sector on market terms, additional measures are experimented to overcome the bottleneck in this specific segment. Municipalities started to offer the possibility to rent land for longer terms to incentivise construction of student housing, thereby decreasing the construction costs. In addition, the government plans to grant a subsidy of SEK 50 million to promote the construction of housing for young people([55]).

Several further initiatives are currently being discussed or finalised to boost new constructions by reducing the existing high administrative burden. For instance, a review of the planning process is under way to further cut red tape. The Swedish National Board of Housing, Building and Planning is currently working to limit the possibility of municipalities to introduce specific technical requirements going beyond the common set of rules and to streamline appeal procedures (Statens Offentliga Utredningar, 2012:86). Another review has been commissioned to avoid diverging interpretations of existing legislation linked to noise levels in construction and housing.

There is a general consensus among stakeholders that all these steps are pointing to the right direction, albeit their impact is perceived to be small in the short term. Further measures would be needed to increase the responsiveness of the supply side to demand.

In the longer term, large infrastructure projects planned in the Stockholm region can also have significant impact on raising supply. An agreement was signed in 2013 to extend metro lines in Stockholm which could involve the construction of 78.000 additional homes until 2030 (County Administrative Board of Stockholm, 2013).

From the low levels of 2012, construction activity strongly rebounded in 2013 in particular in the last quarter of the year according to the preliminary estimates of Statistics Swedish. In 2013, the number of dwelling starts in newly constructed buildings increased by 28%, although there is a wide diversity behind the overall growth figure([56]) (Statistics Sweden, 2014c). The dynamically increasing number of housing permits and the increasing confidence index for building and manufacturing suggests that growth in the construction sector might continue at an increasing speed in the near future, driven by the Stockholm region and by the construction of multi-dwelling buildings (National Institute of Economic Research, 2013).

It would be premature to see whether there is a trend change in construction and whether the investment gap in this area will decrease. If the construction level remains suppressed the low level of housing supply and strong demand pressures will further capitalise into housing prices, in particular in the main urban areas.

3.3.4.   Supply side challenges: The rental market and utilisation of the existing housing stock

A well-developed and efficient rental market providing a viable alternative to ownership plays a balancing role by alleviating house price pressures and smoothing housing market dynamics. This is especially the case when it proves to be an affordable platform for young and low-income households, providing them with a viable alternative to a hasty first step into the 'property ladder'.

The Swedish rental market in certain areas is characterised by growing structural imbalances, in particular, in dynamically growing urban areas. Such ineffectiveness can be witnessed in the long waiting queues for rental dwellings, low turnover of rental units, a developing black market (for instance, sub-letting without the permission of the landlord and at rent levels higher than those of equivalent rental units, or trade with rental contracts) and suppressed construction of rental units (Lind, 2013). 

Market inefficiencies are primarily attributed to the high level of rent control. Sweden is characterised by one of the highest levels of rent control among the EU Member States ([57]).Rent levels are negotiated between the Swedish Union of Tenants (Hyresgästföreningen) and the housing companies. Rental prices are based on a rent valuation model based on a set of characteristics defined as the so-called 'utility value' of any given dwelling. These characteristics include factors such as the level of standard, services offered and the condition of the dwelling, and also factors in renovations or increases of quality undertaken in the dwelling. However, factors pertaining to the actual market value of rental leases (for example the relative proximity to the city centre) have not, until recently, been included. On the other side, rental apartments located in relatively unattractive municipalities with limited employment opportunities, may nevertheless have high rents if they are of a high standard (Statens offentliga utredningar, 2012:86).

The rental prices resulting from these negotiations could greatly vary across municipalities. For instance, in some areas rental prices are considered to reflect the market prices (such as the Malmö region), in other areas, a significant gap has opened up between them. In general, the rent to income ratio has been decreasing sharply in Sweden, much sharper than in other EU countries.

The difference between market based and regulated rent levels is apparent in light of the easing of some conditions for newly constructed rental apartments. Landlords of newly constructed rental units may charge rents departing from the utility value system, on the condition that they get the agreement of the Union of Tenants. As depicted in graph below, the rent for newly built apartments is considerably higher than for the existing stock.

Wide divergence between actual and market rents triggers excessive demand for rental units and creates a lock-in effect since existing tenants would not want to give up their favourable conditions for renting the apartment. Inefficient use of the existing rental units could, thus, also contribute to the supply side constraints of housing. A recent report of the Swedish National Board of Housing, Building and Planning (2013b) argues that a welfare loss at the magnitude of SEK 10 billion per year stems from inefficient rental market, while only relativity limited welfare loss can be attributed to a lack of new constructions. The report argues that most of the loss can be attributed to the Stockholm and Gothenburg cities (SEK 8.3 billion and SEK 1.7 billion, respectively) where the rental market is characterised by the largest inefficiencies due to their large deviation from the market price([58]). By contrast, in other parts of the country, where rental prices are more in line with the market prices and/or not characterised by excessive demand for rental units, the potential welfare loss and the underlying inefficiencies are much more marginal. The study also argues that as a result of such inefficiencies, an undersupply of 40.000 rental dwellings has been accumulated in the country (out of which 27.000 in Stockholm).

Increased demand for rental units, however, is not met by increased supply: on the contrary, the number of rental units is decreasing the most where faced with the strongest demand. Between 1990 and 2012, the stock of apartments has increased by more than 500.000 but for rental apartments it increased by a mere 1.650 units, (Mattson-Linnala et al (2013)). The overall stock of rental dwellings has been decreasing steadily since 1997, by roughly 5%, from 1.746.490 apartments to 1.653.347 apartments in 2012. This decrease can be attributed to government incentives for home ownership, conversion of rental units to tenant-owned apartments ([59]), demolition of existing stock and limited new rental dwelling constructions. Again, the decrease has been uneven. Whereas in some parts of the country the number of rental units even increased (such as Uppsala or Örebrö), the number has decreased by almost 20% in Stockholm (Swedish National Board of Housing, 2013b).

The insufficient supply on the rental market creates a pressure on other type of housing as well, most notably on tenant-owned apartments (which are the closest substitute in urban areas) further intensifying their price increase. A recent paper argues that if Sweden did not apply the utility value system for the rental market, house prices would likely be lower, and private indebtedness in 2013 would have reached 157% of the disposable income instead of the current level of 172% (Evidens, 2013).

The Swedish government has taken several steps on the rental market to tackle the above referred shortcomings.

The government started to relax rules to allow tenant-owners to demand rents that would make private letting profitable. In principle, private lettors can charge prices outside the utility value system([60]) subject to the agreement of the housing cooperative association. These changes entered into force on 1 February 2013 and there are some expectations of rising rent levels for privately let apartments in urban areas, in particular in Stockholm city.

As of 1 February 2013, private renters are also involved in the rent setting negotiations. Despite some initial expectations, however, this has not led to different rental price setting in privately owned dwellings due to the opposition of the Tenants' Union and subsequent legal challenges.

There have also been some attempts by landlords to increase rents on the stock of rental dwellings through the so-called Stockholm model (Stockholmsmodellen). The model aims at better reflecting the location of the apartments in the utility value setting, thereby increasing the rental prices for the most attractive areas. However, the model has been challenged by the Swedish Tenants' Union at the Rent Tribunal, which rejected the model. The decision of the Rent Tribunal has been appealed by the Swedish Property Federation; thus, further legal disputes are expected.

The impact of these recent initiatives cannot yet be fully assessed. In light of the forthcoming elections, considering the sensitivity of the rental market issues, no new measures are expected in this field this year.

3.3.5.   Outlook and conclusions

In the absence of major macroeconomic turbulences or government interventions, further price increases are expected on the housing market in the near future driven by favourable demand conditions. In the medium term, increasing construction levels, higher repo rates and the introduced macroprudential measures would dampen house price increases, but other favourable demand factors (most notably favourable taxation) and supply inefficiencies will provide further strong upward pressure on house prices.

Housing market developments in Sweden, in particular the absence of a house price crash, have so far contributed to limit the negative impact of the crisis by sustaining private consumption. The relatively low level of construction also suggests that the risk for sharp house price adjustments is relatively small (National Institute of Economic Research, 2013). On the other hand, house prices need to further stabilise to avoid driving private indebtedness further, to avoid overheating and abrupt volatility that could renew concerns on imbalances and of macroeconomic risks. The renewed growth of private indebtedness in 2013 driven by mortgage loans in this sense raises some concerns([61]).

Besides macroprudential measures, other effective measures are likely to be necessary to keep house price developments on the stabilisation track. In particular, the high levels of tax incentives are perceived as key drivers of housing prices. Accordingly, limiting the strong drive from tax incentives is perceived as a more effective measure in curbing demand than, for instance, macroprudential ones (see for instance Kuttner and Shim (2013)). Rebalancing the current low level of property taxes could also help to stabilise demand (see for instance Johannesson-Linden and Gayer (2012)), while lower transaction taxes could help to have more transactions and a better use of the existing stock.

Addressing the supply side constraints, including the most pressing rental market inefficiencies and supporting new constructions (e.g. by supporting the building of housing for special sub-segments, notably students dwellings, affordable houses, rental houses) could further smoothen house price dynamics and could ease imminent economic and social constraints on a shorter term.

A wide debate is currently taking place in Sweden involving all major stakeholders in this area to analyse the main challenges of the housing market. The analyses and the debate resulting from these wide discussions could pave the way for further initiatives to tackle these issues on the housing market.

4.1.      Banking sector overview

Sweden has a large banking sector whose total assets amount to about 400% of GDP. The banking market is dominated by four large groups, Nordea, Svenska Handelsbanken, SEB and Swedbank, which together have about 70% market share in terms of loans or deposits. The level of foreign ownership and public ownership in the banking sector is relatively low. On the asset side, high and increasing exposure to households, mainly through mortgage loans, is the key credit risk of Swedish banks. It mirrors the problem of high household indebtedness analysed in chapter 3.2. On the liability side, the banks are exposed to funding risks as they finance their business predominantly on the international wholesale funding markets, creating significant maturity and currency mismatches in their balance sheets. Analysis of these key risks constitutes the focus of this chapter.

The Swedish banking groups, together with Danske Bank and the Norwegian DNB, dominate the financial sector in the Nordic-Baltic region. Total foreign exposure of Swedish banks amounts to 158% of GDP (and about half of their total lending) and is mainly directed towards Denmark, Finland, US, Germany and UK. Exposure to troubled euro area economies is marginal. Exposure to the Baltic countries, which corresponds to about 10% of GDP, plays a relatively minor role in Swedish banks’ balance sheets (accounting for around 5% of the loan book). At the same time, these banks constitute main pillars of the Estonian, Latvian and Lithuanian financial sectors ([62]). Hence, soundness of Swedish banks is fundamental to financial stability in the whole region.

Loans constitute about 55% of banking sector assets. Traditionally, Swedish banks were financing domestic enterprises, facilitating expansion of the export-oriented industry. Currently, corporate loans make up 17% of the balance sheet, down from 22% ten years ago. In recent years, loans to households have been gaining share relative to loans to non-financial corporations on the back of higher growth rates (Graph 4.1). Mortgage loans amount to 21% of the balance sheet, compared to 18% in 2004. This change in balance is linked with the growing household indebtedness and increasing house prices (see Chapter 3).

Deposits account for only 30% of the banking sector liabilities. Due to their large size relative to the Swedish economy – and the local savings – Swedish banks are bound to rely extensively on market funding, for the most part foreign. This also results from the traditional saving patterns in Sweden, with a low share of savings held as bank deposits and a high share placed in investment funds, pension funds and insurance products. Since market financing tends to be more volatile, the funding risk is one of the focuses of this chapter.

The capital adequacy of banks measured by standard regulatory ratios is high, with an average Tier 1 ratio ([63]) at 11.2% (Table 4.1). However, the leverage ratio, which shows the relation of capital to total (non-risk weighted) assets, is relatively low. This is linked with the issue of risk weights applied by Swedish banks, which is examined below in the context of credit risk analysis.

Swedish banks feature good profitability that they were able to maintain despite the crisis. The average return on equity (RoE) amounts to 11.3% and the return on assets (RoA) to 0.5% (Table 4.1), putting the Swedish banks on top of EU peers (Graph 4.2). The high average asset quality, testified by the low non-performing loan ratio, is the key factor behind banks’ profitability. It is also supported by high cost-efficiency. The average cost-to-income ratio of the sector is below 60%.

The financial supervisor (FSA) is responsible for macro-prudential policy. Following internal debate on the organisation of macro-prudential supervision, the government decided in August 2013 to grant responsibility for macro-prudential oversight to the FSA, thus concentrating micro- and macro-prudential tools in a single authority. However, the Riksbank and the Ministry of Finance will remain involved in the areas of their competencies, not least through participation in the Financial Stability Committee established by the government decision in December 2013 ([64]).

4.2.      Credit risk

Lending rates are positive but much below the pre-crisis levels. In 2009-2010, banks deleveraged temporarily the Swedish corporate sector, while housing loans were decelerating. The year-on-year growth of credit to corporations as of October 2013 amounted to 1.3% and to households’ to 4.9%, in which mortgage loans grew by 5.3% (Graph 4.1) ([65]).

The crisis dented banks’ asset quality. Non-performing loan (NPL) ratios peaked in 2009 but have declined since (Graph 4.3), except for the Nordea group due to its losses on the Danish exposures and the shipping portfolio. By September 2013, NPLs were decreasing in all banks. The major banks maintained various levels of loan loss provisions to cover impaired assets. The coverage ratios ranged from 40% to 130%.

The high and increasing exposure to the household sector is the main credit risk of Swedish banks. While the financial soundness indicators (Table 4.1) present a sound picture of the banking system, they are by their nature backward looking and do not fully capture the inherent risks. A slump in households’ disposable income or an external shock could translate in deteriorating asset quality of the Swedish banks (see Chapter 3 for more detailed analysis). The possible overpricing in the Swedish property market poses a risk to banks’ balance sheets through a shrinking value of collateral of mortgage-backed loans. This would put some borrowers in negative equity ([66]), which for banks could increase losses on foreclosures and create funding problems, in particular linked to the use of covered bonds ([67]).

The banks are required to hold capital buffers to cover unexpected losses resulting from a tail scenario. At the request of the FSA banks have been increasing their capital adequacy ratios. The Tier 1 ratios of the major banks increased from about 6% to 8% in 2007 to 10% to 12% in 2013 (Graph 4.4). However, the share of Tier 1 capital in the total balance sheet of banks (leverage ratio) remained stable or slightly declined in the medium term (Graph 4.5). This means that capital held by banks grew at similar pace as their balance sheets. Further analysis reveals that the increase of capital adequacy ratios resulted from the decrease of risk-weighted assets (RWA) relative to total assets of banks, called RWA density ([68]).

The shrinking RWA density is a specific concern for Swedish banks. It has resulted from the growing share of mortgage exposures, which are characterised by generally low risk weights, in total assets. On the other hand, the particularly low risk weights applied by Swedish banks in their internal risk assessment models (about 6% on average), justified by long historical data series without defaults, dating back to mid-1990s, magnified the impact. These factors prompted the Swedish FSA to take corresponding measures.

Sweden has advanced the full implementation of Basel III capital requirements. Ahead of the 2019 universal deadline, Swedish banks are required to hold 10% Core Equity Tier 1 ratio (CET1) as of 2013 and 12% CET1 as of 2015.

In order to address the concerns about the capital adequacy calculation, the FSA introduced a 15% floor on risk weights on mortgage exposures in May 2013. The floor was introduced in the framework of the supervisory review (the Pillar 2 process) following analyses and discussions with various stakeholders. It translates into increased capital requirements on top of the regulatory minimum. The add-on ranges from 0.3% Common Equity Tier 1 ratio for Nordea and 0.5% for SEB to 1.4% for Handelsbanken and 2.2% for Swedbank ([69]). In 2013, banks' actual capital adequacy ratios were above the Pillar 2 requirements including the add-ons.

In November 2013, the Riksbank recommended that the risk weight floor is raised further to 25%. The FSA may turn it into a supervisory requirement once the CRD4 ([70]) is implemented in Sweden, which is foreseen for mid-2014. The increase of the risk weight floor is communicated as an alternative to higher counter-cyclical capital buffers. The central bank has also recommended that banks report quarterly their leverage ratios in accordance with CRR ([71]) definition.

Some other initiatives have been launched to contain risks stemming from private sector indebtedness (see Chapter 3.2). In 2010, a 85% loan-to-value (LTV) ratio was imposed on mortgage lending in Sweden in order to encourage more cautious credit risk assessment by banks and responsible borrowing by consumers. Since 2010, the Swedish Bankers' Association recommends that loans with a LTV above 75% are amortised to this level. More recently, the FSA recommended that banks present to their customers individual amortisation plans (IAP) to promote the long-term benefits of amortisation.

The effects of the recent FSA initiatives have to be monitored. The impact of the new risk-weight floors on mortgage exposures on the banks' capital adequacy ratios and RWA density should be immediate but the long term effect of limiting mortgage credit growth remains to be seen. So far, the effect of the floors on loan pricing has been limited due to the competition in the market.  The undesired effects of the floors might include an increase of risk-appetite on the part of lenders and down-ward revisions of risk weights on corporate exposures in banks internal models ([72]). The FSA plans to monitor the effectiveness of the recommendation through its impact on the actual amortisation levels and take further measures if necessary.

4.3.      Funding risk

The Swedish banking sector is characterised by a high share of market funding in its liabilities, complementing the deposit funding. Customer deposits constitute about 40% of total funding (excluding equity and derivatives). For this reason, Sweden has one of the highest loan-to-deposit ratios in the EU (Graph 4.6). It demonstrates the high degree of the banking sector leverage and its potential vulnerability to market shocks. The remaining 60% of funds are raised on the wholesale financial market. The wholesale market funding of Swedish banks is also linked with currency risks as banks issue the bulk of their debt abroad. About 60% of the major banks market funding is denominated in foreign currencies ([73]), mainly US dollars and euros.

The average maturity of market funding instruments held by the banks was increasing in recent years. Market funding is split between short and long term as per maturities of the financial instruments. Their shares were quasi-equal in the past few years, but the share of long term funding increased significantly in 2013 (Graph 4.7). This trend is confirmed by analysis of maturity structure of the large banks' funding in 2012 and 2013. The share of instruments with maturities of 4 to 6 years and above 7 years increased in their total funding, while the share of instruments with maturities below 3 months dropped (Graph 4.8). This is likely a result of both the Swedish authorities' actions and the favourable market conditions for issuing long-term debt in the Nordic countries, which were regarded by international investors as a safe haven during the sovereign debt crisis in the euro area.

Swedish banks benefit from low funding costs linked with low risk premia charged by investors. It is related to high ratings and low CDS spreads of the main Swedish banks, supported by the triple-A rating and low CDS spreads of the sovereign. The low funding costs contribute to the banks' high profitability. In 2013, yields on the banks' long term bonds rose in line with global market trends. Improving market conditions in the euro area, in particular a better risk assessment of banks in the distressed countries following the 2014 asset quality review and stress test conducted under the Single Supervisory Mechanism, may in the future diminish the "safe haven" advantage enjoyed so far by the Swedish banks.

 

In 2011, the Riksbank recommended that major banks should reduce their structural liquidity risks and approach the minimum level of 100 per cent in the Net Stable Funding Ratio (NSFR) ([74]). Banks have started to adapt their funding structure in the wake of the 2008 financial crisis extending the average maturity of their funding instruments. From 2013, the FSA requires banks to abide by the minimum Liquidity Coverage Ratio (LCR) ([75]). As per September, banks complied with the requirement, reporting LCR between 114% and 147%. However, there were concerns that the average LCR ratios mask insufficient LCRs for Swedish kronas ([76]). Pending the final definition of the NSFR, the Riksbank reports its own proxy structural liquidity measure, according to which banks on average are below the 100% minimum, despite progress from 74% in 2008 to 92% in 2013.  In November 2013, the Riksbank recommended quarterly reporting of NSFR and LCR in Swedish kronas ([77]). The FSA considers making the NSFR binding once its final definition is set at the EU level.

In 2012, the Riksbank decided a 30% increase of its foreign exchange reserves. An amount of SEK 100 billion was added to the existing pool of SEK 314 billion. The motivation to increase the reserves was inter alia to provide a backstop for banks' growing demand for liquidity in foreign currencies. The amount was estimated on the basis of liquidity stress tests carried out for the four major banks and was funded by the National Debt Office. The strengthening of the Riksbank's foreign reserves complements its recommendations aimed at reducing the liquidity risks. Currently, the government in agreement with the Riksbank and the FSA envisages transferring the costs for maintaining the additional reserves to banks in proportion to the foreign currency exposures in their liabilities. The objective is to incentivise banks to reduce their currency risk.

4.4.      Conclusions

The level of credit risk and funding risk in the Swedish banking sector remains stable. The funding risk is linked to the persistently high share of market and foreign currency funding and the credit risk manifests itself in the increasing private sector indebtedness. These risks are inter-related with a potential to reinforce each other in case some tail scenarios materialised. Problems in the Swedish housing market might trigger risk aversion of foreign investors and lead to problems in bank funding. On the other hand, funding problems might be transformed by banks into limited credit availability, which in turn might suppress economic growth with a negative impact on the housing market.

The Swedish authorities have been taking appropriate actions to address the key risks relating to financial stability. These measures include the high capital adequacy requirements, with due attention paid to the impact of risk weighted assets density, the 85% cap on loan-to-value ratios, the recommendation for individual amortisation plans, the advanced implementation of the Basel III liquidity ratios and the increase in foreign exchange reserves, with the plan to make banks share the costs. Whereas some of these measures have already proved to be effective (the LTV cap), others (concerning risk weights, amortisation and liquidity) still have to yield the expected improvements.

Developments in the financial sector and the effectiveness of the applied measures call for continued monitoring. A decrease in the credit risk levels would be visible in a lower private indebtedness level, lower numbers of new amortisation-free loans and longer average maturity of mortgages as well as adequate capital buffers held by banks. For funding risks, the progress may be measured by the evolution of the average maturity of bank liabilities as well as banks' compliance with the regulatory liquidity ratios (LCR and NSFR) and requirements (funding of the currency reserve).

The analysis in this report indicates that macroeconomic developments in the areas of private debt and the housing market are the main challenges in Sweden. The report also analyses some issues linked to the sustainability of the Swedish economy, such as the current account surplus and Sweden’s export market shares (EMS).

It should be recalled that these challenges were identified under the MIP in the first and second IDR and relevant policy responses were reflected and integrated in the country-specific recommendations (CSR) issued for Sweden in July 2013 (CSR 2 and CSR 3). The assessment of progress in the implementation of those recommendations will take place in the context of the assessment of the National Reform Programme and the Convergence Programme under the European Semester. Against this background, this section discusses different avenues that could be further envisaged to address the above challenges.

The challenge of high private debt: Possible policy avenues

Over the last years, Sweden has introduced a number of measures aiming at reducing the high indebtedness of the private sector (including both households and corporations). Nevertheless, the unfavourable situation remains and further reforms would be desirable in order to address the high debt levels which constitute a risk to macroeconomic developments, primarily through potential consumption and investment effects. This risk was reflected in the Council's CSR 2 in July 2013, urging Sweden to continue to address risks related to private debt and to foster prudent lending.

Promoting sound lending practices to households and reviewing taxation issues

Since the last IDR, signals that the amortisation behaviour of households is shifting have strengthened. The LTV-cap at 85% seems to have been effective in limiting households' exposure to mortgage debt, and banks are gradually becoming more consistent in requesting amortisation. The increased risk weight floor to 15% may also have some effect. Notwithstanding these positive shifts in lending practices, additional tools may be needed.

Many households take unsecured loans at higher interest rates to top off the mortgage loan and young people find themselves having to rely on their parents to finance the remaining 15% with their assets as the underlying security. Hence, a look at the global indebtedness of households needs to be taken. The FSA's signals of further risk weight increases and possible amortisation requirements should self-regulation fail could be seen in this perspective. It could also be useful to push for more amortisation than as is currently the case. In addition, one should be observant on the net lending of households. Sound underwriting practices can be further promoted. Banks may also be requested to base lending on real repayment ability rather than on the potentially volatile value of the underlying asset, taking into account actual, and not only presumed, living expenses when assessing the residual income. While a recent government proposal should decrease the charges paid by a consumer in case of earlier repayments of mortgage loans, it remains costly for borrowers to switch between fixed and variable mortgage rates.

Last year's analysis regarding the debt-bias in taxation creating incentives for households to finance their investment in housing through debt rather than own savings remains valid. The Swedish tax system continues to offer generous deductibility of interest payments on mortgages. At the same time, property fees are very low and cannot counterweigh the effect of the tax relief on mortgages. Due to the reduced relative price of debt financing, households often prefer to save in other forms (such as investing in the stock market) rather than repaying their mortgages. This has contributed to an increased leverage of the household sector and could be addressed via a global reform gradually reducing the tax deductibility of interest payments while at the same time strengthening recurrent property taxation. As regards the tax deductibility of interest payments, the government could consider a reduction in the deductible rates or limit the maximum deductible amounts. Reducing the scope of tax deductibility might be also considered: for instance tax deductibility of interest rate payments could be limited to amortising mortgage loans. Although reforming taxation away from the current debt-biasing structure may be sensitive, it seems appropriate to clearly address the link between the current debt-bias and households’ indebtedness at the present moment, and take some concrete steps towards a phased-in implementation of a reformed taxation system.

Although the housing market appears to have regained in stability, reforms in this area obviously need to be carried out gradually and transparently, not to unintentionally feed nervousness in the housing market. It may be opportune to attempt to enshrine the need to address the debt bias in taxation and indebtedness already at the present moment, in order to be prepared for a sequenced roll out.

Addressing corporate indebtedness

Resulting from the analysis in section 3, it seems that although corporate indebtedness remains high in Sweden, the characteristics of the indebtedness do not appear to give rise to macroeconomic risks. When assessing corporate indebtedness in Sweden, it is relevant to focus on consolidated debt (or even super-consolidated as described in section 3.2.1). This is because a combination of generous tax deductibility of interest payments and a historically relatively high corporate income tax in Sweden has inspired advanced tax planning by multinational companies and foreign owners of Swedish companies, driving up total indebtedness in particular in the form of intra-group loans from abroad. However, even when considering only consolidated figures, the tax deductibility has traditionally made it more beneficial also for Swedish, local companies to finance their investments with debt rather than equity.

Swedish authorities have taken some steps to counteract the tax minimisation practices of multinational companies. The corporate tax, reduced from 26.3% to 22% in January 2013, is now more in line with the EU average (ca. 23% in 2013). Moreover, tax deductibility of intragroup interest rate payments was capped. Since the 2013 accounting year, only "sound business reasons" are acceptable grounds to maintain the deductibility under certain, rather strict conditions. However, the lack of precision in the formulation of the new rules may result in many companies being uncertain about the amended "sound business reasons" test and its applicability. Court proceedings in relation to this matter cannot be excluded.

The so-called investor's deductibility introduced as from December 2013, is not expected to substantially shift the balance away from debt-financing on the aggregate level. Nevertheless, the measure, which gives individuals acquiring shares in a new or expanding SME the possibility to deduct half of the amount of the purchase up to SEK 650,000 per person and year, can be of importance for start-ups and innovative SMEs.

Within the debt segment, the observed corporate bond financing trend calls for continued monitoring. Having sprung from the contraction in traditional bank lending during the financial crisis, it has given rise to structural changes affecting the debt portfolios of corporations. While naturally not shifting the debt/equity balance, bond financing diversifies the funding structure of corporations and decreases their reliance on traditional bank loans, with banks having to compete with bond financing on larger deals.

Summing up, it remains hence to be seen whether the above-mentioned changes will eventually have a decisive effect on corporate debt. In any case, the on-going governmental inquiry on corporate taxation due in June 2014 will certainly address the issues of debt/equity neutrality and other aspects of the Swedish corporate tax system. In this context, further ways to reduce the debt-bias in company taxation could be considered, such as a generally applicable cap on the tax relief for interest payments.

The Housing market challenge: Possible policy avenues

The structural inefficiencies linked to under-supply of dwellings have been built up over the course of decades. They have macroeconomic impacts mainly via increased household indebtedness due also to a lack of available rental apartments. Naturally, the unrelenting demand, which is not met by housing supply, has created a strong upwards pressure on property prices. House price developments have gained momentum over the course of the last year and started to climb again. The development of apartment prices has been even more pronounced, in the wake of a strong urbanisation and immigration trend. Particular pressure is being put on Stockholm County to accommodate the new arrivals.

The macroeconomic risks associated with housing market inefficiencies were reflected in the Council's country-specific recommendation (CSR) 3 in July 2013, asking for continued reforms of the rent setting system, promotion of competition in the construction sector and improvements to the planning, zoning and appeal processes. In spite of some recent reforms in this area, a more overarching perspective is desirable when addressing the housing market shortcomings and policy avenues along the lines presented below could be appropriate.

Rental market

As demonstrated in last year's in-depth review, an important factor underlying the inefficient housing market in Sweden is the regulated rental system which limits supply and pushes house prices upwards as consumers are left with limited options. During many decades, rents have been set according to the so-called utility value system, which does not reflect households' real willingness to pay and effectively subsidises the rent for rental units in attractive city locations, limiting the profitability of developing the supply. Most municipalities located in urban areas face a shortage of available rental apartments and administer queuing systems for allotting tenancies. But the number of rental units is decreasing rather than meeting demand, especially in the greater Stockholm area. A gradual move towards market-clearing rents would enhance the functioning of the rental market and create incentives for developers to invest.

The rental market has seen some limited reform over the latest years. Restrictions on private letting of tenant-owned apartments have been eased since February 2013 and this has to some extent contributed to increasing supply. Such rental contracts may be signed on more market-based terms, but are still subject to the agreement of the housing cooperative.

There have been several governmental initiatives suggesting the need to reduce the negative effects of rent regulation, e.g. through a public inquiry explicitly calling for reforms (SOU 2012:88), a parliamentary committee set up in 2013 to address ways in which to promote construction and through a report on the rental market and the efficiency losses of the utility-value based system published by the Swedish National Board of Housing, Building and Planning in November 2013. However, in spite of these initiatives, no decisive legislative proposals have come forward. To allow market forces to establish an optimal supply of rental housing at an adequate price, several additional steps might be taken.

These could include deregulating rent setting in the sense that individual tenants and landlords would be able to agree on mutually acceptable rent levels and hence increase the freedom of contract. A first step could be to allow rents of newly produced rental apartments to fully reflect tenants' willingness to pay. Eventually, rents of the remaining stock could be adapted.

Housing supply

On the supply side, the analysis has also identified several factors that hold back property development in a market basically controlled by a limited number of construction and property development companies. Protracted and opaque zoning and planning processes at the municipal level demotivate all but the biggest and most well-established companies from engaging in negotiations at the local level. The practice of municipalities to pose additional, disparate technical requirements in addition to the rules applicable on the national level constitutes effective barriers to entry for smaller and/or foreign companies.

Municipalities currently lack sufficient incentives to allow construction, as new housing is linked to infrastructure and child care investments that are primarily to be borne by the incumbents. They represent a local planning monopoly and face no direct consequences when projects are delayed or fail to materialise. Construction companies also hesitate to invest in rental units due to the reasons outlined above. They may also choose to wait to develop land in their possession and for which they already have building permits, in the hope that prices of tenant-owned apartments will continue to climb. The land may in fact be more valuable as an option for future developments. Finding clear incentives for municipalities and construction companies seems therefore to be a priority.

The Swedish authorities have taken several positive steps to address these inefficiencies, for instance through a recent governmental proposal in view of facilitating construction of housing for students and young people through changes to the Planning and Building Act, through a governmental initiative to ease the requirements of the standard procedures proceeding building permit approvals, through a review of the rules governing noise levels with a view to facilitate construction as well as through a governmental proposal aiming at limiting the possibility for municipalities to pose additional, technical requirements going beyond national rules, etc. The additional funds allotted in the 2013 budget for speeding up appeals' procedures also seems to have given some effect at the level of the county administrative boards.

Nevertheless, these initiatives do not seem sufficient to address the current challenges as the fundamental issues linked to the rent-setting system as well as to the construction barriers have not fully been challenged by them. It would seem that a more strategic, overarching approach to solve the inefficiencies weighing on the housing market would be useful, in addition to individual measures targeting specific but limited areas thereof. On the detailed level, further streamlining of the planning and zoning processes could be considered. In particular, these processes could be rendered more efficient by more standardised building requirements across the country and by decreasing the extensive zoning and planning requirements. More transparent land allotment procedures at municipal level could be another priority as well as facilitating the access to tenders for smaller or foreign developers to increase competition. Identifying clear mechanisms to incentivise developers to construct on land which has already been subject to detailed planning could also be considered. In a similar vein, the incentives for municipalities to support new constructions could also be re-assessed.

The measures highlighted in this In-Depth Review are interlinked and reinforce each other.  Furthermore, the timeframe of their impact could also differ substantially. Avoiding abrupt policy changes in these areas is crucial due to their pivotal macroeconomic impact: gradual implementation, well-considered timing, wide political and public support and continuous evaluation of the impacts would be necessary. These imbalances have been built up over a longer time horizon, thus their unwinding cannot take place overnight. Nevertheless, a forthcoming stronger economic growth period and the ongoing wide debate in Sweden on these issues could pave the way for further sound policy actions in these areas. 

Andrews, D., Caldera-Sánchez, A. and Johansson, Å., Housing Markets and Structural Policies in the OECD, OECD Economics Department Working Papers No 836, 2011.

Blomberg, G. and Falk, M., How do large current account surpluses co-exist with a weak international investment position?, Riksbank Economic Review 2006:1.

Börestam, A. and Schmiedel, H., Interchange Fees in Card Payments, ECB Occasional Paper Series No 131, September 2011.

Caldera Sánchez, A. and Johansson, Å., The price responsiveness of housing supply in OECD countries, OECD Economics Department Working Paper No 837, 2011.

County Administrative Board of Stockholm, (Länsstyrelsen i Stockholms län), Läget i länet: bostadsmarknaden i Stockholms län 2013 (Stockholm County Housing Market 2013), Report 2013:14, 2013.

Cuerpo Caballero, C., Demertzis, M., Fernández Vilaseca, L., Pontuch, P., Assessing the dynamics of house prices in the euro area, Quarterly Report of the Euro Area, 4/2012.

Cuerpo Caballero, C., Drumond, I., Lendvai, J., Pontuch, P., Raciborski R., Indebtedness: Deleveraging dynamics and macroeconomic adjustment, European Economy - Economic Papers 477, 2013.

European Commission, Current account surpluses in the EU, European Economy 8/2012.

European Commission, Developing an indicator of innovation output, Staff Working Document, No 325, 13.09.2013a).

European Commission, Education and training monitor 2013, Vol. 1, 2013b).

European Commission, European Economic Forecast, Autumn 2013c).

European Commission, European Economic Forecast, Winter 2014.

European Commission, Innovation Union Scoreboard, 2013d).

Evidens, Om hushållens skuldsättning och bostadsmarknaden (On household debt and the housing market), Market research, December 2013.

Hüfner, F. and Lundsgaard, J., The Swedish Housing Market – Better Allocation Via Less Regulation, Economics Department Working Paper, No 558, OECD, 2007.

International Monetary Fund, Balance of Payments Manual, 1993.

International Monetary Fund, Sweden–Staff Report for the 2013 Article IV consultation, 2013.

Jahnson, D. and Lundberg, J., Fördelningseffekter av utbudsrestriktioner på bostadsmarknaden, (Distribution effects of supply restrictions in the housing market), The Reform Institute, September 2013.

Johannesson-Linden, Å. and Gayer, C., Possible reforms of real estate taxation: Criteria for successful policies, European Economy - Occasional Papers 119, October 2012.

Kuttner, K and Shim, I., Can non-Interest Rate Policies Stabilise Housing Markets/ Evidence from a Panel of 57 Economies, BIS Working Papers No 433, November 2013.

Lind, H., Social housing in Sweden, Social Housing in Europe, London School of Economics, forthcoming.

Lind, H., What is Happening on the Housing Market, SNS Analys No 10, March 2013b).

Mattsson-Linnala, S., Lind, H., Weiss, L., Bengtsson H., Så ökar vi bostadsbyggandet (The way to increase housing construction), Premiss Förlag 2013.

Ministry of Social Affairs, En effektivare plan- och bygglovsprocess (A more effective planning and building permit process), Government Official Reports (SOU) 2013:34, 7 May 2013.

Ministry of Social Affairs, Nu blir det enklare att bygga (Now it will be easier to build), Debate article of 13 February 2014.

Müge Adalet McGowan, Housing, Financial and Capital Taxation Policies to Ensure Robust Growth in Sweden. OECD Economics Department Working Papers No 1024, 2013.

National Institute of Economic Research (Konjunkturinstitutet), The Swedish Economy, December 2013.

Peterson Institute for International Economics, Estimates of Fundamental Equilibrium Exchange Rates, Policy Brief No 13-29, November 2013.

Statens Offentliga Utredningar (Swedish Government Official Reports), Ökat bostadsbyggande och samordnade miljökrav – genom enhetliga och förutsägbara byggregler (Increased housing and coordinated environmental requirements - through consistent and predictable building regulations), 2012:86.

Statistics Sweden, (Statistiska centralbyrån), Sharp decrease in conversion of rental dwellings, Nr 2013:137, 30 May 2013.

Statistics Sweden, (Statistiska centralbyrån), Financial Market Statistics December 2013, 2014a.

Statistics Sweden, (Statistiska centralbyrån), Real estate prices up to and including December 2013: Increasing house prices, Nr 2014:6, 13 January 2014b.

Statistics Sweden, (Statistiska centralbyrån), Large increase in dwelling starts in new constructions, Nr 2014:39, 20 February 2014c.

Stockholms stads bostadsförmedling (Stockholm Housing Service), Bostadsförmedling i Stockholm AB: Köstatistik för bostadskön (Housing Service in Stockholm: queue statistics), December 2013.

Sveriges Riksbank, Financial Stability Report 2013:2, 2013a).

Sveriges Riksbank, The Swedish retail-payment market, Riksbank Studies, June 2013.

Sveriges Riksbank, The Riksbank's inquiry into the risks in the Swedish housing market, 2011.

Swedbank, Bostadsbristen kostar 21 miljarder per år. (Housing shortage costs SEK 21 billion a year), Makrofokus Sverige, Makroanalys, 11 November 2013.

Swedish Competition Authority (Konkurrensverket): Competition in Sweden (Konkurrensen i Sverige) 2013. Rapport 2013:10.

Swedish Council for Cooperation on Macroprudential Policy, Minutes, 1 October 2013.

Swedish Financial Supervisory Authority (Finansinspektionen), How FI can decrease the risks inherent in household debt, Ref. 13-12811, 14 November 2013a).

Swedish Financial Supervisory Authority (Finansinspektionen), Risk weight floor for Swedish mortgages, Ref. 12-11920, 21 May 2013b).

Swedish Financial Supervisory Authority (Finansinspektionen), The Swedish Mortgage Market 2013, Ref. 13-2825, 7 March 2013c).

Swedish Financial Supervisory Authority (Finansinspektionen), Risks to financial stability and individual households in Sweden associated with household indebtedness and amortisation behaviour, Memorandum (Author: Robert Boije), Ref. 13-11430, 1 October 2013d).

Swedish Ministry of Justice, Överskuldsättning i kreditsamhället? (Over-indebtedness in the credit society), Government Official Reports (SOU) 2013:78, 26 November 2013.

Swedish National Board of Housing, Building and Planning (Boverket), Bostadsbristen ur ett marknadsperspektiv (The housing crisis from a marketing perspective), Rapport 2012:18, 2012.

Swedish National Board of Housing, Building and Planning (Boverket): Are house prices driven by a housing shortage? Market report, February 2013a).

Swedish National Board of Housing, Building and Planning (Boverket): Bostadsbristen och hyressättningssystemet - ett kunskapsunderlag (The housing shortage and the rent setting system -a knowledge base), Market report, November 2013b).

Swedish National Board of Housing, Building and Planning (Boverket): Vilken typ av bostäder är det brist på? (What type of housing is lacking?), 2013c).

Swedish Property Federation (Fastighetsägarna), Swedish Union of Tenants (Hyresgästföreningen) and Swedish Association of Public Housing (SABO), Conditions for economic equilibrium – a tax reform for rented properties, 2013.

Swedish National Board of Housing, Building and Planning (Boverket), Bostadsbristen ur ett marknadsperspektiv (The housing crisis from a market perspective), Rapport 2012:18.

Transparency International, Corruption Perceptions Index 2013, 2013.

World Bank, Doing Business 2014, 2013.

World Economic Forum, Global Competitiveness Report 2013-14, 2013.

([1])  ECFIN Country Focus, Sun spots on the Swedish labour market? Pavlína Žáková, May 2013

([2])  European Commission (2014).

([3])  Servicification refers to the increased use of services in productive processes, as well as the increased (cross-) selling of services to customers.

([4])  According to Eurostat data, from 2000 to 2011 immigration flows into Sweden have been, on average, twice as large as emigration outflows.

([5])  Defined as gross fixed capital formation divided by GDP (current price data).

([6])  See Peterson Institute for International Economics (2013) and International Monetary Fund (2013).

([7])  World Economic Forum (2013).

([8])  World Bank (2013).

([9])  Transparency International (2013).

([10]) European Commission (2013a).

([11]) 2011 data

([12]) European Commission (2013b).

([13]) European Commission (2013). [Education and training monitor 2013 - Volume 1]

([14]) The symmetric revealed comparative advantage indicator shown in Charts 3.10 and 3.11 is an indicator of specialisation of a country's exports relative to the world. The indicator ranges from -1 to +1. Values greater than zero imply specialisation of the country in the corresponding sector.

([15]) See European Commission (2012), Table 4.2.

([16]) European Commission (2014).

([17]) Statistics Sweden data (financial accounts by counterpart sector).

([18]) Part of the domestic stock market capitalisation is "owned" by foreign investors, and therefore counted as an external liability for accounting purposes. Additionally, domestic bond prices co-move with domestic interest rates.

([19]) As measured, e.g., by the Euronext index.

([20]) See Blomberg, G. and Falk, M (2006).

([21]) Consolidated data are more appropriate than non-consolidated in assessing the debt burden of Swedish corporations due to a wide use of domestic intra-group lending.

([22]) For the concept of notional leverage, see Cuerpo et al. (2013).

([23]) These tax minimisation practices are further detailed in last year's Review, p47.

([24]) See Cuerpo et al. (2013) for more details on the methodology to filter assets from their valuation effects.

([25]) FSA, Memorandum, 21/05/2013

([26]) FSA, Swedish Mortgage Market 2013

([27]) In practice loans with interest rate adjustments every third month.

([28]) Sveriges Riksbank (2011), The Riksbank's inquiry into the risks in the Swedish housing market

([29]) FSA, The Swedish Mortgage Market 2013

([30]) Discretionary income is the income net of housing costs and subsistence costs.

([31]) FSA memorandum 2013-11-14.

([32]) FSA Memorandum 2013-10-01, FI Dnr 13-11430

([33]) Prop 2013/14:44 of 7 November 2013.

([34]) Swedish Council for Cooperation on Macroprudential Policy: Minutes, 1 October 2013.

([35]) FSA, Memorandum 2013-11-14.

([36]) The methodology to construct the indicators of overall loan supply pressure and overall loan demand pressure is explained in Cuerpo et al. (2013).

([37]) The financial soundness indicators are further elaborated in chapter 4.

([38]) Only Norway and New Zealand  reached higher real house price growth rates from the OECD countries in the same period.

([39]) The AMR scoreboard is available at: http://ec.europa.eu/economy_finance/indicators/economic_reforms/eip/sbh/

([40]) Mean-reversion properties are not confirmed by empirical evidence in several countries.

([41]) Nominal price increase is of little information given the general increase in prices and income. So-called real series compare house prices with the consumer price index (CPI), but it is still a limited measure. First, the CPI is not the best price index to compare to house price inflation: monetary policy keeps overall inflation low and stable, and hence relatively independent from the long and deep housing cycle. Second, nominal prices are not corrected for quality increases, which can be substantial in this sector. In short, even if the price to income ratio also misses the corrections for quality, it is probably one of the best indicators to assess house price level developments.

([42]) A Vector Error Correction Model has been estimated for Sweden, using a system of five fundamental variables; the relative house price, total population, real housing investment, real disposable income per capita and real long-term interest rate. For further details see Cuerpo C., M. Demertzis, L. Fernández, P. Pontuch (2012).

([43]) Tenant-owned apartments ("bostadsrätter") are apartments owned by an association, in which the respective residents own a share giving them the right to reside in a particular apartment. This means that the inhabitants do not own a particular apartment and decisions regarding more substantial renovations and sub-letting of particular apartments have to be taken at collective level.

([44]) Rental dwellings are mainly owned by private landlords and public housing companies and the distribution is fairly even between them. A very small share is also owned by the Swedish state, counties and directly by municipalities.

([45]) The average property price for a one- or two-dwelling building during the fourth quarter of 2013 was nearly SEK 2.2 million in Sweden, SEK 2.8 million in Greater Malmö and almost 4 million in Greater Stockholm (Statistics Sweden, 2014a).

([46]) According to a paper of Sweden's National Board of Housing, increasing incomes and population/dwelling ratio can account for more than half of the total increase in house prices between 1996 and 2011. For more details, see Swedish National Board of Housing, Building and Planning, 2013a.

([47]) In contrast, the third major metropolitan centre of Sweden, Malmö did not experience the same level of house price increases in the last decade. During the early 2000s, many Danes chose to move to the Malmö region, as the newly constructed Öresund bridge enabled easier commuting to Copenhagen, and benefitted from the lower level of house prices in Malmö compared to Copenhagen. However, Danish house prices have been falling sharply since 2007, reaching 2003 levels, and the migration flow has reversed putting downward pressure on house prices in the Malmö region (Swedish National Board of Housing, Building and Planning, 2013a).

([48]) Since many years, almost the entire population growth in Sweden, amounting to roughly 70,000 persons per year, takes place in metropolitan areas and university cities. The population of the Stockholm region has increased by more than 20% in the last 10 years, due to favourable demographic factors, internal migration and a growing number of refugees.

([49]) Riksbanks' repo rates stayed at 4.75% in September 2008, and stands at 0.75% in January 2014.

([50]) The repo rate has been decreased to a record low of 0.75% on 17/12/2013 (and the Riksbank's timing of initial tightening has been postponed until 15-1Q).

([51]) The property tax was lowered from 1.2% of the cadastral value (the cadastral value amounts to 75% of the market value) to the lower of either 0.75% of the cadastral value or SEK 6,512 (or roughly EUR 700, a very low ceiling that would apply to the vast majority of houses), which drastically reduced the taxation of housing.

([52]) Currently more than 430.000 people are waiting for a rental apartment in Stockholm and the average waiting time could be more than 10 years (Bostadsförmedling I Stockholm AB, 2013).

([53]) According to OECD estimations, Sweden has built up an investment gap in the construction of dwellings of SEK 1,500 billion compared to OECD average (Hüfner and Lundsgaard, 2007).

([54]) On average, 42,000 dwellings were built annually from 1980 to 1990, while this figure more than halved in the next decades (based on Statistics Sweden figures).

([55]) A summary of the government initiatives are available at: http://www.regeringen.se/sb/d/14867.

([56]) Most of the growth can be attributed to the Stockholm region where, for instance, dwelling starts in newly constructed buildings increased by 70% in 2013 compared to 2012. By contrast, construction activity in the Goteborg stayed flat (+1% growth) and decreased in the Malmo region (-22%), while the rest of Sweden outside the three main metropolitan areas experienced 18% growth regarding new dwellings compared to the same period of 2012. Regarding dwellings types, the growth has been driven by multi-dwelling buildings with an increase of 36% in 2013 compared to the similar period in 2012.

([57]) For further details on the methodology considering the degree of flexibility in setting rental levels for new contracts and the definition of the rental updating methods within tenures, see Cuerpo C., M. Demertzis, L. Fernández, P. Pontuch (2012).

([58]) The model developed in the paper assumes that the difference between rental prices and market prices in Stockholm could be at the magnitude of 68% and of 25% in Gothenburg, while in the rest of the country this figure stays within the magnitude of 5%.

([59]) The number of rental dwellings that were converted to tenant-owned dwellings declined in 2012 (4.216 dwellings) compared to 2011 (7.100) and 2010 (about 20.000). Since 2000, approx. 160.000 rental dwellings have been converted to tenant-owned dwellings in the country. Of these, 72% or about 115.000 are in Stockholm (see Statistics Sweden (2013a)).

([60]) Private lettors and tenants may agree on rent levels based on the actual costs borne by the lettor, and as a general guidance for rent-setting in the future a 4% return on the market value plus the monthly fees will be applicable.

([61]) In December 2013, loans to households increased at an annual rate of 4.9% primarily driven by mortgage loans to households (Statistics Sweden, 2014b).

([62]) Following introduction of the Single Supervisory Mechanism in the euro area (November 2014), the ECB will become a "host" supervisor for the Finnish, Estonian and Latvian subsidiaries of the Swedish banking groups. Formally, the ECB will be subject to the coordination powers of the Swedish lead supervisor. However, it may become a relatively stronger counterpart in the colleges of supervisors than its national “host” predecessors.

([63]) Tier 1 regulatory capital to risk weighted assets.

([64]) The FSC will also include the FSA and the Swedish National Debt Office. The first of its official semi-annual meetings is foreseen for June 2014. The establishment of the FSC formalises inter-institutional cooperation on financial stability and crisis management dating back to 2005.

([65]) ECB consolidated bank balance sheets in local currency.

([66]) Situation whereby the value of a loan exceeds the value of the house.

([67]) The declining collateral value would on the one hand force banks to replenish the collateral pools that secure covered bonds with additional good quality loans and on the other hand could result in an increase of covered bond prices and thus the banks' funding costs.

([68]) It is a matter of a denominator effect, given that the Tier 1 ratio equals Tier 1 capital divided by RWA.

([69]) Sveriges Riksbank (2013) Financial Stability Report 2013:2

([70]) Capital Requirements Directive 4 introducing Basel III in the EU

([71]) Capital Requirements Regulation introducing Basel III in the EU

([72]) Similarly to mortgage exposures, banks base their models on historical data showing very low default levels. A number of banks are currently revising their Internal Rating Based models.

([73]) Sveriges Riksbank (2013) Financial Stability Report 2013:2

([74]) The Basel III Net Stable Funding Ratio (NSFR) seeks to calculate the proportion of long-term assets which are funded by long term, stable funding. Detailed features of the NSFR are still being calibrated: an EU consultation was launched in January 2014. The 100% minimum will be binding by 2019.

([75]) The Basel III Liquidity Coverage Ratio (LCR) requires banks to hold liquid assets sufficient to cover cash outflows over a 30 day period. The minimum coverage is 100%. The Basel III / CRD4 foreseen enforcement deadline for the EU Member States is 2015.

([76]) See Riksbank Financial Stability Report 2013:2 for more details.

([77]) On top of its prior recommendation to report LCRs for USD, EUR and all currencies mixed.