Sweden: 2019 Article IV Consultation—Press Release; Staff Report; and Statement by the Executive Director for Sweden
Author:
International Monetary Fund. European Dept.
Search for other papers by International Monetary Fund. European Dept. in
Current site
Google Scholar Close
Publication Date:
26 Mar 2019
eISBN:
9781498305259
Language:
English
Keywords:
ISCR; CR; holding; rate; IMF staff calculation; liability positions vis-à-vis nonresident; inflation prospect; Swedish krona; household mortgage debt growth; core CPIF inflation; headline inflation; Housing prices; Inflation; Housing; Wages; Income; Global; HICP inflation; statistics Sweden
Download PDF (7.9 MB)

Sweden’s solid growth appears to have slowed during 2018 as domestic demand cooled and exports were flat for most of the year. Wage rises remained subdued even as labor markets tightened, contributing to low underlying inflation and krona weakness. A mild decline in housing prices and slowing household lending have eased risks, yet construction is slumping despite high housing price levels, partly due to reduced availability of pre-sale financing for small developers.

Abstract

Sweden’s solid growth appears to have slowed during 2018 as domestic demand cooled and exports were flat for most of the year. Wage rises remained subdued even as labor markets tightened, contributing to low underlying inflation and krona weakness. A mild decline in housing prices and slowing household lending have eased risks, yet construction is slumping despite high housing price levels, partly due to reduced availability of pre-sale financing for small developers.

Recent Developments

Sweden’s solid growth appears to have slowed during 2018 as domestic demand cooled and exports were flat for most of the year. Wage rises remained subdued even as labor markets tightened, contributing to low underlying inflation and krona weakness. A mild decline in housing prices and slowing household lending have eased risks, yet construction is slumping despite high housing price levels, partly due to reduced availability of pre-sale financing for small developers.

1. Sweden’s growth has moderated yet has still generated strong job gains in recent years. Impacted by the euro area double-dip recession, the Swedish economy stagnated in 2011–13, leading to a widening output gap and low inflation. European recovery and domestic monetary easing saw Swedish growth surge to 3.5 percent in 2014–15, with labor productivity rebounding sharply. Growth eased to a still robust 2.4 percent in 2016 to 2018:Q3, driven by domestic demand growing by around 3 percent y/y on average. Job creation accelerated as participation rose from already high levels, while labor productivity growth slowed to about 0.6 percent, a pace similar to that in other advanced economies.

Real GDP Growth

(Percent, y/y annualized, seasonally-adjusted for 2018)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: Statistics Sweden; and IMF staff calculations.

2. But weaker demand and exports are dragging on growth more recently. The third quarter saw a notable decline in domestic demand as both dwelling investment and private consumption fell, with the latter declining even excluding sharp tax-related swings in auto sales. This slowdown in domestic demand is likely partly in response to deteriorating housing market sentiment following the late 2017 housing price decline discussed below. At the same time, exports of goods were flat in 2018—until a sharp jump in Q4—consistent with global factors weakening exports in other European countries, and possibly also reflecting slowing growth in EU markets.

Housing Market Sentiment and Domestic Demand

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: SEB, Haver Analytics, and IMF staff calculations.Note: 1/ Share of households expecting housing price rises increases, net of those expecting decreases, in the next 12 months.

3. Despite strong employment gains, especially among the foreign born, wage growth remains low (Figure 3). Employment growth has averaged nearly 2 percent since 2015, led by 6.4 percent growth for foreign born persons, who accounted for 60 percent of employment gains. Labor force participation (of 15–74 year olds) has increased to above 73 percent, led by a 2 percentage point increase in the participation rate of the foreign born to reach levels close to those born in Sweden. At the same time, unemployment has fallen to a post-crisis low of 6.2 percent in 2018:Q4 and job vacancy rates have risen to the highest level this century.1 Despite this tight labor market, wage rises have only edged up to 2.6 percent y/y in 2018, from a subdued 2.4 percent y/y in 2015–17, moving similarly to German wage growth since late 2012.

Figure 1.
Figure 1.

Sweden: Macroeconomic Indicators

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Figure 2.
Figure 2.

Sweden: Inflation and Monetary Policy

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Figure 3.
Figure 3.

Sweden: Labor Market Developments

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Unemployment and Vacancy Rates

(LHS, percent of labor force; RHS, percent of occupied jobs plus vacancies)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

1/ With full-time students removed from labor force and ranks of unemployed.Sources: Eurostat; Statistics Sweden; and IMF staff calculations.

Wage Growth in Sweden and Germany

(Percent change y/y; 6-month moving average)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: Haver; and IMF staff calculations.

4. Underlying inflation remained broadly flat at below target rates in 2018. Headline inflation measures rose to around 2 percent from mid-2017, with energy prices playing a significant role (Figure 2).2 Core HICP inflation—which excludes energy and unprocessed food—also rose to peak at 2.1 percent y/y in July 2017, only to fall back to 1.3 percent y/y by December 2018. This sharp temporary spike in core HICP inflation mainly reflects methodological changes for travel (Appendix I). An adjusted core HICP that excludes these items has been notably more stable in recent years, at around 1–1¼ percent y/y, and the estimated trend in the core HICP has also been stable at 1¼ percent y/y. The Riksbank has a suite of core CPIF indicators, with the median of these running at 1.6 percent y/y, little changed from early 2017 before the impact of temporary factors.

Core HICP Inflation

(Percent change, y/y)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

1/ Excludes energy and unprocessed food.2/ Also excludes package holidays, and bank/social protection/air transport services.Sources: Eurostat; and IMF staff calculations.

Indicators of Core CPIF Inflation 1/

(Percent change, y/y)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

1/See Riksbank Monetary Policy Report, Oct. 2018, for list of indicators.Sources: Riksbank; and IMF staff calculations.

5. Housing prices went through a mild correction and housing starts have slumped even as housing price levels remain high. From their August 2017 peak, housing prices declined 6 percent in seasonally adjusted terms by end 2017, owing to a surge in apartment completions at the luxury end of the market. This price decline led to a sharp fall in household interest to purchase pre-sale rights to apartments yet to be completed. Price rises slowed to 2 percent y/y in 2018 and household expectations for housing price gains have fallen to levels well below 2013–17 norms. Relatedly, household credit growth eased to 5.3 percent y/y in December from 6.8 percent a year earlier (Figure 4). Housing starts have dropped about 16 percent in 2018—partly reflecting the reduced availability of pre-sale financing to smaller property developers—presaging the 14 percent fall in dwelling investment from 2018:Q1 to 2018:Q4.

Figure 4.
Figure 4.

Sweden: Housing Market Developments

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

House Prices

(Index, 2010=100, SA)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: Haver Analytics; and IMF staff calculations.

Residential Investment and Housing Prices

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: Haver Analytics; and IMF staff calculations.

6. The current account surplus has narrowed, but the krona remains weak. The Swedish current account balance has trended down since 2009 (Figure 8) as investment recovered to over 25 percent of GDP. From 2.8 percent of GDP in 2017, preliminary estimates are that the current account surplus fell to 2.0 percent in 2018, but this decline may be overstated given the large discrepancy between trade balance data from the national accounts and the BOP. Moreover, as factors not captured by the EBA model may also be driving Swedish savings-investment balances, assessments of Sweden’s external position give greater weight to estimates from the EBA REER models and the ULC-based REER. The krona depreciated by 4.3 percent in effective terms in 2018 (Figure 5), possibly reflecting political uncertainties and prospects for slower rate hikes. The EBA REER models find an exchange rate gap of 17–19 percent in 2018, while the ULC-based REER was about 6 percent below historical average in 2018. Overall, staff assesses that the external position remained moderately stronger than the level consistent with medium-term fundamentals and policies in 2018 (Appendix II).

Figure 5.
Figure 5.

Sweden: Select Financial Market Indicators

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Figure 6.
Figure 6.

Sweden: Cross-Country Household Balance Sheet Indicators

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

* Comparator countries have similar income levels to Sweden and high home ownership.
Figure 7.
Figure 7.

Sweden: Banking Sector Developments

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Figure 8.
Figure 8.

Sweden: Current Account and Savings-Investment Decomposition

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

External Position and Investment

(Percent of GDP; seasonally -adjusted; 2-quarter moving average)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: Statistics Sweden; and IMF staff calculations.

External Position of Surplus Countries

(Percent of GDP;2017)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Source: IMF World Economic Outlook database.

Real Effective Exchange Rate: ULC Basis

(Percent deviation from 1993–2018 average)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Source: OECD Monetary and Financial Statistics.

Prospects and Risks

7. Sweden’s growth is expected to slow in 2019. Domestic demand will ease further in 2019 as lower housing starts continue to feed into investment and the consumption softening in 2018 may take time to unwind. Weaker trading partner growth prospects suggest export growth will remain low in the near-term. Overall, real GDP growth is expected fall to about 1.2 percent in 2019, before returning to near the estimated 2 percent rate of potential growth over the medium term. Unemployment is expected to remain around the current low levels with the modestly positive output gap closing by 2020. A gradual increase in inflation reflects a combination of slowly rising inflation in the euro area and some rise in Swedish unit labor cost growth, in part reflecting a recent pick up in German wages. Eventual interest rate normalization will likely be associated with some krona appreciation. Medium-term projections for the current account surplus to stabilize at around 3 percent of GDP are consistent with prospects for the age structure of the population (Appendix II).

Selected Macroeconomic Indicators

article image
Sources: IMF WEO, Riksbank, Swedish Ministry of Finance, Statistics Sweden, and Fund staff calculations.

8. Downside risks to growth have increased. The fall in dwelling investment could exceed projections, depending in part on how the construction sector responds to the reduced availability of pre-sale financing. Weaker-than-expected global growth could stem from several factors, including a sharp tightening of global financial conditions and trade disruptions—perhaps related to a no-deal Brexit—with growth vulnerabilities in the euro area particularly relevant (Appendix VI). Although goods exports are equivalent to over one-quarter of Sweden’s GDP, analysis suggests that the direct impacts of tariffs are modest, with potential declines in global growth having most impact.

9. Although the housing market has cooled, household debt remains a macroeconomic vulnerability, and commercial property risks could be emerging. Aggregate household debt levels are not especially high (Figure 6), with the average loan-to-value ratio of 55 percent (and 67 percent for new loans) buffering house price declines. Although the risk of such declines appears to have moderated, vulnerabilities persist such as potential spillovers from financing difficulties among smaller property developers.3 Analysis by the Finansinspektionen (FI) finds households have sizable buffers against interest rate and other shocks, limiting the impact on debt service to banks. However, a macroeconomic vulnerability remains from the portion of households with high debt relative to their income (see section E), who may reduce consumption and reinforce an economic downturn, especially as almost three-quarters of mortgages have variable rates. Commercial property prices have been rising rapidly and valuations appear stretched by historical standards, although rental yields remain favorable relative to bonds which may reduce risks (Appendix III). However, a scenario with prolonged low inflation and interest rates could promote housing and commercial property price rises and exacerbate debt vulnerabilities.

10. In the next decade, Sweden also faces significant demographic shifts at both ends of the age spectrum. Sweden’s old age dependency ratio is relatively high, but the increase in coming decades is at the lower end (Appendix V). However, the population over 80 years old is expected to rise 55 percent by 2030 while the total population rises 9 percent. The cohort aged 10–19 years also rises relatively quickly. In addition to having significant fiscal costs, these demographic shifts require higher employment in social services (section D).

Population Growth from 2018 to 2030, by Age Cohort

(Percent)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: Statistics Sweden; and Ministry of Finance.

Authorities’ Views

11. The authorities also expect growth to slow, with the Riksbank projecting growth at 1.3 percent in 2019, and they saw similar risks to the outlook with uncertainty regarding the strength of global and domestic demand. The Riksbank agrees that the Swedish krona could be weaker than its long-term equilibrium, but there is substantial uncertainty surrounding the exchange rate in the longer term. Estimates from October 2018 imply that the krona can be expected to appreciate by about 5–15 percent in real terms over the next 5–10 years.

Policy Discussions

12. The recently-formed government has agreed a reform agenda to promote inclusive growth while maintaining Sweden’s prudent fiscal policy. Based on the “January agreement” with two center-right parties, a minority coalition government was formed in January following elections in September 2018. Some of the agreed measures will be adopted in the forthcoming Spring budget and the budget for 2020, but most reforms will be further developed and implemented during the term of the government. Key policy elements of the agreement include:

  • Fiscal: while observing the medium-term fiscal framework, environmental taxes are to be raised to finance income and payroll tax cuts, an austerity-related levy on high incomes is to be abolished, and taxation and regulations for micro-enterprises will be simplified.

  • Labor market: the social partners are asked to review employment protection, especially for SMEs, while maintaining legal certainty and protection against arbitrariness. The Public Employment Service is to be reformed to use private contractors for job placement.

  • Migrant integration: training for refugees, including in Swedish, is to be intensified and citizenship will include a Swedish language test.

  • Housing: liberalization of rent controls on new apartments is envisaged together with abolishing interest on deferrals of capital gains taxes on housing to facilitate mobility.

This consultation focuses on (i) labor markets including wage responsiveness at macroeconomic and sectoral levels (section B); (ii) housing markets including high housing prices (sections C, D, and E); and (iii) fiscal challenges, including demographic prospects in the coming decade (sections B and D).

A. Monetary and Exchange Rate Policies

13. Monetary policy remains accommodative following the hike in the repo rate in late 2018. At the time of the 2017 Article IV consultation, the Riksbank expected to begin rate increases in mid-2018, but the first hike of 25 basis points was deferred until year end, lifting rates to -0.25 percent. The Riksbank noted that the need for highly expansionary policy had decreased slightly with headline inflation and expectations established around 2 percent. At the same time, it emphasized that policy stimulus will be decreased slowly and it further lowered forecasts for the repo rate. For similar reasons, in February 2019 the Riksbank’s Executive Board chose not to extend a mandate that facilitated rapid intervention on the foreign exchange market, that had originally been adopted in early 2016 to signal its willingness to act if exchange rate developments put its price stability objective at risk.4 The Riksbank concluded government bond purchases in December 2017, having acquired some 40 percent of outstanding Swedish government bonds. Reinvestments of principal and coupon payments will continue until further notice.

Riksbank Forecasts of Repo Rate

(Percent)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: Sveriges Riksbank; and IMF staff calculations.

Monetary Instruments 1/

(Percent, left; SEK bil., right)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: Sveriges Riksbank; and IMF staff calculations.1/ Orange bars represent the cumulative amount of government bonds that the Riksbank had purchased by December 2018.

14. Inflation is expected to increase gradually, but significant uncertainty remains around the timing of durably reaching the inflation target. Aided by recent krona depreciation, the Riksbank projects CPIF excluding energy inflation to rise from 1.5 percent in 2018 to 1.9 percent in 2019, and to remain around that level subsequently. Staff expects a more gradual return to target, inflation reaching 1.7 percent by end-2019, and then rising toward the 2 percent target by 2021–22.5 This gradual inflation rise reflects projections for euro area inflation and an assumption that the recent increase in German wage growth continues, which, along with tight labor market conditions, would lift Swedish wage rises over time. The significant uncertainty around this central inflation outlook partly reflects uncertainties around euro area growth and inflation, including German wage rises.

Core Inflation Projections

{Percent change, y/y)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: Eurostat, Riksbank, Statistics Sweden, and IMF staff calculations.1/ Excludes package holidays, bank service fees, social protection services, and air transport services, in addition to energy and unprocessed food.

15. A data-dependent approach to monetary policy should continue in view of these uncertainties. Headline inflation measures and inflation expectations have risen close to target in recent years. But, with underlying inflation low and expected to rise only gradually, headline CPIF inflation could fall below target during 2019, depending on global and domestic developments. It is essential to keep inflation expectations well anchored, especially with the next round of centralized wage negotiations approaching in early 2020. It is therefore appropriate that the Riksbank has deferred further rate increases until the second half of 2019, and has made them conditional on the economic outlook and inflation prospects evolving as it expects.

16. The innovative evaluation of issuing a central bank digital currency (CBDC) is garnering global attention. Cash usage in Sweden has fallen dramatically, with most payments made via credit/debit cards or mobile phones using SWISH (Appendix VII). Increasingly, retailers and even bank branches no longer accept cash, making it less attractive. The potential for a few private providers to dominate payments without a viable alternative prompted the Riksbank to launch the e-krona project in 2017 to evaluate over a few years the issuance of a CBDC. This work has included an initial assessment of the demand for e-krona, at roughly 1–2 percent of GDP.6 An exploration of financial stability implications notes that banks would be no more vulnerable to institution-specific deposit runs, as they already face the risk of deposit transfers to other banks. In further work, the authorities should also explore regulatory options to ensure reliable and efficient private payments, such as standards for business continuity and regulation of fees.

Authorities’ Views

17. The Riksbank shared similar views on monetary policy, especially regarding the need to take into account uncertainty around developments abroad, the strength of domestic demand and the sustainability of underlying inflation pressures. Regarding an e-krona, consultation responses on the Riksbank’s second e-krona report indicate a need to involve a wider set of institutions and additional policy measures in further analysis of how to address challenges from changes in payment patterns and payment markets in Sweden. Since this issue is of considerable importance, a broader inquiry into the challenges posed by the changes in the payment landscape may be warranted, including the possible role of an e-krona.

B. Labor Market Policies

18. Sweden has adopted labor market reforms in recent years, but challenges remain. Seeking to raise employment of the low-skilled and migrants, the budget for 2018 boosted resources for education and streamlined active labor market policies with the aim of increasing employer participation. Nonetheless, unemployment rates of the foreign born and low-skilled much exceed that of natives, partly reflecting high minimum wages in collective agreements and relatively strict employment protection.7 Despite Sweden’s free floating exchange rate, nominal wage rises have become closely linked to those in Germany, with implications for inflation and monetary policy, and rigidities in relative wages impede mobility across sectors.8

Unemployment Rate by Skill and Nationality

(Percent)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Source: Statistics Sweden.

19. Further reforms, including of employment protection and public employment services, should support employment of the low-skilled and migrants, aided by enhanced education and training. The social partners’ plan for entry agreements, which is to be implemented in 2020, will help address these challenges by providing jobs for migrants and the long-term unemployed at a lower cost to employers as salaries will be supplemented by a state benefit for up to two years. Participants can take part in training courses agreed with the employer, including in Swedish, with the aim of gaining a permanent, full-time position with the employer. It is also important that the “January agreement” includes steps to strengthen the integration of migrants together with measures to enhance education and training. When the social partners review employment protection in collective agreements, in addition to considering increased exceptions for small- and medium-sized enterprises, other steps to facilitate labor market entry by the low-skilled and migrants should be evaluated, such as extending trial periods. Reforms of the Public Employment Service to use private providers for job placement need to be carefully designed and monitored to improve job matching and work skills development to ensure that migrants and other more difficult cases are not left behind.

20. Wage formation led by the industrial sector seeks to protect competitiveness. The agreement between industrial sector employers and unions sets a benchmark wage rise that other sectors are expected not to exceed, despite the industrial sector employing only 11 percent of workers. This centralized agreement is the principal wage driver, as “wage drift” over the agreement has declined notably. In practice, since 2012 nominal wages have remained quite stable relative to those in Germany (see chart) as the industrial sector seeks to maintain international competitiveness.9 In recent years the resulting wage rises have been subdued relative to the tightening of labor markets and rising inflation expectations. This has contributed to prolonging low inflation and low interest rates, which may also weaken the krona.

Swedish Wages Follow German Wages in National Currency Since 2012

(Index, 2012=100, Industry excl. construction)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: Haver Analytics, Labor Cost Index, and IMF staff calculations.

Wages Flat Despite Rising Shortages and Inflation Expectations

(LHS, percent, 12m moving avg; RHS, percent balance)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: TNS Prospera; National Mediation Office; NIER.

21. The social partners should consider updating wage formation to reflect structural changes in the Swedish economy, including the rise of other sectors such as services. One option is to broaden the range of sectors that set the benchmark for wage rises, with the National Institute for Economic Research continuing to provide information to the social partners. Such an approach would likely tighten wage linkages to domestic economic trends such as inflation expectations, productivity, and business profitability, enhancing Swedish macroeconomic stability.10

22. Scope for wage variation across sectors should also be increased. Sweden’s pattern bargaining tends to limit wage differentiation across sectors, limiting flexibility to adapt to shifting labor demand including from demographic change.11 To facilitate greater flexibility at the sectoral level to deviate up and down from the national benchmark, the mandate of the National Mediation Office in mediating sectoral wage negotiations should be amended to drop adherence to the industrial sector wage benchmark, to enable greater focus on its goals for real wage growth, higher employment, reducing labor market conflicts, and facilitating changes in relative pay.

Sectoral Wage Differentiation 1/

(Percent, 2017 Q2 – 2018 Q1)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: Eurostat; IMF staff calculations.1/ Standard deviation across sectors of sectoral average wages measured in percent of the national average wage.

Authorities’ and Social Partners’ Views

23. While recognizing the need for further progress on labor market challenges, such as decreasing the large differences in unemployment between domestic and foreign born, the authorities noted that Sweden’s integration framework along with a high level of general welfare, such as the right to daycare, had contributed to labor participation rates among the foreign-born that were relatively high compared with national averages in European countries. The authorities are alert to “cherry picking” and other risks in using private employment services and they intend to study international experience on how to mitigate such problems. Unions emphasized that Sweden’s employment protection was more flexible in practice than might appear, and they agreed that mediators sought agreements— including the value of terms and conditions—near the industrial wage benchmark. The social partners saw no barrier to broadening the range of sectors setting the wage norm, but they noted that the current system had delivered positive results in terms of employment, real income growth, and low labor disputes in the past two decades.

C. Housing Policies

24. High housing prices and high market rents increase vulnerabilities and inequality. Despite their recent moderation, house prices have tripled in real terms since the mid-1990s, lifting the price-to-income (PTI) ratio to almost 30 percent above its 20-year average, with Stockholm’s PTI nearly twice the national average and among the highest worldwide. New purchasers must take on high debts relative to income (DTI), typically at floating rates, a macrofinancial vulnerability (section E). Moreover, long queues for rent-controlled apartments meant that those unable to purchase housing had to pay much higher rents on subletted or newly constructed apartments, that are estimated to be 65 percent higher on average. An “insider-outsider” problem arises as labor mobility to the main centers is most impaired for those without parents able to assist with large down payments, impeding growth and exacerbating intergenerational and regional inequality.12

25. The “January agreement” includes promising steps but comprehensive housing reforms are needed given the scale of the problem. Key elements of a reform package include:

  • Making the rental market work: in addition to fully liberalizing rents of newly constructed apartments, there is a need to phase out existing controls. A common approach is to apply market rents when there is a change in tenant. Access to the housing allowance could be expanded to cushion this adjustment (see below), while also applying a temporary “windfall” tax on significant rental income gains. To help reduce market rents more quickly, rental supply should be increased, such as by reducing impediments to sub-letting and to households renting out their own apartments, while containing macrofinancial risks from buy-to-let housing.

  • Taxing property to rebalance the housing market: Sweden’s property tax was capped in 2008 to be among the lowest in the OECD, limiting the cost to households of occupying housing beyond their needs. A broad-based increase in the ceiling on the property tax would be most efficient, but an increase targeted to the main centers could be considered to incentivize mobility where it is most needed. Abolishing the interest on deferrals of capital gains taxes will help ease deterrents to mobility. An additional step to be considered is to tax only a portion of the capital gains on primary dwellings. Implementing a phase out of mortgage interest deducibility—which boosts housing demand and prices—would have limited impact on household finances currently given the low level of interest rates.13

  • Producing housing that is affordable: it is important to simplify the planning process to reduce the cost of construction which have risen by over 28 percent in real terms over the past 15 years, compared with 10 percent in the euro area. Productivity in the construction sector should be enhanced by strengthening competition, including by harmonizing land sale procedures across the municipalities and preventing requirements beyond national building standards in the approval process. The government should also expand existing subsidies for construction of affordable rental apartments, together with those for student and elderly housing in view of changing demographics.

House Prices and Construction Costs

(Percent change from 2002 to 2017)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: Eurostat, Haver Analytics, OECD, and IMF staff calculations.

26. Protecting households in the transition is important to help build broad consensus on comprehensive reforms. On average, the share of rental expenditure in disposable income would increase from 24 percent to 31 percent, with most impact on households in the lowest income decile, which could be cushioned by expanding payments under the existing housing allowance. However, this allowance is seldom paid to households in the second or third income quintiles, contributing to relatively low total expenditure on housing allowances in Sweden. There is a need to review the coverage and amounts of housing allowances to give confidence that a transition to market rents will be manageable.

General Government Spending on Housing

(In percent of GDP)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: Eurostat; and IMF staff calculations.

Share of Household Receiving Housing Allowances

(In percent, 2014)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: OECD; and IMF staff calculations.

Authorities’ Views

27. The authorities recognize the long-standing structural weaknesses in the Swedish housing market. They emphasized that housing policies were politically contentious in Sweden, limiting the feasibility of some reforms that could be most effective in principle, such as raising property taxation. Nevertheless, implementing the measures included in the “January agreement” during the term of the government would represent a step forward.

D. Fiscal Policy

28. Sweden should maintain its practice of allowing full operation of automatic fiscal stabilizers.14 A surplus of about 0.8 percent of GDP is expected in 2018, implying a cyclically-adjusted surplus of about 0.5 percent.15 Public debt is projected at 38 percent of GDP in 2018, declining in a baseline scenario to below 30 percent by 2024 (Appendix IV) such that Sweden has substantial fiscal space were it to be needed. During the process of forming a new government, the parliament approved a 2019 budget proposed by the Alliance of center-right parties, which included modest tax cuts (0.3 percent of GDP). Given the subsequent reduction in the growth outlook for 2019, the surplus will likely be below the original estimate of 0.9 percent of GDP as the automatic fiscal stabilizers come into play. Sweden’s practice of allowing these stabilizers to operate fully is especially useful given the downside risks to growth in 2019.

29. A lower medium-term surplus target is in effect from 2019. In 2016, Sweden modified its fiscal framework, to lower the medium-term fiscal surplus target from 1 percent of GDP to 0.33 percent, with effect from 2019, and it also established a debt benchmark of 35 percent of GDP.16 This reduction in the surplus target still allows Sweden to maintain prudent fiscal buffers against adverse economic shocks. In particular, meeting the surplus target implies that the net financial worth of the general government, of 27 percent of GDP in 2017, will be broadly stable, declining only a few percentage points over 10 years in baseline projections.

30. Transitioning to the new medium-term target by 2020 would release modest resources to facilitate structural reforms to support inclusive growth. Undertaking this small stimulus (about 0.2 percent of GDP) in the next few years entails little risk over overheating given the low level of inflation, especially with domestic demand expected to slow. Staff assumes the surplus would be reduced to about 0.5 percent of GDP in 2019, mostly reflecting lower growth, but much depends on the specific measures adopted in the Spring budget and when they become effective.

31. Looking further ahead, a temporary cut in the surplus target could help address higher public investment needs from demographic shifts or other sources. Demographic changes in the coming decade are relatively rapid.17 Studies by the Ministry of Finance and the Swedish Association of Local Authorities and Regions indicate needs for at least 1,400 additional schools, 700 housing facilities for the elderly, and 93 healthcare facilities through 2026. Major infrastructure investments such as new railway lines are also being considered. If these factors raise public investment beyond past norms (see chart), a temporary cut in the medium-term surplus target should be considered.

Investment of General Government

(In percent of GDP)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: Statistics Sweden; and IMF staff calculations.

Authorities’ Views

32. The authorities confirmed that Sweden would allow the automatic fiscal stabilizers to operate fully. They noted that the transition to the new medium-term fiscal target had started in 2018, such that the remaining resources to be released from this transition were small. Preparations for the Spring budget were just getting underway, but they saw little room for measures without new financing while meeting the new surplus target. The Fiscal Policy Council considered that any adjustments to the medium-term surplus target, should be made at the eight-yearly reviews, which could take into account factors such as public investment needs.

E. Macrofinancial Stability

33. The macroprudential framework was strengthened and a well-targeted tightening of macroprudential measures was adopted. Addressing a key recommendation of the 2016 FSAP to enable timely and effective macroprudential action, the FI’s macroprudential mandate was expanded from February 2018, giving it authority to apply measures subject to government approval but no longer requiring parliamentary approval. A stricter amortization requirement became effective in March 2018, raising minimum amortization on new mortgages with a loan-to-income (LTI) over 450 percent. Noting increased risk taking in the financial sector, the FI announced in September an increase of ½ percentage point in the countercyclical capital buffer to 2½ percent from September 2019, the highest in Europe.

Countercyclical Capital Buffer Ratio

(Percent of risk-weighted assets)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: ESRB; and BIS.Note: Information as of September 2018.

34. Macroprudential policy should remain on alert and household level data collection enhanced to help evaluate its effectiveness. The recent decline in housing prices and subsequent stability have been associated with some slowing in household credit growth. As the stricter amortization requirements take hold, the FI expects the share of new high LTI mortgages to decline substantially, helping to contain vulnerabilities over time. The use of higher amortization targeted at high LTI loans provides greater flexibility for households than an outright LTI ceiling, but its effectiveness will need to be monitored closely, especially if low interest rates are prolonged. To facilitate such assessments, it is essential that household-level balance sheet data are collected.

Mortgage Borrowers with High Loan-To-Income Ratio

(Share of new mortgage loans exceeding 450 percent of households’ gross income, percent)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Source: FI (The Swedish Mortgage Report, 2018).

35. Recent developments reinforce the need for strong anti-money laundering and counter-terrorist financing (AML/CFT) frameworks and for close regional cooperation. Swedbank, the largest lender in Sweden, is reported by the news media to have handled about US$4.3 billion in suspicious transfers in the Baltics between 2007 and 2015. Any issues in Swedish institutions should be identified and addressed. The authorities should continue to work to correct any remaining deficiencies in Sweden’s AML/CFT framework. Given the highly integrated financial sector in the region, strong cooperation and frameworks are key to the effectiveness of the AML/CFT regime.

36. The banking system is robust, yet it faces challenges from financial innovation, requiring steps to contain prudential risks without preventing competition. Swedish banks maintain high profitability and capital levels, and sound loan books (Figure 8). However, new entrants into the mortgage market are capitalizing on innovations in financial services technology, relatively high profit margins on mortgages, and the ready availability of market funding from institutional investors.18 These nonbank entities provide low-cost mortgages to high-quality borrowers (with less than 60 percent LTV for instance) and transfer the credit risk to end-investors such as life insurers. Although these nonbanks are currently a small share of the mortgage market, it is important that they meet the same macroprudential and consumer protection requirements, while ensuring compliance costs are manageable. Commercial real estate is another area where nonbank financing is playing a growing role (Appendix III), but bank exposure remains material, and FI should continue to closely review the adequacy of banks’ risk management along with the financial health of commercial property borrowers. In September 2018 one clearing member in Nasdaq Clearing defaulted, as collateral was insufficient to cover losses from trading power derivatives to which it was highly exposed. The FI should assess the adequacy of measures adopted to contain default risks and improve the efficiency of default management procedures.

Authorities’ Views

37. The authorities agree that after the recent tightening of macroprudential measures, it would be appropriate to review the effectiveness of the steps taken before launching further measures. After exploring options to enhance household data collection, the authorities have concluded that new legislative authority would be required. The authorities take very seriously recent reports of banks in the Nordic-Baltic region being used for money laundering. The authorities agree on the need for broader regional cooperation and the Swedish and Estonian financial supervisors have initiated a joint investigation, with the participation of the authorities of Latvia and Lithuania. The authorities note that Sweden’s AML/CFT framework has been strengthened in recent years, including through new legislation in 2017, and that FI has previously allocated additional resources to AML/CFT supervision. At the same time, they recognize that further action may be necessary.

Staff Appraisal

38. After sustained strong performance, Sweden’s economy is slowing. Accommodative monetary policy supported strong domestic demand gains in recent years, driving solid economic growth while narrowing the current account surplus. Job creation was rapid, especially among those born outside Sweden, although wage growth remained subdued and underlying inflation was broadly flat. Low inflation contributed to krona depreciation and the external position remained moderately stronger than the level consistent with medium-term fundamentals and policies in 2018. However, falling dwelling investment, softening consumption, and weaker exports are likely to slow growth notably in 2019, with material downside risks from the global economy and domestic demand.

39. A data-dependent monetary policy should continue. Aided by some depreciation of the krona, CPIF inflation excluding energy is expected to rise modestly by end 2019 and to gradually converge to target in following years, benefiting from a further pickup in wage rises. Yet this inflation outlook is subject to heightened uncertainty considering the risks to growth. The Riksbank’s deferral of further rate increases until the second half of 2019, depending on whether the economic outlook and inflation prospects are as it expects, was appropriate in these circumstances.

40. Fiscal policy should also support Sweden’s economic resilience. Given Sweden’s strong fiscal buffers, its relatively strong automatic fiscal stabilizers should operate fully to cushion lower growth in 2019. Reducing the cyclically-adjusted surplus to the new medium-term fiscal target by 2020 would release resources to facilitate reforms. The resulting fiscal stimulus would be small, with low risk of overheating at a time of slowing growth and still low inflation. Looking ahead, if demographic or other factors require public investment above past norms, a temporary cut in the medium-term surplus target should be considered to avoid tax or spending adjustments while protecting fiscal buffers.

41. Labor market reforms should aim to support employment of the low-skilled and migrants, aided by enhanced education and training. The unemployment rate of the foreign born and low-skilled remains very high. The social partners’ plan for “entry agreements” to enable migrants and the low-skilled to combine work and training at reduced cost to employers will help to address these challenges. When reviewing employment protection arrangements, the social partners should seek to facilitate labor market entry by the low-skilled and migrants. Reforms of the Public Employment Service need to be carefully designed and monitored to improve job matching and work skills development, especially for migrants and the low-skilled.

42. The social partners should consider updating wage formation to reflect structural changes in the Swedish economy, including the rise of other sectors such as services. Swedish wage rises tend to follow those in Germany, resulting in wage rises that are subdued relative to the tightness of domestic labor markets in recent years, contributing to low inflation, thereby prolonging low interest rates and a weak krona. One option is to broaden the range of sectors that set the benchmark for wage rises, which would likely tighten linkages to domestic economic trends such as inflation expectations and productivity, enhancing Swedish macroeconomic stability.

43. Comprehensive housing market reforms are essential to lower barriers to labor mobility and growth and to contain inequality. Tackling Sweden’s dysfunctional housing market requires reforms of rent controls, tax policies, and construction regulation. In addition to fully liberalizing rents of newly constructed apartments, there is a need to phase out existing controls, such as by applying market rents when there is a change in tenant. Access to the housing allowance could be expanded to cushion this adjustment, while also applying a temporary “windfall” tax on significant rental income gains. To incentivize efficient property allocation, recurrent property taxes need to be increased—including by phasing out mortgage interest deductibility—while reducing taxes incurred when moving. To reduce construction costs, it is important to simplify the planning process, but productivity in the construction sector should also be enhanced by strengthening competition, including by reducing barriers at the municipal level.

44. Financial sector risks should continue to be monitored closely. The macroprudential stance is appropriate following the adoption of stricter mortgage amortization requirements, yet risks should remain under close watch. Hence, household level balance sheet data should be collected to enhance monitoring of macrofinancial risks and facilitate the design and evaluation of measures. In view of stretched commercial property valuations, the authorities should continue to review the adequacy of banks’ risk management and the health of commercial property borrowers. Any AML/CFT issues in domestic institutions must be addressed, the authorities should continue to work to correct any remaining deficiencies in Sweden’s AML/CFT framework, and regional cooperation should be strengthened. In future inquiries related to the e-krona, the authorities should also explore regulatory options to ensure reliable and efficient private payments.

45. It is proposed that the next Article IV consultation with Sweden take place on the standard 12-month cycle.

Table 1.

Sweden: Selected Economic Indicators, 2016–24

article image
Sources: IMF WEO, Riksbank, Swedish Ministry of Finance, Statistics Sweden, and Fund staff calculations.

Data for 2019 is as of January.

Mortgage rates for new contracts.

Data for 2018 are as of Q4 2018.

Data for 2019 is as of February.

Based on relative unit labor costs in manufacturing.

Table 2.

Sweden: General Government Statement of Operations, 2016–24

article image
Sources: The 2019 Budget Bill; and Fund staff calculations.

Structural balance takes into account output gaps.

Table 3.

Sweden: Public Sector Balance Sheet, 2009–17

article image
Sources: Eurostat; Statistics Sweden; and Fund staff calculations.
Table 4.

Sweden: Balance of Payments Accounts, 2016–24

article image
Sources: Statistics Sweden and Fund staff calculations.

Positive number indicates an accumulation of foreign assets.

Percent changes of exports of G&S and imports of G&S are calculated using numbers in USD terms.

Table 5.

Sweden: Financial Soundness Indicators: Banks, 2012–17

article image
Sources: Financial Supervisory Authority, Riksbank, and Fund staff calculations.

The calculations follow rules under Basel II, including transition rules as reported by the Riksbank. Without transition rules, the capital ratios would currently be higher due to lower risk-weighted assets (the result of banks’ implementation of the IRB approach).

On consolidated basis.

The calculation of liquid assets follows the EBA guideline.

Table 6.

Sweden: Financial System Structure, 2015–17

article image
Sources: Financial Supervisory Authority, Riksbank, and Fund staff calculations.

Includes all major nonbank financial institutions of the banking groups and branches in abroad.

Not including minor local companies

Market value of funds

Number of institutions is computed on unconsolidated basis.

Appendix I. Underlying Inflation Developments

1. Headline inflation has returned to the 2 percent target, boosted by energy price increases. Both the CPI with fixed interest rates (CPIF)—used by the Riksbank to assess performance against its inflation target—and the internationally-comparable HICP have varied around 2 percent y/y since mid-2017. However, energy price increases have contributed heavily to headline inflation since 2016, together with food prices accounting for nearly half the headline rate in 2018.

Measures of Headline Inflation

(Percent change y/y, 3-month average)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: Haver Analytics; and IMF staff calculations.

Decomposition of CPIF Inflation

(Percentage point contribution to y/y inflation)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: Eurostat; Statistics Sweden; and IMF staff calculations.

2. Traditional core inflation measures exhibit an unusual temporary spike in mid-2017, making them less useful as policy guides. Core measures--excluding energy and unprocessed food--of both the CPIF and the HICP jumped to about 2 percent in mid-2017, only to decline quickly to as low as 1.0 and 0.8 percent y/y respectively by August 2018. This suggests that temporary factors other than energy and food drove up core inflation, making core inflation a less informative indicator of underlying inflation.

Traditional Measures of Core Inflation

(Percent change y/y, 3-month average)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: Haver Analytics; and IMF staff calculations.

3. A change in methodology and one-off price hikes were the main contributors to the spike in core inflation. A change to the measurement of package holiday prices resulted in a large (about 20 percent m/m) increase in this component in June and July 2017. Despite a weight of only 1.6 percent in the core HICP, package holidays alone contributed 0.5 ppts to y/y core HICP inflation of 2.1 percent in July 2017. One-off increases in financial services fees (3.2 percent weight) in June 2017 and social protection services (1.5 percent weight) in March 2017 together contributed an additional 0.3 ppts to the July 2017 peak in core HICP inflation. In addition, the air transport services series (with a 1 percent weight) has made highly volatile contributions from month to month during this period.

Contributions to Core HICP Inflation 1/

(Percentage point contribution to 12-month inflation rate)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

1/ Core HICP inflation defined as HICP inflation ex. energy/processed food.Sources: Eurostat; and IMF staff calculations.

4. A measure excluding these four components is notably more stable in recent years, at levels significantly below the inflation target. Removing these four components from the core HICP index produces an “adjusted core HICP.” As illustrated, the adjusted core HICP inflation rate is more stable in recent years, remaining around 1–1¼ percent y/y since 2015. It is notable that inflation in the traditional core HICP has recently declined to around this level.

Additional Measures of Core Inflation

(Percent change y/y, 3-month average)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

1/ Ex. energy, unprocessed food, air travel, holidays, bank services, and social protection services.Sources: Haver Analytics; and IMF staff calculations.

5. Another approach to assessing underlying inflation is to use an estimate of the trend in the price index. The aim is to separate the temporary price level movements—such as the mid-2017 shocks—from a more stable trend. In a state space model, the core HICP is modelled as the sum of a stochastic trend with a slope that follows a random walk, plus a stationary AR(1) component for the temporary deviations from that trend. A Kalman filter estimate of the slope tends to be smooth, suggesting it captures the more persistent component of inflation. It shows a broadly flat profile around 1¼ percent over the last two years. Nonetheless, there is significant uncertainty around this estimate, with the standard deviation of 0.2 percent giving a 66 percent confidence interval of 1.1 to 1.5 percent.

Core HICP Inflation

(Percent change, y/y)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

1/ Excludes energy and unprocessed food.2/ Also excludes package holidays, and bank/social protection/air transport services.Sources: Eurostat; and IMF staff calculations.

Appendix II. External Sector Assessment

article image

Notes on the External Sector Assessment

6. Revisions to the specification of the 2018 external balance assessment (EBA) model increased Sweden’s EBA current account (CA) norm. The 2018 External Sector Report undertook significant refinements of the EBA framework for assessing countries’ external balance position. For Sweden, this resulted in an almost 3 percentage point increase in the EBA current account norm, to 1.8 percent of GDP in 2017 up from -1 percent of GDP under the previous model.

7. Nonetheless, based on the historical performance of the EBA CA model, staff gives greater weight to exchange rate models in assessing Sweden’s external position. The fitted values of the EBA CA model for Sweden have been below the actual current account balance for the past two decades. These prolonged and substantial deviations suggest that factors not captured by the EBA model may also be driving Sweden’s savings-investment balances. Given uncertainties related to the EBA CA model gap estimates for Sweden, staff gives greater weight to estimates from the EBA REER models and the ULC based REER position.

8. The wide historical swings in Sweden’s current account balance are associated with the age structure of the population. Staff analysis of Swedish household saving found the old and young age dependency ratios to be significant explanatory variables.1 Similarly, the major rise and decline in Sweden’s current account broadly mirrors the share of prime age savers (persons aged 45–64), aside from a temporary deviation during the property boom-bust in the late 1980s and early 1990s.

Prime Age Savers and Current Account

(Percent, left; and percent of GDP, right)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: Statistics Sweden; and IMF staff calculations.

9. Demographic factors appear to be statistically robust drivers of trends in Sweden’s current account. A two-variable model that includes the share of children (0–14 year olds) in addition to the share of prime savers, was estimated for the 30-year period 1968–97. Its 20-year out of sample projections match the historical rise and fall in Sweden’s current account without large deviations. Based on official demographic projections by age group, this simple model suggests the current account will average about 3 percent of GDP over the medium-term.

Two-Variable Demographic Model of Current Account 1/

(Percent of GDP)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: Statistics Sweden; and IMF staff calculations.1/ Explanatory variables are trie shares of 0–14 and 45–54 year olds in the population.

Appendix III. Commercial Real Estate Market1

The commercial real estate (CRE) market is substantial in Sweden, utilizing sizable financing including bank lending. Strong growth in prices relative to rents has led to stretched valuations, especially if required rates of returns revert to historical norms. Increased reliance on bond financing and short interest rate fixing periods call for continued close monitoring to protect macrofinancial stability.

1. Commercial real estate is important for Swedish financial stability. The commercial property market has experienced substantial volatility historically, with implications for macrofinancial stability.2 Indeed, the 1990s banking crisis in Sweden followed an asset price boom that was amplified by financial deregulation in the mid-1980s, leading to a collapse in commercial property prices exceeding 60 percent in 1990–92.3 Defaults in the commercial property sector, such as construction companies and CRE developers, were the main source of bank credit losses, contributing to the wider banking-currency crisis which had major adverse macroeconomic impacts.

2. Sweden has a relatively large commercial property sector. MSCI estimates the size of the CRE market at nearly SEK 1.7 trillion ($200 billion) in 2017, or around 36 percent of GDP, making it among the largest in the world in relation to GDP.4 By comparison, the value of Swedish residential dwellings was about 150 percent of GDP in 2017. Unlike the U.K. and Ireland, Sweden faced a mild and partly temporary decline in CRE market value during the global financial crisis, and following valuation increases in recent years, total CRE market values are broadly in line with pre-crisis levels.

Commercial Real Estate in Select Economies

(Percent of GDP)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: IPD MSCI; and IMF World Economic Outlook database.

Commercial Real Estate in Select European Economies

(Percent of GDP)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: IPD MSCI; and IMF World Economic Outlook database.

3. Marketable commercial real estate in Sweden is primarily office space. In the subset of commercial property assets that are sufficiently marketable to be included in MSCI’s commercial property index,5 about half are in the office segment, predominantly in the three largest cities, Stockholm, Gothenburg, and Malmö. Some 16 percent of commercial real estate assets are in residential real estate (via multi-family dwellings, rental housing, and special-purpose housing) while 14 percent are retail properties.

Sweden: Commercial Real Estate Capital Value 1/

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Source: IPD MSCI.1/ Capital value of the IPD MSCI index for Sweden. The index tracks $95 bn of assets out of the $196 bn of total Swedish CRE assets (36 percent of GDP).

4. The financing structure of CRE has been changing but banks still have substantial exposure. Commercial property operators and developers rely primarily on debt financing from banks and nonbank financial institutions. But borrowing from financial markets has risen, with the FI estimating that one-quarter of the total interest-bearing debt of real estate companies was in bonds in 2017, up from about 5 percent in 2006. This direct exposure to wholesale funding can amplify market shocks to CRE borrowers and restrict their ability to rollover debt if rates rise suddenly. The availability of equity financing has also increased, often via cross-border inflows. Nonetheless, Swedish banks have significant lending exposure to commercial property, at 14 percent of their loan book in 2017. Yet, banks are funding a declining share of the CRE market, with bank loans declining from 45 percent of CRE market value in 2012 to 35 percent in 2017.

Credit Portfolio Composition of Major Swedish Banks

(Percent of total)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Source: Fl.

5. Recent property price and rental developments imply a decline in rental yields to low levels by historical standards, but spreads over bonds remain significant. Prices of CRE assets rose by almost 30 percent in 2015–17, with prices of office and retail properties recovering more strongly than the rest of the commercial property market from the 2008 slump. The price gains are far ahead of the steady 3 percent annual average increase in commercial property rents over the last 20 years. As a result, nominal yields of CRE assets declined to around 4 percent in 2017, down from an average of 5 percent in 2005–15. This brings CRE yields back to near their early 1990s lows, ahead of the property bust. But, at the same time, the spread of CRE yields to 10-year government bond yields has increased since the global financial crisis from ¼ percentage point to 3¾ percentage points currently. This spread increase could reflect domestic and external QE reducing bond yields with less spillover to less liquid assets like CRE, but it may partly be owing to a need for greater compensation due to investor caution. The FI notes that the combination of low yields and rapidly rising prices have historically been an indicator of future price falls, therefore increased monitoring of imbalances is needed.

Sweden: CRE Growth and Value Indices and Yield

(“Index, 1997 = 100: and percent, right axis)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Source: IPD MSCI.

6. Commercial property prices appear somewhat stretched, with estimated valuation gaps sensitive to changes in required returns. Staff analysis using standard market value and mortgage lending value models suggests some 13 percent overvaluation of the MSCI broad CRE index at current yield levels, assuming that rental values and net operating income yields remain unchanged.6 If yields were to revert to their 15-year average (approximately 5 percent), the estimated overvaluation would increase to 35 percent. Office properties appear to be overvalued by 10 to 38 percent under similar assumptions.

Commercial Real Estate Property Values and Valuation 1/

(Index, 2007= 100)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: MSCIIPD; and Fund staff calculations.1/ Thevaluation is the average of the market value and mortgage lending value methodologies (see Crosby, N. and Hughes C (2011), “The basis of valuations for secured commercial property lending in the UK.” Journal of European RealEstate Research, Vol 4, No. 3, pp 225–42).

7. Analysis by the authorities of borrowers’ repayment capacity suggests that risks to financial stability are contained, but the sensitivity to rising rates has increased. Analysis by the Riksbank of commercial property firms listed on the Swedish stock exchange finds that large interest rate shocks and substantial revenue declines are required for these firms to generate insufficient cash flows to meet their short-term debt liabilities. Similarly, in its 2018 Stability in the Financial System report, the FI notes the increase in the interest coverage ratio of CRE firms, as well as the reduction of their debt/equity leverage ratio, and concludes that the quality of commercial property credit is satisfactory. Yet, the FI warns that real estate companies are subject to greater refinancing risk due to their large debt and relatively short interest rate fixation periods, so it intends to continue monitoring developments.

Appendix IV. Debt Sustainability Analysis

Sweden Public Sector Debt Sustainability Analysis (DSA) – Baseline Scenario

(in percent of GDP unless otherwise indicated)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Source: IMF staff.1/ Public sector is defined as general government.2/ Based on available data.3/ Long-term bond spread over German bonds,4/ Defined as interest payments divided by debt stock (excluding guarantees) at the end of previous year.5/ Derived as [(r -π(1+g) – g + ae(1+r)]/(1+g+π+gπ)) times previous period debt ratio, with r = interest rate; π = growth rate of GDP deflator; g π real GDP growth rate; a – share of foreign-currency denominated debt; and e — nominal exchange rate depreciation (measured by increase in local currency value of US dollar).6/ The real interest rate contribution is derived from the numerator in footnote 5 as r – π (1 +g) and the real growth contribution as -g.7/ The exchange rate contribution is derived from the numerator in footnote 5 asae(1 +r).8/ Includes asset changes and interest revenues (if any). For projections, includes exchange rate changes during the projection period.9/ Assumes that key variables (real GDP growth, real interest rate, and other identified debt-creating flows) remain at the level of the last projection year.

Sweden Public DSA – Composition of Public Debt and Alternative Scenarios

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Source: IMF staff.

Appendix V. Sweden’s Intertemporal Fiscal Position

Sweden’s fiscal position is sustainable in the face of population aging given its sound initial balance sheet condition and relatively small expected increases in total aging-related expenditures. The latter is largely attributed to reforms that result in falling public pension spending as a share of GDP as the pension system switches from public to private schemes. Although other social expenditures, especially long-term care, are rising more rapidly than public pensions, the expected increases are lower than in peer countries thanks to more favorable demographic trends.

1. Sweden’s general government balance sheet is sound. Over the past two decades Sweden’s gross debt has fallen from around 70 to 40 percent of GDP, a relatively low level by international standards. Net financial wealth, defined as financial assets less gross debt, turned positive in mid-2005, reaching 29 percent of GDP in 2018. Financial assets consist mainly of public pension funds, holdings in state-owned enterprises, and more liquid investment assets. The Swedish government also has sizable non-financial assets, bringing net wealth up to 90 percent of GDP, higher than most advanced economies.

Government Debt and Net Financial Wealth

(in percent of GDP)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: NIER

General Government Balance Sheet, 2016

(in percent of GDP)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: October 2018 Fiscal Monitor, NIER

2. Sweden’s population aging is smaller than in many advanced economies. Sweden has an old age dependency ratio of about 30 percent, similar to Germany. Yet, the increase in this ratio by 2040 is among the lowest in advanced economies, at 10 percentage points, about half that expected in Germany. Sweden’s relatively favorable demographic prospects reflect a combination of higher fertility and immigration.

Old Age Dependency in Select Economies 1/

(Percent and percentage points on right axis)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: United Nations; and Fund staff calculations.1/ Old age dependency = 65+ year olds / 15–64 year olds.

3. The expected increase in total aging-related expenditures is among the lowest in advanced EU countries. By 2070, the total aging-related expenditures are expected to increase by only 1.6 percentage points (ppts) of GDP in Sweden, much lower than Norway (6.2 ppts), Germany (4.1 ppts) and the U.K. (4.5 ppts). Looking into the individual components, long-term care contributes most to the increase, with its expenditure share rising by 1.7 ppts from 2016 to 2070. Health care (excluding long-term care) expenditure is also expected to increase, by 0.8 ppts. Public pension expenditure, on the other hand, is projected to fall as a share of GDP, as discussed below. Expected changes in other social spending such as education and unemployment benefits are relatively marginal.

Expected change of aging-related expenditures from 2016 to 2070

(in percent of GDP)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: European Commission 2018 Ageing Report

4. Sweden’s pension reforms are the key driver of the projected decline in public pension spending as a share of GDP.1 A decomposition exercise from the European Commission shows that the change in public pension expenditures is mainly driven by four factors—the dependency ratio, coverage ratio, benefit ratio, and labor market dynamics.2 In Sweden, a falling benefit ratio contributes the most to the deceleration of public pension expenditures, primarily because a smaller share of pension payments come from the public pensions (the income pension), and more from the premium pension (private investment funds not included in general government).3 As the new system grows in importance, the public part of pension expenditure declines throughout the projection period, even as the total pension spending still rises faster than GDP. 4 A secondary contributor is the “sustainability factor” introduced during the pension reform that requires the pension system to smooth out the pension payments over a longer period of retirement in anticipation of rising life expectancy.5 The relatively favorable demographic outlook also contributes to smaller increase in public pension spending in Sweden.

Contributions to change in public pension expenditures, 2016–2070

(in percent of GDP)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: European Commission 2018 Ageing Report

Income and premium pension expenditures, 2000–40

(in percent of GDP)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: Statistics Sweden and NIER

5. Other social spending is projected to rise more rapidly, especially long-term care. Health care excluding long-term care spending, currently at around 7 percent of GDP, is close to the EU average. It is projected to rise as a share of GDP to 7.7 percent by 2070, which is relatively modest thanks to Sweden’s favorable demographics with the population becoming healthier and aging at a slower pace than its peers. Long-term care spending, largely affected by the old and very old (over 80 year old) population, is projected to increase more rapidly, to almost 5 percent of GDP in the next five decades from just over 3 percent 2016. This can be due to longer life expectancy, higher coverage rates of long-term care recipients, and more formally qualified care-givers.

6. Despite the rising fiscal cost from an aging population, the government’s balance sheet remains sustainable in the long run. The net present value (NPV) of the change in future public spending on pensions and health care is negative, contributing positively to Sweden’s intertemporal net wealth.6 Even accounting for the rise in long-term care spending, the total increase in aging costs is not expected to weigh heavily on the fiscal position going forward. Based on the projections by the National Institute of Economic Research (NIER), total government consumption (including publicly-funded welfare services such as health care, schools and elderly care), is projected to rise from 26 to 28 percent of GDP, which is offset by falling social transfers (including public pensions) from 14 to 12 percent of GDP. Government investment is expected to remain broadly unchanged at around 4.5 percent of GDP. The revenue ratio would also stay flat at around 44 percent of GDP assuming unchanged tax policy. This spending and revenue outlook yields positive net lending throughout the projection period at around 0.5 percent of GDP. Hence, NIER projects Sweden’s net financial wealth to rise from 26 percent in 2016 to close to 30 percent in 2040.

Net Present Value of Age-related Spendings Change, 2015–50

(in percent of GDP)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: Oct 2018 Fiscal Monitor

Appendix VI. Update on the 2016 FSAP Recommendations

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

C = continuous; I (immediate) = within one year; NT (near term) = 1–3 years; MT (medium term) = 3–5 years

Appendix VII. The Riksbank’s E-Krona Project

Several central banks have begun to explore the merits and risks of issuing a Central Bank Digital Currency (CBDC). The e-krona project initiated in early 2017 has attracted much international attention and put the Riksbank at the forefront of the development of CBDCs.

1. Declining use of cash is making it less useful. From already low levels in percent of GDP, cash circulating in Sweden has almost halved over the last five years, diverging from trends in some comparator economies. In addition, the number of automated teller machines (ATMs) has fallen by over 20 percent since 2012, outpacing the decline in all comparator countries, although this could also be a cause of the decline in cash usage or related to other factors, such as a merging of ATM networks across banks. Riksbank surveys find that consumers increasingly experience situations in which they cannot use cash, and the share of consumers that use cash has fallen significantly over the last four years, as they instead rely on mobile apps and credit/debit cards. Moreover, survey-based estimates show that two-thirds of Swedish retailers expect to no longer accept cash by 2030.

Coins and Notes in Circulation

(Percent of GDP)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Sources: BIS Statistics; and IMF staff calculations.Note: comparator countries include Australia, Canada, Euro Area, Hong Kong SAR, Japan, Korea, Singapore, Switzerland, and the United States.

Change in the Number of ATMs

(2012 = 100)

Citation: IMF Staff Country Reports 2019, 088; 10.5089/9781498305259.002.A001

Source: BIS Statistics; and IMF staff calculations.Note: comparator countries include Australia, Belgium, Canada, France, Germany, Italy, Japan, Korea, Singapore, Spain, Switzerland, and the United Kingdom.

2. What problems might arise if cash disappears? In time, it is possible that cash can no longer be used for payments with virtually all payments being made digitally using privately owned and controlled systems. From the Riksbank’s perspective, such a scenario is concerning as:

  • First, strong network externalities lead to high market concentration in the digital payment market, potentially allowing significant inefficiencies.

  • Second, economic vulnerabilities from disruptions in existing payment systems would increase. The Riksbank’s options to address such disruptions are limited when cash is no longer used, especially as much private payment infrastructure is located abroad.

  • Third, parts of the population are unable to use the technology for digital payment systems or could be denied access by payment system operators in future, effectively excluding them from being able to make payments efficiently.

  • Fourth, trust in the monetary system could be eroded in a financial crisis because bank deposits can no longer be converted into risk-free central bank money.

  • Finally, future developments in digital payment systems would be guided by private profit-maximization rather than societal benefits.

3. The e-krona project seeks to address these concerns. The e-krona would complement physical cash and be a digital claim on the central bank. The e-krona could offer a competitively neutral infrastructure which could lower payments costs. The e-krona payment system could be built in a more robust way with some offline functionality ensuring continuity when private systems fail. The e-krona could also be designed in a simple and user-friendly way to address the needs of parts of the population. The availability of e-krona would enable the general public to access risk-free assets at all times. Finally, with an e-krona available, the state would be able to better influence the design of payment systems.

4. Understanding the demand for e-krona as a store of value is essential from a financial stability perspective. During a financial crisis, the e-krona, which as a central bank money, is a risk-free asset, could become more attractive leading to deposit flight. However, Juks (2018) argues that historical evidence makes such a scenario unlikely. He further suggests that the e-krona be designed in a way to mitigate adverse effects on financial stability.

5. Preliminary estimates suggest that demand for the e-krona to make payments will be modest. Segendorf (2018) estimates that demand for the e-krona to meet transaction needs will not exceed the demand for cash amounting to 1–2 percent of GDP. His estimation relies on the assumption that a fixed share of payments will be made in e-krona, together with estimates of the volume of payments and assumptions on liquidity management.

6. However, such estimates are subject to significant uncertainty. First mover advantages of existing payment systems, and other technological innovations and improvements in the payment market, could weigh on the use of e-krona to make payments, at least initially. At the same time, demand could be higher depending on the additional benefits and incentives relative to existing payment systems which will be driven by the exact design of the e-krona. In addition, the extent to which the government uses e-krona in making and receiving payments will shape demand as well.

7. Future work plans. The Riksbank has already published two reports on the e-krona in 2017 and 2018, respectively.1 But the exact design of the e-krona is still being discussed, and currently, separate workstreams are evaluating and specifying the legal, technical and economic details of issuing e-krona which is scheduled to be tested in 2020. Once the tests are completed, the Swedish parliament would have the option to adopt new legislation to underpin e-krona issuance.

Appendix VIII. Risk Assessment Matrix1/

(Scale—high, medium, or low)

article image

The Risk Assessment Matrix (RAM) shows events that could materially alter the baseline path (the scenario most likely to materialize in the view of the IMF staff). The relative likelihood of risks listed is the staff’s subjective assessment of the risks surrounding this baseline (“low” is meant to indicate a probability below 10 percent, “medium” between 10 and 30 percent, and “high” between 30 and 50 percent).
1

Excluding full-time students would lower Sweden’s unemployment rate by about 2 percentage points.

2

Since September 2017, the Riksbank’s inflation target is based on the national CPI with fixed interest rates (CPIF).

3

Staff estimates an overvaluation of only 2 percent in 2018:H2 based on the methodology in Nan Geng, 2018, “Fundamental Drivers of House Prices in Advanced Economies,” IMF Working Paper, 18/164.

4

The Riksbank remains able to intervene in the foreign exchange market with the approval of the Executive Board.

5

The staff projection for CPIF ex. energy is based on models of adjusted core HICP inflation, which was 1.1 percent in 2018:Q4 and is projected to rise by about one-quarter of a percentage point per annum in coming years.

6

Sveriges Riksbank, Economic Review 2018:3, Special issue on the e-krona.

7

Giang Ho and Kazuko Shirono (2015), The Nordic Labor Market and Migration, IMF Working Paper WP/15/254.

8

The role of German wages in Swedish wage formation is analyzed in “Recent Wage Moderation in Sweden” in the 2017 Article IV Selected Issues paper. Box 1 of this paper provides a fuller description of wage formation in Sweden.

9

In practice, Sweden’s free floating krona implies that matching Swedish nominal wage growth to that in Germany is not sufficient to stabilize relative ULC and maintain competitiveness.

10

The OECD notes the possibility to take account of productivity and price developments in all tradable sectors, page 109 in Employment Outlook 2018.

11

See Swedish Labor Policy Council, 2016, Time for Larger Wage Dispersion?

12

On housing prices impediments to labor mobility, see Poghosyan, 2018, “Regional Labor Mobility in Finland,” IMF WP/18/252, and Hsieh and Moretti, 2015, “Housing Constraints and Spatial Misallocation,” 21154, NBER.

13

See studies of distributional effects of lower MID and higher property tax by the Swedish Fiscal Policy Council.

14

Sweden’s automatic stabilizers are relatively strong (see Figure 2.2 in the 2015 Spring Fiscal Monitor).

15

Fiscal data for 2018 are not yet available and estimates are subject to uncertainty. The National Institute of Economic Research estimates a decline in the structural fiscal balance of 0.3 percentage points in 2018 after adjusting for cyclical changes in the composition of tax base.

16

The fiscal framework is subject to review every eight years. The debt benchmark is not an additional operational target, rather the government is required to explain deviations of more than 5 percent of GDP from this anchor.

17

See the Ministry of Finance’s June 2018 Economic Update and associated demographic studies.

18

See the Riksbank on new players on the mortgage market and the FI’s stability in the financial system report.

1

See “High Household Savings in Sweden,” 2017 Article IV Selected Issues paper.

1

This appendix benefits greatly from discussions with Neil Crosby (University of Reading) and Christina Gustafsson (KTH Royal Institute of Technology in Stockholm).

2

Commercial real estate (CRE) refers office buildings, shopping malls, hotels, restaurants, etc. that are often leased or rented to businesses. Some definitions treat industrial properties as a subset of CRE.

3

See Englund (2015), The Swedish 1990s banking crisis: A revisit in the light of recent experience, Stockholm School of Economics Working Paper).

4

Earlier estimates by the Riksbank put the size of the CRE market closer to 40 percent of GDP.

5

Marketable CRE assets are those with more turnover and better availability of price and rental data. MSCI tracks SEK 814 billion in marketable CRE assets, or about half of the total market size (SEK 1.7 trillion) as of 2017.

6

Following the methodology described in Crosby and Hughes (2011), the basic models assume that the property rents at its rental value and the rental value estimate is also the net operating income (NOI). The capitalization rates for the mortgage lending value model are estimated using a long-term trend estimate for the NOI yield via an HP filter with parameter equal to 100 (annual data).

1

“A Reformed Pension System,” Memorandum of the Pension Working Group,1992, Edward Palmer, “The Swedish Pension Reform Model: Framework and Issues,” 2000.

2

European Commission, “The 2018 Ageing Report,” May 2018.

3

The investment funds consist of securities or shares that pensioners select.

4

“The Pension System and Pension Projections Until 2060,” Ministry of Finance, 2014.

5

“Fiscal Sustainability Report 2018,” National Institute of Economic Research, 2018.

6

The calculation for long-term care is not available from the October 2018 Fiscal Monitor.

1

These reports are available here: www.riksbank.se/en-gb/payments--cash/e-krona/

Same Series

Other IMF Content

Other Publishers

  • Collapse
  • Expand
Cover IMF Staff Country Reports
Sweden: 2019 Article IV Consultation-Press Release; Staff Report; and Statement by the Executive Director for Sweden
Author:
International Monetary Fund. European Dept.
Volume/Issue:
Volume 2019: Issue 088
Publisher:
International Monetary Fund
ISBN:
9781498305259
ISSN:
1934-7685
Pages:
69
DOI:
https://doi.org/10.5089/9781498305259.002

Real GDP Growth

(Percent, y/y annualized, seasonally-adjusted for 2018)

Housing Market Sentiment and Domestic Demand
Figure 1.

Sweden: Macroeconomic Indicators
Figure 2.

Sweden: Inflation and Monetary Policy
Figure 3.

Sweden: Labor Market Developments

Unemployment and Vacancy Rates

(LHS, percent of labor force; RHS, percent of occupied jobs plus vacancies)

Wage Growth in Sweden and Germany

(Percent change y/y; 6-month moving average)

Core HICP Inflation

(Percent change, y/y)

Indicators of Core CPIF Inflation 1/

(Percent change, y/y)
Figure 4.

Sweden: Housing Market Developments

House Prices

(Index, 2010=100, SA)

Residential Investment and Housing Prices
Figure 5.

Sweden: Select Financial Market Indicators
Figure 6.

Sweden: Cross-Country Household Balance Sheet Indicators
Figure 7.

Sweden: Banking Sector Developments
Figure 8.

Sweden: Current Account and Savings-Investment Decomposition

External Position and Investment

(Percent of GDP; seasonally -adjusted; 2-quarter moving average)

External Position of Surplus Countries

(Percent of GDP;2017)

Real Effective Exchange Rate: ULC Basis

(Percent deviation from 1993–2018 average)

Population Growth from 2018 to 2030, by Age Cohort

(Percent)

Riksbank Forecasts of Repo Rate

(Percent)

Monetary Instruments 1/

(Percent, left; SEK bil., right)

Core Inflation Projections

{Percent change, y/y)

Unemployment Rate by Skill and Nationality

(Percent)

Swedish Wages Follow German Wages in National Currency Since 2012

(Index, 2012=100, Industry excl. construction)

Wages Flat Despite Rising Shortages and Inflation Expectations

(LHS, percent, 12m moving avg; RHS, percent balance)

Sectoral Wage Differentiation 1/

(Percent, 2017 Q2 – 2018 Q1)

House Prices and Construction Costs

(Percent change from 2002 to 2017)

General Government Spending on Housing

(In percent of GDP)

Share of Household Receiving Housing Allowances

(In percent, 2014)

Investment of General Government

(In percent of GDP)

Countercyclical Capital Buffer Ratio

(Percent of risk-weighted assets)

Mortgage Borrowers with High Loan-To-Income Ratio

(Share of new mortgage loans exceeding 450 percent of households’ gross income, percent)

Measures of Headline Inflation

(Percent change y/y, 3-month average)

Decomposition of CPIF Inflation

(Percentage point contribution to y/y inflation)

Traditional Measures of Core Inflation

(Percent change y/y, 3-month average)

Contributions to Core HICP Inflation 1/

(Percentage point contribution to 12-month inflation rate)

Additional Measures of Core Inflation

(Percent change y/y, 3-month average)

Core HICP Inflation

(Percent change, y/y)

Prime Age Savers and Current Account

(Percent, left; and percent of GDP, right)

Two-Variable Demographic Model of Current Account 1/

(Percent of GDP)

Commercial Real Estate in Select Economies

(Percent of GDP)

Commercial Real Estate in Select European Economies

(Percent of GDP)

Sweden: Commercial Real Estate Capital Value 1/

Credit Portfolio Composition of Major Swedish Banks

(Percent of total)

Sweden: CRE Growth and Value Indices and Yield

(“Index, 1997 = 100: and percent, right axis)

Commercial Real Estate Property Values and Valuation 1/

(Index, 2007= 100)

Sweden Public Sector Debt Sustainability Analysis (DSA) – Baseline Scenario

(in percent of GDP unless otherwise indicated)

Sweden Public DSA – Composition of Public Debt and Alternative Scenarios

Government Debt and Net Financial Wealth

(in percent of GDP)

General Government Balance Sheet, 2016

(in percent of GDP)

Old Age Dependency in Select Economies 1/

(Percent and percentage points on right axis)

Expected change of aging-related expenditures from 2016 to 2070

(in percent of GDP)

Contributions to change in public pension expenditures, 2016–2070

(in percent of GDP)

Income and premium pension expenditures, 2000–40

(in percent of GDP)

Net Present Value of Age-related Spendings Change, 2015–50

(in percent of GDP)

Coins and Notes in Circulation

(Percent of GDP)

Change in the Number of ATMs

(2012 = 100)