Belgium: Staff Report for the 2017 Article IV Consultation
Author:
International Monetary Fund. European Dept.
Search for other papers by International Monetary Fund. European Dept. in
Current site
Google Scholar Close
Publication Date:
17 Mar 2017
eISBN:
9781475587630
Language:
English
Keywords:
ISCR; CR; center; report; IMF staff calculation; competitiveness loss; investment strategy; wage norm; sectoral wage negotiations; Labor markets; Labor taxes; Global
Download PDF (1 MB)

Following successful reforms during the government's initial year in office, the year 2016 proved to be more difficult. The terror attacks in Paris and Brussels had a significant, albeit temporary, effect on the economy. The fiscal strategy veered off track, with a sizeable overshoot of the deficit target. Growth prospects for 2017 and beyond are modest, as in other euro area countries. The Belgian labor market remains severely fragmented.

Abstract

Following successful reforms during the government's initial year in office, the year 2016 proved to be more difficult. The terror attacks in Paris and Brussels had a significant, albeit temporary, effect on the economy. The fiscal strategy veered off track, with a sizeable overshoot of the deficit target. Growth prospects for 2017 and beyond are modest, as in other euro area countries. The Belgian labor market remains severely fragmented.

Context—A more Difficult Year

1. After a strong start, reform progress has slowed, particularly on fiscal policies. The four-party right-leaning coalition government took big steps forward during its first year in office. The 2015 pension reform and the ongoing efforts to contain rising health care costs have been major advances toward addressing the cost of ageing. Wage moderation—including through the temporary suspension of indexation (saut d’index)—and the targeted cuts in the labor tax wedge under the “tax shift” have significantly reduced the labor cost gap with neighboring countries. Further labor market reforms to increase flexibility at the sector and firm levels, as well as a reform of the wage setting process to prevent wage competitiveness losses, are underway, although a comprehensive strategy for addressing the fragmentation of the labor market is still lacking. The government plans to reform the corporate income tax regime, but agreement has yet to be reached. Measures needed to achieve the government’s ambitious fiscal consolidation objectives are largely lacking, particularly those to sustainably reduce the high level of government spending.

2. Growth was held back by external factors and weak consumer spending. Real GDP growth is estimated to have declined to 1.2 percent in 2016, partly reflecting the weakness of the euro area recovery, as well as the significant, though temporary, impact of the Paris and Brussels terrorist attacks. Moreover, private consumption contributed markedly less to growth than in peer countries. This likely reflected in part wage moderation combined with the tax shift, which raised certain consumption taxes to offset labor tax wedge cuts. As a result of these tax policies, inflation came closer to the ECB’s target than in most other euro area countries (Figure 1). Unemployment, averaging 8 percent in 2016, has fallen closer to pre-crisis levels, supported by solid employment growth in the services sector.

Figure 1.
Figure 1.

Belgium: Macroeconomic and External Developments

Citation: IMF Staff Country Reports 2017, 069; 10.5089/9781475587630.002.A001

Sources: Haver Analytics, Eurostat, National statistical authorities, and IMF staff calculations.
A01ufig1

Growth Contributions of Selected Components

(In percent year-year, 2015-16 average of quarterly data)

Citation: IMF Staff Country Reports 2017, 069; 10.5089/9781475587630.002.A001

Sources: Haver Analytics and IMF Staff calculations.

3. Credit conditions have remained supportive. Corporate lending growth (in particular for manufacturing, real estate, and health services) has picked up alongside investment demand and interest rate reductions, while mortgage lending growth has slowed through 2016Q3 as reductions in tax deductibility of interest in certain regions and a higher VAT rate on renovations have become effective. Lending rates, particularly long term, have declined since the ECB’s quantitative easing policies were put in place.

A01ufig2

Growth of Credit to the Non-Financial Private Sector

(In percent contribution, excluding securitization)

Citation: IMF Staff Country Reports 2017, 069; 10.5089/9781475587630.002.A001

Sources: Haver Analytics and IMF Staff calculations.

4. Low oil prices and solid export growth brought the current account back into surplus (Figure 1). Export growth, including in chemicals and transport equipment, may have benefited from wage moderation policies. Nominal unit labor costs have been broadly flat since 2013, closing the cost competitiveness gap with neighboring countries based on an hourly labor costs measure. Meanwhile, the primary income balance has turned negative, despite Belgium’s strong net international investment position; the predominance of portfolio and other investment assets with low returns has been outweighing outgoing income flows. Nevertheless, the current account has moved back toward a moderate surplus and closer to its norm of 2.7 percent of GDP. Staff’s preliminary assessment under the External Balance Assessment suggests that Belgium’s external position is moderately weaker than medium-term fundamentals and desirable policies would suggest (see Appendix II), with a current account gap of -2¾ to -¾ percent and the real effective exchange rate overvalued by 2 to 7 percent.

5. Fiscal consolidation stalled in 2016, with a sizeable overshoot of the deficit target. Despite falling interest costs, the fiscal deficit widened to 2.7 percent of GDP, against an original budget target of 2.1 percent of GDP. Many factors have contributed, including higher expenditures from the impact of the terror attacks as well as revenue shortfalls reflecting some over-optimism on certain revenue yields and possible shifts in the tax base (such as from the withholding tax increase).1 There was also delay in policy implementation, such as on certain new tax measures and savings in health spending. As a result, both the headline and structural primary balance appear to have worsened, rather than improving as envisaged in the April 2016 Stability Program.

Text Table.

Budget and Projected Outturns1

article image
Sources: Haver Analytics, Authorities, and IMF staff estimates.

DBP = Draft Budgetary Plan, submitted to the European Commission in October preceding the budget year. Stab prog = April Stability program submission, consistent with revised budget submitted in May.

A Moderate Rebound Ahead but also Risks

6. Belgium’s economy is projected to grow at a moderate pace of around 1½ percent for the foreseeable future. Within the context of a lackluster euro area growth outlook, external demand is not expected to contribute significantly to growth in the medium term. Inflation should ease back as the impact of tax increases and recent energy price increases recede, but gradually increase over the medium term as the output gap closes. Potential output growth is estimated at around 1¼ percent, rising over the medium term as reforms already undertaken help raise participation rates, support continued employment growth, and reduce unemployment.

7. A protracted period of low growth poses significant risks in the context of a highly uncertain external environment. Given the economy’s openness and dependence on Europe, risks arise from both global and regional factors. There are uncertainties relating to upcoming elections in neighboring countries, the United Kingdom’s negotiations to exit the European Union, and the direction of U.S. policies, in particular on trade, corporate income taxes, and interest rates. Moreover, if domestic reform efforts and fiscal adjustment stall, within an environment of structurally low growth in advanced economies, Belgium faces the risk of persistent high public debt that could ultimately undermine confidence, especially if markets were to expect a gradual withdrawal of the exceptional monetary accommodation of the ECB. Under a “stagnation” scenario, with lower nominal GDP growth, public debt could turn explosive even if interest rates remain low, and there would be no room for fiscal maneuver (see below for a more detailed discussion on fiscal scenarios). This could become a source of macroeconomic instability and a deterrent to growth.

8. Prolonged low growth and interest rates would also create financial sector vulnerabilities. Banks’ earnings would be constrained because of fixed rate long term assets with low yields, while life insurers and pension funds need to generate sufficient returns to meet long-term liabilities. Staff simulations suggest a significant negative impact on bank profitability of a low growth scenario (see Section C below and Appendix IV). Low profitability would reduce retained earnings and therefore internal capital generation, which could have an adverse impact on future credit supply, and expose banks to shocks. Such shocks could be external, through tighter and more volatile global or regional financial conditions, likely affecting Belgium indirectly, through effects on larger neighbors. Domestically, banks would also become more vulnerable to housing price shocks. Banks are currently relatively well protected (see below), but the rise in household indebtedness bears close watch. With mortgage loans on full recourse terms, in the event of a shock households would likely try to avoid default and instead cut consumption. An unemployment shock could depress private consumption sharply, slowing growth, which could raise default rates more generally and reduce banks’ asset quality.

Deepening and Broadening Reforms

9. Belgium faces the risk of a high debt/low growth trap in the years ahead. Growth has recovered, but is expected to remain sub-par relative to pre-crisis. With a still weak recovery in the euro area, a slowdown in global trade, and political winds shifting against free trade, external demand may no longer serve as a strong engine of growth. Four broad structural challenges are weighing on economic prospects:

  • Belgium’s strong record of large primary fiscal surpluses was undone by the crisis, and progress in reducing high public debt has been slow since then, despite record low interest rates and the government’s ambitious consolidation goals. High and rising expenditure has undermined revenue-based consolidation and precluded a return to the pre-crisis fiscal dynamics.

  • There is still a long way to go in raising labor market participation rates, and addressing structurally high unemployment among certain vulnerable groups and in the Brussels and Wallonia regions.

  • There has also been little movement in product market reforms, while the business environment is affected by further devolution of regulatory power to the regional level under the sixth reform of the state.

  • At the same time, the financial sector is being challenged by the low growth/low interest rate environment, while also having to adapt to technological changes.

10. Addressing these challenges is all the more difficult in Belgium’s unique political economy and governance system. Besides the federal government, the system also comprises three regional authorities with significant autonomy, while at the same time recognizing three language communities (cutting across the regions), with separate competencies (e.g., in education). Authority to tax and spend is thus spread across different levels of government, with complicated sharing arrangements and sometimes overlapping responsibilities. This complex political and institutional setting means that a consensus on economic issues is difficult to achieve. In this context, the European fiscal governance framework has provided a welcome anchor to forge agreement on difficult decisions, which has allowed Belgium to exit the Excessive Deficit Procedure in 2014. However, it also means that fiscal policy needs to be negotiated across many different dimensions, and that deeper reforms, both fiscal and structural, can be challenging to adopt.

11. Policy discussions focused on how a broad and well-coordinated reform strategy could raise the country’s growth potential and bring down public debt sustainably. A central task is to balance the budget across all levels of government, with consolidation based on an ambitious and credible medium-term strategy—agreed between all levels of government—that delivers a gradual reduction in spending while safeguarding fiscal revenues. However, fiscal adjustment alone, especially if it is of low quality (e.g. based on across-the board containment), could hurt growth, thus yielding only limited benefits for fiscal sustainability. Given the potential impact on domestic demand, consolidation should thus be underpinned by high quality fiscal measures accompanied by reforms to raise Belgium’s growth potential, including by raising the low rates of employment among certain groups, pursuing a coherent infrastructure investment strategy, and strengthening competition in services. Flexibility in the sequencing of specific measures will be important, given Belgium’s governance context. The majority of such reforms (see below) should have a positive impact on growth, even in the short run. Fiscal and product market reforms should also generally help improve the budget position, while certain labor market reforms (such as more flexible wage setting, a youth minimum wage, improved education, training) could have a negative fiscal impact while still being positive for employment and growth. Combining the three reform areas is thus the approach that is most likely to be a net positive for growth, employment, and fiscal sustainability.

A. Consolidation Underpinned by Growth-Friendly Fiscal Reforms

12. Staff expressed concern about fiscal slippages, especially given the high level of debt and the uncertain outlook. After the sizeable budget deficit overshoot in 2016, the authorities have targeted a structural adjustment of around 1 percent of GDP for 2017, to be achieved via both revenue (e.g. raising withholding taxes further) and expenditure measures (mainly on health spending), supported by a further decline in the interest bill as well as continued savings from earlier reforms on wages and social benefits. In staff’s view, this ambitious target is hard to achieve without additional measures. As missing budget targets by significant margins two years in a row could harm credibility, staff emphasized the importance of implementing a budget that is based on a credible deficit target, backed by realistic revenue and spending assumptions, and supported by high quality measures.

13. A broad-based fiscal strategy would yield better growth and debt reduction outcomes. Under baseline projections with largely unchanged policies, public debt declines only modestly over the next ten years. In the event of a “stagnation” scenario in which nominal GDP growth is lower by 0.6 percentage points per year, the debt ratio becomes explosive (see Figure 2, and Debt Sustainability Analysis in Appendix V). This illustrates how sensitive Belgium’s debt is to the macroeconomic environment (including interest rates) and how little fiscal space Belgium has to address shocks, especially if outside the context of a coordinated European response.

  • With a view to gradually rebuilding fiscal buffers, staff recommended efficiency-oriented spending reforms (see below), anchoring primary spending growth to stay broadly flat in real terms over the medium run, together with revenue neutral but growth-oriented tax rebalancing and a package of labor and product market reforms. This could bring the structural fiscal position to near balance by 2021.

  • Such an approach (see scenario “high quality with reform” below) would have much stronger growth and debt reduction pay offs compared to a “low quality” adjustment scenario of across-the-board spending containment. A well-designed consistent policy package would prevent output losses while bringing the debt-GDP ratio almost 20 percentage points below the baseline over ten years (and 10 percentage points below a low-quality adjustment scenario without reforms).

  • As for near-term budget policies, staff strongly cautioned against overoptimistic adjustment assumptions that are not backed by quality measures. On balance, additional measures to achieve around ½ percentage point of GDP of structural primary fiscal adjustment per year would appear appropriate. Since high quality structural measures take time to put in place and yield results, a decisive start on such measures would be important, including already from 2017, in the context of a revised budget if deemed necessary.

Figure 2.
Figure 2.

Scenarios for Debt and the Deficit

Citation: IMF Staff Country Reports 2017, 069; 10.5089/9781475587630.002.A001

Stagnation: nominal GDP growth on average 0.6 percentage point lower than baseline.Low quality: flat real spending growth, multiplier of 1.15, no other reforms.High quality: flat real spending growth, multiplier of .85, product and labor market reforms.Multipliers informed by Kilponen et al, “Comparing fiscal multipliers across models and countries in Europe”, National Bank of Belgium Working Paper No. 278, 2015.

14. A sustainable strategy for growth-friendly medium-term consolidation would require broad reforms that are coordinated across levels of government. With further tax cuts scheduled for 2018–20 under the tax shift, staff projects that additional measures of about 2–2½ percent of GDP would be needed to achieve the goal of a balanced budget over the medium term. For high-quality adjustment, this should rely primarily on measures to improve spending efficiency. As shown in the text table, spending levels exceed those of peers in a number of areas, and significant savings could be achieved. Revenue-neutral tax reforms could help in promoting employment and growth. Agreement and implementation of such measures would require cooperation and participation across all levels of government. Current coordination mechanisms could be strengthened by binding commitments to meet agreed deficit targets, backed up by stronger budget processes, including centralized intra-year monitoring of outcomes across all levels of government.

Options for Fiscal Reforms

(In percent of GDP)

article image
Sources: Eurostat, and IMF staff calculations.

Effect estimated as a convergence to EU average.

Through attrition.

Subsidies are about 2.1 percent of GDP higher than the EU average and 2.0 percent of GDP higher than the three neighboring countries’ average (France, Germany, Netherlands). Reforms should focus on business subsidies that are high by European standards (see Hallaert and Nowak, 2015, Country Report 15/71, and the NBB’s Economic Review, September 2016).

Assuming that the redistributive power of social benefits is increased to the EU average, Belgium could achieve the same reduction in income inequality at a fiscal cost lower by 3¼ points of GDP. This number is reduced by the planned reforms of social security of 1.2 percent of GDP over 2015-18. Measures could include (1) tightening the eligibility rules of sickness and disabilities schemes and unemployment to avoid their use as early retirement scheme and (2) increasing the use of means testing. For details (See Hallaert, 2016, Country report 16/78).

Convergence to Euro area average.

15. Expenditure is at the heart of the consolidation challenge. After years of rapid growth, public expenditure amounted to nearly 54 percent of GDP in 2015, which does not necessarily translate into better services or social outcomes. There is significant scope to make spending more efficient, in particular by (i) reducing the high level of subsidies, (ii) making the division of labor between levels of government more efficient, (iii) accelerating ongoing reductions in public employment, (iv) enhancing means-testing and stronger control of social benefits, with a view to better targeting the most vulnerable.2 Greater efficiencies in spending would also make room for much-needed public investment spending (see below).

Figure 3.
Figure 3.

Belgium: Fiscal Context

Citation: IMF Staff Country Reports 2017, 069; 10.5089/9781475587630.002.A001

Sources: Haver Analytics, NBB, IMF World Economic Outlook, and IMF staff calculations.

16. Further tax reform could support growth while safeguarding revenues. The targeted personal income tax and social security contribution cuts planned for 2018–20 will appropriately further reduce the still high tax burden on labor. Staff suggested that well-targeted offsetting measures could help ensure revenue neutrality, including by strengthening environmental taxation, and eliminating deductions and exemptions, including on VAT and for company cars. The authorities’ plan to lower the relatively high corporate income tax (CIT) rate has clear merits. However, as discussed with IMF tax policy experts, this should be part of a broader and revenue-neutral reform of business and investment income taxation in order to maximize the growth impact without jeopardizing the consolidation strategy (Box 1).3 A key principle would be to create a more level playing field across business and investment activities, including through a review of profit tax deductions, rules against tax avoidance, capital gains taxation, taxes on interest and dividends, tax treatment of rental income and real estate, and tax preferences on savings accounts.

A01ufig3

Tax Revenue and Social Contribution, 2015

(In percent of GDP)

Citation: IMF Staff Country Reports 2017, 069; 10.5089/9781475587630.002.A001

Growth-Friendly Corporate Income Tax Reform

CIT reform is currently under consideration by Belgium’s coalition government. The relatively high CIT rate of 34 percent makes the tax base vulnerable to outbound profit shifting and can deter foreign investors, especially in a global environment of potentially intensifying tax competition. Belgium has historically attracted mobile businesses through targeted tax incentives, such as for intangible assets and financial services. Some of these are being revised to align Belgium’s tax system with EU state-aid rules, the new European Anti-Tax Avoidance Directive, and new standards under the Base Erosion and Profit Shifting project of the G20/OECD.

Belgium’s fiscal constraints limit the scope for lowering overall revenue from business and investment income taxation, especially as the labor tax wedge is still high. However, there should be room for reducing the statutory CIT rate since it is significantly higher than the average effective rate (see Figure), reflecting tax provisions such as for depreciation allowances, the tax credit for research and development, the innovation box, and the notional interest deduction. While Belgium’s average effective rate is similar to peer countries, the marginal effective tax rate is comparatively low, suggesting that the tax disincentive to additional investment by Belgian firms is more limited than in other countries.

A01ufig4

Statutory and Effective Tax Rates, 2015

(In percent)

Citation: IMF Staff Country Reports 2017, 069; 10.5089/9781475587630.002.A001

Source: Zentrum für Europäische Wirtschaftsforschung (2016).

CIT reform could be an opportunity to improve the broader business and investment income tax regime by reducing distortions and thus better balancing growth and revenue objectives. Areas for potential efficiency gains include the preferential rate for small businesses, the patent box, measures against international tax avoidance, and some specific deductions and exemptions. Some other forms of base broadening, however, could hurt rather than support economic growth, such as limiting R&D incentives or loss carry forward. Another reform element could be Belgium’s notional interest deduction (NID). The NID plays an important role in reducing the corporate debt bias by encouraging equity funding (IMF 2016), as it lowers the cost of capital and supports investment. Yet, it has also been used for tax planning via special-purpose vehicles of multinational companies, which could be addressed through strengthened anti-avoidance rules.

Finally, to ensure a level playing field across business and investment activities, the reform could usefully be broadened to include capital gains and dividend taxation, with a view to maintaining neutrality between different business forms and limiting tax arbitrage.

Authorities’ views

17. The authorities emphasized the critical importance of sticking to the coalition agreement, reiterating their determination and commitment to reach structural budget balance in the near term. They agreed that most of the consolidation effort should come from expenditures, and acknowledged that further measures could be required in this area. They agreed that budget credibility is critical, and are assessing the reasons for the deficit overshoot in 2016. In this context, they noted that some revenue measures under the tax shift had taken more time to implement, such as the tax on real estate investment vehicles, and an agreement with the regions on regularization, and hence the effects will only be fully felt from 2017. The authorities appreciated the broader perspective that Fund staff brought on corporate income tax reform, without detracting from the urgency of such reform.

B. Priorities for Boosting Potential Growth and Productivity

18. Public investment in infrastructure will be critical to support economic activity following years of erosion of the public capital stock. At 2.3 percent of GDP in 2015, public investment is well below EU average (3.9 percent). With new investment barely keeping up with the depreciation of fixed assets since the late 1980s, the capital stock has been declining relative to GDP, and is now one of the lowest in Europe (Figure 4). There is also scope for redirecting public investment to more productive investment projects such as transport infrastructure. Transport infrastructure, both rail and road, is perceived to be of lower quality than in the three neighboring countries and has declined in recent years, with traffic congestion an increasingly serious problem. Since network density is already high, the priority appears to be maintenance and efficiency improvements. To ensure efficiency, it will be critical to develop a comprehensive and prioritized infrastructure strategy that is agreed between all levels of government, and reflected in medium-term budget plans.

Figure 4.
Figure 4.

Public Investment, Productivity, and Regulation in Services

Citation: IMF Staff Country Reports 2017, 069; 10.5089/9781475587630.002.A001

Source: Eurostat, WEO and Public Investment Management Assessment (PIMA) databases, OECD, Federal Planning Bureau, and IMF staff calculations.

19. Greater competition in services would enhance productivity, while helping to reduce prices paid by consumers and firms. Competition could be increased in sectors such as telecommunications, retail, and the legal, architectural, and accounting professions. Protection of incumbents, including through the action of self-regulatory bodies that have, by law, regulatory power, and complex and lengthy bankruptcy procedures, contribute to a low rate of entry and exit in many services sectors. As a result, the services sector’s productivity growth is substantially weaker than in neighboring countries and services’ prices are comparatively high and dynamic, affecting the competitiveness of the tradable sector for which services are a crucial input (Figure 4). Some studies suggest that, if Belgium were to reduce the regulatory burden affecting services to the level of best practice observed in the OECD, TFP could grow by 1¼ percentage point more per year during five years. 4 Most of the gains would come from reform of professional services and retail trade.

20. Labor market reforms could provide a major boost to Belgium’s growth potential. Building on the reforms of unemployment benefits and pensions implemented since in 2012 (See Appendix VI) the government is undertaking several labor market reforms that should help keep Belgium’s labor costs in line with its neighbors, while increasing flexibility at the sector and firm level (Box 2). In particular, the reform of the wage setting mechanism (the so-called “1996 law”) aims to preserve the recent gains from wage moderation by linking wage growth to broader labor market and economic conditions. In this context, in setting the ceiling for wage growth, it would be important to consider not only wage developments in neighboring countries but also the differences in labor productivity growth. Moreover, while wage moderation has been successful in improving Belgium’s competitiveness, it has weighed on private consumption and growth performance. Going forward, it will be important to also focus on non-cost competitiveness by fostering innovation, improving education and training, and labor market performance.

Labor Market Reforms

The government has proposed several labor market reforms to Parliament. The main elements focus on:

Adjusting the wage setting mechanism. The 1996 law on competitiveness establishes, every two years, a “wage norm” to keep wage developments aligned with those of the three main partner countries.1 Recent wage moderation has closed the wage gap accumulated since 1996 and the government is reforming the calculation of the wage norm to avoid that rapid wage increases erode the effects of the temporary de-indexation and to close the wage gap accumulated before 1996 (the “historical handicap” that remains to be quantified by the Conseil Central de l’Economie). Specifically, the reform would (i) build in the calculation of the wage norm a “security margin” to avoid that, as in the past, forecast errors contribute to a wage gap (if not used, the margin would allow faster wage increase in the subsequent wage norm), (ii) require social partners to correct past slippages when establishing a new norm, and (iii) increase the fine for firms and sectors that do not respect the norm. The cut in employers’ social contributions implemented under the tax shift would be used to reduce the historical handicap and not to allow a faster increase in wages. To better take into account economic conditions, the calculation of the wage norm could take into account deviations in labor productivity growth between Belgium and its partner countries. The 2017–18 wage norm has been agreed in the context of the new mechanism.

Re-instating a minimum wage for the young. To tackle youth unemployment, a lower minimum wage for employees younger than 21 (eliminated in 2015) will be re-introduced. The minimum wage would be reduced by 6 points per year below 21 (6 percent reduction for a 20-year-old, up to a 30 percent reduction for a 16-year-old). The young would receive compensation from the government so as not to receive lower pay. The authorities2 estimate that between 13,300 and 17,660 wage earners could be concerned and the measure would have a direct fiscal cost of about €40.5 million (0.01 percent of GDP).

Increasing flexibility in work time. Legal work time would continue to be set at 38 hours per week on average, but the average would now be calculated on an annual basis rather than a quarterly basis. This increased flexibility would remain constrained by a maximum working time of 9 hours per day and 45 hours per week (with wage premium starting at 44 hours). However, a branch agreement could increase the limits to 11 hours per day and 50 hours per week. In sectors facing international competition (industry and services), it will be possible to calculate the 38-hour week average over a period of up to 6 years. This system already exists for some sectors such as the automobile industry. Possibility for night work would be extended to the E-commerce. Moreover, the administrative regulation of part time jobs will be simplified.

Changing training requirements. The legal requirement would shift from dedicating 1.9 percent of the wage bill to training to providing at least 5 days of training by equivalent full time employee.

1 Wages are indexed to inflation, but the wage norm sets a ceiling on real wage negotiations based on expected wage developments in Germany, France, and the Netherlands. The national wage norm then guides sectoral wage negotiations. 2 Masure, L., M. López-Novella, M. Saintrain, and P. Stockman, Herinvoering van de degressiviteit van de mininumlonen van min-21-Jaringen, Federal Planning Bureau, Rapport 11337, October 2016.

21. Raising the low employment rates among certain groups would help reduce the severe fragmentation of the labor market. While employment growth has been solid, the labor market remains severely fragmented across regions and skill levels (Figure 5). Many younger and older persons are not working, and non-EU immigrants are less integrated in the labor market than in other European countries (Box 3). Only about half of non-EU immigrants aged 25–55 are employed, compared to over 80 percent for Belgian-born residents. The reform of the minimum wage and of training are first steps that will need to be complemented by improvements in education and on-the-job training, in particular given the need for competency in multiple languages and unmet demand for certain technical skills.

Figure 5.
Figure 5.

Labor Market Fragmentation

Citation: IMF Staff Country Reports 2017, 069; 10.5089/9781475587630.002.A001

Sources: Haver Analytics, Eurostat, and IMF Staff calculations.

Authorities’ views

22. The authorities concur that there is a need to boost quality public investment, which adds to the fiscal consolidation challenge. Since transport projects such as the completion of the Antwerp ring or the 10-year plan for the public transport system in Brussels are typically large and occur at irregular intervals, the authorities argue that more flexibility in European fiscal rules is needed to avoid deterring such investment, in addition to the welcome recent clarification by Eurostat of the treatment of Public Private Partnerships. Moreover, the Prime Minister has announced that an investment plan for Belgium will be elaborated in consultation with sub-federal levels of government, which undertake most public investment. The government has also initiated plans to speed up the integration of immigrants who come from outside the EU. The Minister of Economy has commissioned a study to analyze the reasons of stronger services inflation than in neighboring countries, which could provide guidance on where greater competition may be needed. While emphasizing the importance to gains from wage moderation, the authorities agreed on the need to foster also non-cost competitiveness.

Which Groups Are Most Vulnerable in Belgium’s Fragmented Labor Market?

Belgium suffers from low labor market participation and high (un)employment disparities across regions and population subgroups. To better gauge the factors underlying labor market fragmentation, the relative likelihood of being out of or in a job conditional on belonging to a certain socioeconomic group was explored using a standard probit regression model estimated on microeconomic data for Belgium and various other countries from the European Union Labor Force Survey (EU LFS). This allows controlling for overlap between vulnerable sub-groups (e.g., low-skilled young workers living in Brussels) to assess which individual characteristics appear to be most important for determining the probability of employment or unemployment.

At the aggregate level, the results confirm significant group “penalties” for conditional unemployment probabilities affecting the young, the low-skilled, and non-EU immigrants. But while the youth and skill penalties are broadly in line with comparator countries (a bit higher than in stronger labor markets and a bit lower than in weaker labor markets), the unemployment probability penalty for immigrants from outside the EU is considerably higher than in most other countries. Moreover, the vulnerability of this group decreases more slowly with years of residency than in other countries.

article image

Coefficients represent the change in the probability of being unemployed compared to the base category unless otherwise noted. Base categories are shown in actual probabilities. All probabilities are conditional on the factors above and the following: sex, household composition, and years of residency.

Countries with significant labor market impact following the 2008 crisis: Italy, Ireland, Portugal, and Spain.

Countries with mild labor market impact following the 2008 crisis: France, Netherlands, the U.K., and Austria.

Comparing pre- and post-crisis probabilities indicates that the crisis has not exacerbated these penalties tangibly, which contrasts with EU countries whose labor markets were hit harder by the crisis. Estimates reaffirm strong regional disparities in labor market performance among different socio-economic groups, with Flanders’ residents of any socioeconomic group largely outperforming comparable groups in Brussels and in Wallonia. These results are broadly confirmed when looking at conditional employment probabilities, although immigrants seem equally placed to find a job across Belgium in the very first year upon arrival.

Mismatches in labor supply and demand and low geographical mobility have been advanced as factors explaining persistent underperformance of labor outcomes for young, immigrant and low-skilled workers in the non-Flanders regions (High Council of Employment, 2013). This suggests the need for a comprehensive strategy involving traditionally advocated measures on labor market flexibility, complemented by activation policies, migrant integration, and steps to increase labor mobility in order to help reduce fragmentation and unlock Belgium’s large untapped employment potential.

A01ufig5

Probability of Being Unemployed and Marginal Penalties

(Percent)

Citation: IMF Staff Country Reports 2017, 069; 10.5089/9781475587630.002.A001

Note: Marginal probability of the first category (lighter) over the second category (darker). The total reflects the probability of the first category.

C. Profitability Challenge to The Financial Sector

23. Private balance sheets appear solid in aggregate, although there may be pockets of vulnerability in households. Compared to other countries, Belgian households hold very high levels of financial assets, with net financial wealth reaching 238 percent of GDP in 2016H1. Much of this is held abroad, as reflected in the strong net international investment position.

  • Household debt has increased steadily since the mid-2000s, reaching 60 percent of GDP in 2016H1 and now above the average for the euro area. The debt service ratio remains contained at 7½ percent of gross disposable income. However, around 14 percent of indebted households had debt to income ratios above 3, and, in aggregate, debt to disposable income has increased to over 100 percent.

  • Corporate debt on a consolidated basis reached 119 percent of GDP in 2016H1, up from 106 percent of GDP in 2014, although this headline figure captures intragroup funding for multinationals locating treasury centers in Belgium via special purpose vehicles, including for tax reasons. Excluding these operations reduces this ratio by around half. Interest coverage ratios look solid, with the median firm earning (before interest, taxes, and depreciation) more than seven times the interest paid on financial debt in 2014, and the 25th percentile earning more than double.

A01ufig6

Household Net Financial Assets

(In percent of GDP)

Citation: IMF Staff Country Reports 2017, 069; 10.5089/9781475587630.002.A001

Sources: Haver Analytics and Eurostat.

24. The banking sector faces the challenge of a low growth/low interest rate environment. The banking system has restructured post crisis and built more comfortable capital buffers, with the Tier 1 capital ratio reaching 15½ percent, and NPLs remain low at 3.5 percent (Figure 6). Banks have become more focused on the domestic market, with legacy issues from Dexia waning as the bank is wound down. The system’s customer deposits comfortably exceed its loans, unlike many peers in neighboring countries. However, the low rate low growth environment has brought new challenges:

  • While the overall net interest margin has been broadly resilient in the face of negative/low interest rates (Appendix IV), margins on mortgage loans have been declining, and the prevalence of fixed rate loans and strong refinancing trends imply lower future returns for the banks through loan repricing. Moreover, the term premium on outstanding mortgage loans has come down close to zero. The resulting squeeze on profitability and internal capital generation could constrain future credit supply, potentially affecting growth. However, banks could also loosen lending standards to increase interest income by boosting volumes, increasing household and corporate leverage. At the same time, banks face challenges from digitalization and competition from non-banks, as well as cyber-security expenses, requiring nimble adaptation to remain competitive.

  • This environment raises the risk of search-for-yield behavior, as investors move funds away from low-yielding bank accounts and into real estate, asset management or other financial vehicles. These risks would stretch beyond the banking system and bear close watching and strong monitoring mechanisms.

  • In such a context, banks will need to shift to other sources of income and cut costs further to boost profitability and internal capital generation, to help them withstand the expected long period of low returns on loans. Indeed, the sector has been consolidating; the number of domestic banks has fallen, and some Belgian banks have reduced their workforce and closed branches with a view to cutting expenses.

Figure 6.
Figure 6.

Belgium: Challenges to Banks

Citation: IMF Staff Country Reports 2017, 069; 10.5089/9781475587630.002.A001

Sources: Haver Analytics, IMF Financial Soundness Indicators, and IMF Staff calculations.

25. Staff’s empirical analysis suggests that a low interest rate environment is harmful to bank profitability even after controlling for general macroeconomic conditions (Appendix IV). Based on a sample of more than 3,000 European banks, the analysis indicates that low nominal GDP growth and low interest rates tend to squeeze various measures of profitability. Drawing on the econometric estimates, a “stagnation” scenario (with a growth slowdown of 0.6 percent per year on average, as discussed above) accompanied by low rates and spreads through 2022, would result in a cumulative decline of ¾ percentage points in banks’ net interest margins, and sharply lower returns on equity. Moreover, given Belgian banks’ greater reliance on fixed-rate mortgage lending and retail deposit funding, the effects could be more pronounced than for the typical European bank.

A01ufig7

Impact on NIM by Different Bank Characteristics

(In cumulative percentage point change by 2022)

Citation: IMF Staff Country Reports 2017, 069; 10.5089/9781475587630.002.A001

Sources: Fitch Connect and IMF Staff calculations.1/ Higher-than-average deposit share in liabilities 2/ Higher-than-average mortgage share in total loan porfolio 3/ Total assets less than $1bn

26. Insurance companies are also vulnerable to the low growth/low interest rate environment. The sector’s profitability has been deteriorating, squeezed by legacy policies offering guaranteed high interest rates and declining yields on assets. And while most Belgian insurance companies comfortably meet the solvency capital requirement, the European insurance regulator’s 2016 stress test found that they would be vulnerable to a prolonged period of low interest rates. Belgian insurers would seem less hurt by a “low-for-long” scenario than many other European insurers, but would be among the more strongly affected by the “double-hit” scenario which combines a low yield environment with a sudden increase in risk premia. The average guaranteed interest rate is declining only slowly, falling to 2.82 percent at end-2015, and past-concluded contracts offering a guaranteed returns greater than 4.5 percent still accounted for 15 percent of inventory reserves. The authorities have taken policy actions to ensure the sector’s adjustment, including setting up action plans for capital strengthening of some smaller firms. More broadly, a new mechanism has been put in place for the annual setting of maximum interest rates on individual long-term life insurance contracts to better reflect current market conditions. A royal decree is expected to be issued soon granting the NBB greater powers to supervise the distribution of profit sharing or dividends by insurance companies.

A01ufig8

Impact of Stress Scenario on Excess of Assets over Liabilities

(In percent change)

Citation: IMF Staff Country Reports 2017, 069; 10.5089/9781475587630.002.A001

Source: EIOPA Report, 2016.

27. Further macro-prudential actions may be needed to address pockets of vulnerability in the housing market. Concerns relate to the continued growth in house prices, combined with rising household indebtedness and significant shares of risky mortgage lending practices5, as well as strong expansion in mortgage credit. Various overvaluation estimates are in the range of 0–20 percent. Policy actions could address these issues from various angles:

  • Direct remedies. In addition to the existing 5-percentage-point add-on to certain domestic mortgage risk weights, the NBB plans to impose additional capital buffers targeted at riskier loans, i.e. mortgage loans with loan-to-value (LTV) ratio higher than 80 percent, effective May 2017 (pending approval from the European authorities). Staff welcomed these measures and called for continued close monitoring of real estate market developments, including possible search-for-yield behavior in the housing market, and discussed possible further measures such as LTV or debt service to income (DSTI) limits to more directly target vulnerable borrowers.

  • Other buffers. An additional capital cushion, the capital conservation buffer, will also be raised by 2019 by eight banks deemed domestically systemically important, among which are the large mortgage lenders. Another buffer that could be deployed should strong credit growth persist is the counter-cyclical capital buffer, which is currently set to zero. On balance, staff considers this appropriate, in view of the various estimates of the credit gap of near zero6. Close monitoring is nevertheless warranted.

Authorities’ views

28. The authorities broadly agreed with staff’s assessment. On the banking sector’s challenges in a low-rate environment, they noted that there is scope for higher commercial margins on mortgage loans, but banks have also diversified to other sources of income (e.g., fees and commissions) and there is room for further adaptation of the cost structure. Regarding life insurance, they noted that the sector is restructuring, with some firms exiting and others no longer offering guaranteed products. In addition, NBB is developing several macroprudential tools, including a national stress test framework, to monitor risks in the life insurance market. On potential further macroprudential measures (such as LTV and DSTI) to more directly address housing market risks, the authorities noted that they plan to wait and see how effective current and upcoming measures are before considering these alternatives.

Staff Appraisal

29. The government continues to make progress on its reform agenda, albeit with less success on the fiscal strategy. Major strides taken during the government’s first year, building on previous efforts, include the saut d’index, the tax shift, and pension reforms. These should contribute to reducing the cost competitiveness gap with neighboring countries, support job creation, and reduce the long-term increase in the cost of aging. Additional labor market reforms are under way, notably a revision to the wage setting mechanism.

30. Deeper reforms would reduce the risk of a low growth/high debt trap. Complementary efforts to increase the efficiency of public expenditure, make the tax system more growth friendly, raise low rates of employment among certain groups, implement a coherent infrastructure investment strategy, and strengthen competition in services, would help raise the headroom for growth through faster job creation and productivity growth, and contribute to faster debt reduction. This will be particularly important in the context of increased global and regional uncertainties and structurally weak growth in key trading partners.

31. Given the high level of public debt, it will be essential for the public sector to be both a source of stability and a contributor to stronger growth. The sizeable fiscal slippage in 2016 has halted the nascent trend in public debt reduction. It will now be critical to implement a credible 2017 budget, backed by realistic revenue and spending assumptions, and supported by high quality measures. Beyond 2017, the main task is to balance the budget across all levels of government, which will require substantial additional measures. It should be underpinned by an ambitious and credible medium-term strategy—agreed and committed to by all levels of government.

32. The bulk of fiscal consolidation must come from the spending side. There is significant scope to make spending more efficient, through measures to reduce subsidies and public employment, reform social benefits to better target to most vulnerable, streamline the division of labor between levels of government, and improve the budget process across all levels of government.

33. Further tax reform should safeguard revenues while making the system more supportive of jobs and growth. The targeted personal income tax and social security contribution cuts planned for 2018–20 will further reduce the high tax burden on labor, and there is a need to ensure revenue neutrality through additional measures, as outlined in paragraph 16. In this context, there would be merit in lowering the corporate income tax rate if it is part of a broader and revenue-neutral reform of business and investment income taxation that is designed to support jobs, growth, and innovation.

34. A comprehensive plan to raise the low employment rates among vulnerable groups would provide a major boost to Belgium’s growth potential. While employment growth has been solid, the labor market remains severely fragmented across regions and population groups. Four broad priorities stand out: (i) preserving the gains from wage moderation by linking wage growth to broader labor market and economic conditions (supported by the reform of the 1996 law on employment and competitiveness); (ii) further reducing the labor tax wedge; (iii) improving education and on-the-job training, in particular given the need for competency in multiple languages and unmet demand for certain technical skills, while also promoting the integration of non-EU immigrants; and (iv) addressing barriers to geographical mobility, including traffic congestion and gaps in the public transport network.

35. Upgrading public infrastructure and strengthening competition in services would raise productivity growth and competitiveness. The backlog in public investment is weighing on economic prospects, with bottlenecks particularly severe in transport infrastructure. To ensure the efficiency of investment, it will be critical to develop a comprehensive and prioritized infrastructure strategy that is agreed between all levels of government, and reflected in medium-term budget plans. In addition, greater competition in services should be fostered—including in telecommunications, retail, and the legal, architectural, and accounting professions—which should help reduce prices paid by consumers and firms, while boosting productivity.

36. The financial sector will have to continue adapting to the low interest rate environment, and regulators should continue to closely monitor pockets of vulnerability in the mortgage market. The possibility of a prolonged period of very low growth and interest rates could put pressure on banks’ profit margins, especially given the high reliance on deposit taking and mortgage lending. To maintain the sector’s soundness and resilience, banks will need to pursue further cost reduction and diversification of revenue sources. The measures undertaken by the authorities and the continued restructuring efforts by both banking and insurance sectors to adapt to these challenges have been appropriate. It will also be important to monitor and address potential risks in the mortgage, life insurance, and shadow bank markets, and stand ready to deploy further macro-prudential measures when appropriate.

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

Table 1.

Belgium: Selected Economic Indicators, 2013–2022

article image
Sources: Haver Analytics, Belgian authorities, and IMF staff projections.

Contribution to GDP growth.

Table 2.

Belgium: Balance of Payments, 2013–2022

article image
Sources: Haver Analytics, Belgian authorities, and IMF staff projections.
Table 3.

Belgium: General Government Statement of Operations, 2013–2022

(In percent of GDP)

article image
Sources: Haver Analytics, Belgian authorities, and IMF staff projections.

Excludes the 2012 Dexia recapitalization.

Table 4.

Belgium: General Government Consolidated Balance Sheet, 2008–2015

article image
Sources: Haver Analytics and IMF staff calculations.
Table 5.

Belgium: Structure of the Financial System, 2010–2016

article image
Sources: National Bank of Belgium, Belgian Asset Managers Association, and Financial Services and Markets Authority.

On consolidated basis.

On company basis. Figure for insurance total assets in 2016 at market value.

For the number of credit institutions, data refer to situation of 30 October. For the number of pensions funds, data refer to situation of 3 November. All other data refer to the end-June 2016 situation.

Table 6.

Belgium: Financial Soundness Indicators of the Banking Sector, 2010–20161

article image
Sources: National Bank of Belgium.

Consolidated data. Data are based on the IAS/IFRS reporting scheme.

Appendix I. Main Recommendations of the 2016 Article IV Consultation and Authorities’ Response

article image
Source: IMF Staff

Appendix II. External Sector Report

article image

Appendix III. Risk Assessment Matrix1

article image

Appendix IV. The Impact of Low Growth and Rates on Bank Profitability1

This annex presents a stylized quantitative analysis of how a protracted period of low growth and interest rates might affect bank profitability in Belgium.

Belgium’s banking sector has undergone significant deleveraging since the Global Financial Crisis, with shifts in asset and liability structures. The scaling down of bank balance sheets has been accompanied by a reorientation of their business activities towards domestic and other core markets and more traditional business lines (National Bank of Belgium, 2016). During this process, banks have expanded their deposit base from an already high level and increased exposure to domestic households through mortgage lending—outstanding mortgage credit in total loans to the private sector has risen from under 30 percent in early 2012 to 43 percent in September 2016.

Bank profitability has held up well. Belgian banks’ returns on assets and equity (ROA, ROE) suffered large declines during the crisis, but most of them have quickly recovered afterwards. Profits have recently been improving—surpassing performance in neighboring countries (France, Germany)—despite sluggish domestic activity and falling interest rates in the euro area. A decomposition of ROA for the average Belgian bank suggests that interest savings from lower funding costs are more than offsetting the compression in interest incomes as a result of low/negative interest rates, supporting net interest margins (NIMs). The resilience in interest incomes observed so far is partly due to the inclusion of one-off penalties associated with the large volumes of refinanced loans, attracted by historically low rates.2 The picture looks similar for the four largest Belgian institutions; whose post-crisis (2013–15) net interest incomes stand higher compared to pre-crisis (2006–07) levels. Some banks have also managed to increase their non-interest revenues such as fees and commissions (BNP Paribas Fortis, KBC Group) or have benefitted from lower loan loss provisions post-crisis (BNP Paribas Fortis).

A01ufig9

Belgium: Components of Bank ROA

(In percent, weighted by bank assets)

Citation: IMF Staff Country Reports 2017, 069; 10.5089/9781475587630.002.A001

Sources: Fitch Connect and IMF Staff calculations.
A01ufig10

Belgian Big-4: Components of ROA

(Change between pre- and post-crisis, percentage points)

Citation: IMF Staff Country Reports 2017, 069; 10.5089/9781475587630.002.A001

Sources: Fitch Connect and IMF Staff calculations.

However, a protracted period of low growth and interest rates could be costly for banks. While the lending-deposit margins on the corporate side have remained largely stable, margins on household loans and in particular mortgages have been shrinking as a result of falling market interest rates and the flat yield curve. Together with the tapering in refinancing volumes (and hence revenues from penalties), banks’ NIMs are expected to decline going forward, especially if growth stays subdued. Recent consolidation efforts by Belgian banks have kept a lid on the cost-to-income ratio, which compares favorably with peer countries. But future profitability prospects will depend to a large extent on banks’ ability and willingness to cut costs further and/or adjust their business models by diversifying into other income sources.

A Quantitative Perspective

The relationship between economic growth/interest rates and bank profits is usually positive. Weak growth is typically associated with sluggish credit demand and deteriorating asset quality for banks, which weigh on profits. In addition, a sustained period of low interest rates associated with a flat yield curve in the context of economic stagnation can reduce bank income from maturity transformation activities by compressing margins. First, there is a limit to banks’ ability to pass on low (or even negative) interest rates to retail depositors, i.e. a downward stickiness in funding costs. Second, the flat yield curve associated with a “low for long” rate environment is detrimental to banks, who typically borrow funds in the short term and make longer-term loans and investments. In general, the degree to which fluctuations in interest rates impact banks’ NIMs and profitability depends importantly on the maturity and repricing structure of banks’ assets and liabilities, and the extent to which banks are able to shift their balance sheet and activities to adapt to interest rate changes. A growing empirical literature finds that low growth and interest rates weigh on banks’ bottom lines.3

Following the literature, we adopt the following empirical specification:

y i j t = β r j t + γ s j t + δ ′ X j t + θ ′ Z i , t − 1 + α i + μ t + ϵ i j t ⁢   ( 1 )

Here, yijt denotes a measure of profitability (e.g., NIM, ROA, ROE) of bank i in country j in year t, rjt the level of short-term (ST) interest rate (the 3-month interbank rate), sjt the slope of the yield curve, or the spread between the ST rate and 10-year government bond yield, Xjt a vector of macroeconomic variables (e.g., nominal GDP growth, real house price growth), and Z(i,t-1) a vector of (lagged) bank-level controls (e.g., asset size, capital, asset and funding structures). The financial and macroeconomic variables are at the country level. A full set of bank and year fixed effects are included, the former to control for unobserved time-invariant bank heterogeneity and the latter to absorb any common shocks. Per the theoretical discussion above, we expect positive coefficients on the ST rate, the spread, and GDP growth.

We explore the hypothesis that growth and interest rates may have differential effects along several dimensions. Following Claessens and others (2016), we test for possible nonlinearities in the interest rate-profitability relationship (with respect to the level of interest rates) by classifying countries each year as being in a low-rate environment if the average 3-month interbank rate is below 1.25 percent. We expect the relationship to become stronger for banks in a low-rate environment, as has been found in previous empirical work. Further, we split the low-rate sample along several banks’ characteristics: (i) asset size, (ii) share of deposits in total liabilities, and (iii) share of mortgages in total loans.4

The model is estimated using a rich cross-country bank level dataset. The bank level data—provided by Fitch Connect—come from annual financial statements containing detailed reported information on balance sheet components, incomes, and costs for each bank. We use data at the unconsolidated level to isolate the effect of a country’s macroeconomic and financial conditions on only the bank’s operations in that country.5 The estimation sample consists of about 3,400 banks from 15 advanced European economies spanning the 2006–15 period.

Results indicate that both lower growth and low interest rates are associated with weaker bank profits, with a significantly stronger effects in a general low-rate environment (Table 1). Specifically:

  • Growth. Using the full estimation sample, a one-percentage-point decline in nominal GDP growth is associated with a modest decline in bank NIMs of 2.4 bps, whereas the impact on overall profits (ROA, ROE) is not statistically significant. However, focusing on the low-rate period, the effect on NIMs increases slightly, and the estimates for ROA and ROE become much larger and highly significant.

  • ST interest rate. In line with findings in the literature, there is a strong empirical association between ST interest rates and bank profits.6 For example, a one-percentage-point decline in the ST rate is estimated to result in a 53 bps decline in NIMs in the low-rate environment compared with only 15 bps using the full sample, and the difference is statistically significant.7 The corresponding estimates for ROE are 4.7 versus 1.6 percentage points.

  • Yield curve. The slope of the yield curve is found to affect bank NIMs only when rates are low, with a flatter curve implying lower NIMs as expected. A one-percentage-point decline in spreads is estimated to reduce NIMs by a modest 8 bps. However, the spread doesn’t appear to have a systematic impact on overall profits.

  • Estimating the model on the sub-components of NIMs reveals that low interest rates tend to compress interest income more than they reduce interest expenses (consistent with the hypothesis of deposit rates being bounded below by zero). In addition, low growth/interest rates are associated with increased loan loss provisions but also higher non-interest income, the latter helping to mitigate the adverse impact on overall profits.8

  • Our results are broadly maintained across a range of robustness tests.9 10

Table 1.

Impact on Bank Profits

article image
Sources: Fund staff estimates.

ST rate below 1.25 percent. Notes: All specifications include bank-level controls (bank size, tangible common equity/tangible assets, deposit/liabilities, loans/assets) and bank and year fixed-effects.

The impact of a low-rate low-growth environment varies with bank size and asset and liability structure (Table 2). For example, we find that smaller banks’ NIMs tend to be more sensitive to GDP growth and the yield curve compared to larger banks—consistent with findings by Genay and Podjasek (2014) for the US, whereas the response to ST market rate is broadly similar across bank size classes. There is also evidence that the NIMs of banks with greater reliance on deposit funding or mortgage lending are affected more by low growth or low interest rates. While not reported here, results for ROA and ROE show similar patterns.

Table 2.

Impact on NIMs by Bank Characteristics

article image
Sources: Fund staff estimates.

Total assets less than $1bn.

Above-median share of deposits in total liabilities.

Above-median share of mortgage lending in total loans.

Notes: All specifications include bank-level controls (bank size, tangible common equity/tangible assets, deposit/liabilities, loans/assets) and bank and year fixed-effects.

Simulation of a Stagnation Scenario

To gauge the impact of a potential protracted low-growth low-rate period on banks, we construct a downside scenario. In this stagnation scenario, nominal GDP growth would be lower than the baseline projection by an average of 0.6 percentage points per year between 2016 and 2022, and the ST interest rate (and spread) would remain low at roughly the current level until 2022.11 This scenario emphasizes medium-term risks of economic stagnation as opposed to a near-term severe shock.

Banks would see their profitability squeezed under such a scenario, and the effect might be more pronounced in Belgium given high reliance on mortgages and deposit taking. Using coefficients from the estimated models (Table 1) and assuming other things stay constant (e.g., no change in bank business models), the scenario results in a cumulative ¾ percentage point decline in banks’ NIMs by 2022, with about two-thirds of the effect coming from the ST rate and one-fifth from growth. The corresponding decline in ROE is larger (almost 5½ percentage points) whereas the decline in ROA is more muted. Simulations conducted on the relevant bank subsamples (i.e. using estimated effects from Table 2) suggests an even more pronounced impact of low rates and low growth on bank profits for smaller-sized banks, or banks which rely more heavily on retail funding or mortgage lending. Thus, notwithstanding the benefits of a stable funding base, the high-deposit, high-mortgage characteristics of Belgian banks may imply greater profitability challenges in a protracted low-growth, low-rate period, given the zero lower bound on deposit rates and compressed mortgage margins.

A01ufig11

Illustrative Medium-Term Impact of Low Growth and Low Rates on Bank Profitability

(In cumul. percentage point change by 2022)

Citation: IMF Staff Country Reports 2017, 069; 10.5089/9781475587630.002.A001

Sources: Fitch Connect and IMF Staff calculations.
A01ufig12

Impact on NIM by Different Bank Characteristics

(In cumulative percentage point change by 2022)

Citation: IMF Staff Country Reports 2017, 069; 10.5089/9781475587630.002.A001

Sources: Fitch Connect and IMF Staff calculations.1/ Higher-than-average deposit share in liabilities 2/ Higher-than-average mortgage share in total loan porfolio 3/ Total assets less than $1bn

Appendix V. Debt Sustainability Analysis

Public debt sustainability risks remain high. Under the baseline scenario, the public debt-to-GDP ratio is projected to have peaked at 106.5 percent in 2014 and to decline to 100.9 percent by 2021 under broadly unchanged policies while economic recovery gains traction in line with neighboring countries. Gross financing needs are estimated at 18 percent of GDP in 2016 and are expected to decline to 16–17 percent in the medium term. The projected decline in public debt is relatively sensitive to macroeconomic shocks, particularly involving combined shocks.

Baseline and Realism of Projections

Under the baseline, economic recovery and broadly unchanged policies result in a small decline in the gross financing need and the public debt ratio over the medium term, broadly unchanged relative to the DSA from the 2016 Article IV consultation.

  • Macroeconomic assumptions. Growth is estimated at 1.2 percent in 2016 and rising to 1½ percent in 2017 on the back of stronger domestic demand. The output gap narrows over the medium term and is projected to close by 2020. Inflation is projected to reach 2 percent by the end of the projection period, reflecting the closure of the output gap.

  • Fiscal outlook. In staff’s baseline projections, the general government deficit comes down from 2.7 percent of GDP in 2016 to 2.3 percent in 2019, before rising again in the medium-term due mainly to unfinanced revenue losses from tax shift. The primary balance is projected to remain in small deficit throughout the projection period. The projected fiscal path results in a structural deficit of 2½ percent of GDP in 2021. This reflects staff’s assessment of the authorities’ plans over the medium-term, in which some measures remain to be identified.

  • Debt levels and gross financing needs. Belgium’s high level of government debt and gross financing requirement calls for using the higher scrutiny framework.1 Government gross debt has increased significantly since 2007, reflecting sizable fiscal stimulus, declining real and nominal growth, and a large recapitalization of (and financial support to) the banking sector. Public debt is estimated to have reached 105.8 percent of GDP in 2015 (helped by early repayment of bank recapitalization funds) and is projected to stay above 100 percent through the medium term. Gross financing needs are estimated at 17.9 percent of GDP in 2017 and are expected to remain around 16–17 percent of GDP in the medium term.

  • Realism of baseline assumptions. The median forecast error for real GDP growth during 2007–15 is 0.11 percent, while that for inflation (GDP deflator) is -0.3 percent, both of which are relatively small. The median forecast error for primary balance suggests that staff projections have been optimistic (a forecast bias of -1.1 percent of GDP).

  • Cross-country experience suggests the projected fiscal adjustment is feasible. The adjustment in the cyclically-adjusted primary balance (CAPB) over the projection period is small, reflecting broadly unchanged policies. Stronger and more sustained adjustment would be feasible if the authorities implemented significant quality structural measures (see discussion in the text).

  • Heat map. Risks from the debt level are deemed high given that the relevant threshold to which Belgium’s values are compared is 85 percent and this threshold is breached under baseline and all stress test scenarios. Belgium’s gross financing needs are just below the benchmark of 20 percent of GDP in 2014-15 and decline over the medium-term. Belgium also faces risks relating to its external financing requirement and a large share of public debt held by foreigners. At 99 percent of GDP, the external financing requirement is significantly above the upper threshold of early warning benchmarks and the share of debt held by foreigners is relatively high at 54 percent of total.

Shocks and Stress Tests

The DSA framework shows Belgium’s sensitivity to GDP growth and real interest rate shocks, and a combined macroeconomic-fiscal shock. They illustrate how difficult reducing public debt could be as the effect of these shocks, based on Belgium’s own historical record, play out. In some cases debt ends the projection period on a rising trajectory.

  • Growth shock. Under this scenario, real output growth rates are lower by one standard deviation starting in 2017–18, i.e. 1.6 percentage points relative to the baseline scenario. The assumed decline in growth leads to lower inflation (0.25 percentage points per 1 percentage point decrease in GDP growth). Under this scenario, the debt-to-GDP ratio increases to 111 percent of GDP in 2018 and declines thereafter.

  • High interest rate scenario. This scenario examines the implications for debt sustainability of an increase in spreads by 470 basis points (a historical high observed in 1995) starting in 2018. The deterioration of public debt and gross financing needs are back-loaded as old debt gradually matures and new higher interest rate debt is contracted. In 2021, the impact on financing needs is 2.9 percent of GDP. Debt dynamics are reversed as the debt ratio no longer follows a downward trajectory.

  • Real exchange rate shock. This scenario assumes 10 percent depreciation in the real exchange rate in 2017. This shock results in small effects relative to the baseline.

  • Primary balance shock. This scenario examines the implications of a revenue shock and a rise in interest rates leading to a cumulative 2 percentage points of GDP deterioration in the primary balance (one standard deviation shock) in 2017–18. This scenario illustrates risks of delayed fiscal adjustment, due to insufficient adjustment measures. This shock leads to a deterioration in the debt ratio of 0.8 percentage points relative to the baseline in 2021 and slightly higher gross financing needs.

  • Combined macro-fiscal scenario. This scenario aggregates shocks to real growth, the interest rate, the exchange rate, and the primary balance while taking care not to double-count the effects of individual shocks. Under this scenario, debt would reach 112.6 percent of GDP in 2018 and remain elevated through 2021. The impact on financing needs would be significant, keeping financing needs high through 2018 and remaining close to 20 percent of GDP through 2021. The higher gross financing need in 2021 reflects maturing debt.

  • Persistent fiscal deficits shock. This scenario a primary balance close to 1 percent of GDP through 2021. This results in the debt ratio hovering at current rates over the medium-term.

The authorities agreed with the conclusion of the analysis, and noted that foreign holdership of public debt can be a source of risk as well as a sign of confidence by investors in Belgium’s public finances. They also observed that the methodology did not take into account projections for the cost of aging, given its relatively short time frame.

A01ufig13

Belgium: Public DSA Risk Assessment

Citation: IMF Staff Country Reports 2017, 069; 10.5089/9781475587630.002.A001

Source: IMF staff.1 The cell is highlighted in green if debt burden benchmark of 85% is not exceeded under the specific shock or baseline, yellow if exceeded under specific shock but not baseline, red if benchmark is exceeded under baseline, white if stress test is not relevant.2 The cell is highlighted in green if gross financing needs benchmark of 20% is not exceeded under the specific shock or baseline, yellow if exceeded under specific shock but not baseline, red if benchmark is exceeded under baseline, white if stress test is not relevant.3 The cell is highlighted in green if country value is less than the lower risk-assessment benchmark, red if country value exceeds the upper risk-assessment benchmark, yellow if country value is between the lower and upper risk-assessment benchmarks. If data are unavailable or indicator is not relevant, cell is white.Lower and upper risk-assessment benchmarks are:400 and 600 basis points for bond spreads; 17 and 25 percent of GDP for external financing requirement; 1 and 1.5 percent for change in the share of short-term debt; 30 and 45 percent for the public debt held by non-residents.4 Long-term bond spread over German bonds, an average over the last 3 months.5 External financing requirement is defined as the sum of current account deficit, amortization of medium and long-term total external debt, and short-term total external debt at the end of previous period.
A01ufig14
A01ufig14

Belgium: Public DSA – Realism of Baseline Assumptions

Citation: IMF Staff Country Reports 2017, 069; 10.5089/9781475587630.002.A001

Source: IMF staff.1/ Plotted distribution includes surveillance countries, percentile rank refers to all countries2/ Projections made in the spring WEO vintage of the preceding year3/ Belgium has had a cumulative increase in private sector credit of 36 percent of GDP, 2012-2015. For Belgium, t corresponds to 2016; for the distribution, t corresponds to the first year of the crisis.Source: IMF staff.4/ Data cover annual obervations from 1990 to 2011 for advanced and emerging economies with debt greater than 60 percent of GDP. Percent of sample on vertical axis.
A01ufig15

Belgium: Public Sector Debt Sustainability Analysis (DSA) - Baseline Scenario

(In percentage of GDP unless otherwise indicated)

Citation: IMF Staff Country Reports 2017, 069; 10.5089/9781475587630.002.A001

Source: IMF staff1 Public sector is defined as General government. Medium-term projections for debt differ somewhat from the baseline due to differing methologies.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 = 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 U.S, 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 as ae(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.
A01ufig16

Belgium: Public DSA - Composition of Public Debt and Alternative Scenarios

Citation: IMF Staff Country Reports 2017, 069; 10.5089/9781475587630.002.A001

Source: IMF staff
A01ufig17

Belgium: Public DSA - Stress Tests

Citation: IMF Staff Country Reports 2017, 069; 10.5089/9781475587630.002.A001

Source: IMF staff

Appendix VI. Labor Market and Pension Reform Measures

article image
article image
article image
article image
article image
article image
article image
article image
Sources: Belgian Federal Public Service Employment, Labor and Social Dialogue, National Bank of Belgium, National Employment Office, Federal Planning Bureau, Ministry of Pensions, and Belgian Stability Program 2012–15.
1

Revenue shortfalls were partly offset by one-off collections such as from the excess profit tax, which was found illegal under European Union state aid rules.

2

See Hallaert, J.J. (2016), Belgium—Making Public Expenditure More Efficient, Washington, D.C.: IMF, Country Report 16/78.

3

Contributors: R. De Mooij and S. Hebous.

4

See B. Renaud, G. Cette, J. Lopez, J. Mairesse, and G. Nicoletti (2010b), The Impact on Growth of Easing Regulation in Upstream Sectors, CESifo DICE Report, vol. 8(3), pp. 8–12.

5

In 2015, 30 percent of new loans had loan-to-value ratios of over 90 percent, and a fifth had debt-service-to-income ratios of over 50 percent.

6

Estimates by staff, NBB, and European Systemic Risk Board, the latter after removing the effects of debt issuance of a large Belgian non-financial corporation in the context of a takeover of a large foreign company in 2016Q1.

1

The Risk Assessment Matrix (RAM) shows events that could materially alter the baseline path (the scenario most likely to materialize in the view of IMF staff). The relative likelihood is the staff’s subjective assessment of the risks surrounding the baseline (“low” is meant to indicate a probability below 10 percent, “medium” a probability between 10 and 30 percent, and “high” a probability between 30 and 50 percent). The RAM reflects staff views on the source of risks and overall level of concern as of the time of discussions with the authorities. Non-mutually exclusive risks may interact and materialize jointly. “Short term” and “medium term” are meant to indicate that the risk could materialize within 1 year and 3 years, respectively.

1

Prepared by Giang Ho (RES).

2

To refinance a loan, the customer has to pay a penalty equal to three months’ interest payments on the outstanding amount. The majority of mortgages in Belgium have fixed-rate contracts.

3

See e.g. Genay and Podjasek (2014) for US, Kok and others (2015) and Jobst and Lin (2016) for euro area, Busch and Memmel (2015) for Germany, Kongsamut and Tressel (2016) for France, Alessandri and Nelson (2012) for UK, and Deutsche Bank (2013) for Japan. On the other hand, Turk (2016) found that bank margins in Sweden and Denmark have been broadly stable despite negative policy interest rates as lower interest income was offset by reductions in wholesale funding costs and higher fee income.

4

Smaller banks may be in a less advantaged position compared to larger banks, who can better cushion the effects of low interest rates on profits by shifting their business focus to activities generating fee and commission income. Also, banks relying more heavily on retail funding and/or mortgage lending may be expected to see relatively more compressed NIMs in a low-rate environment, given the downward stickiness of deposit funding costs.

5

For example, two of the largest Belgian institutions (BNP Paribas Fortis, ING Belgium) are subsidiaries of foreign Global Systemically Important Banks (G-SIBs), and KBC has extensive operations in Ireland and Eastern Europe, so that using consolidated bank accounts would introduce many confounding elements.

6

Low short-term rates are typically associated not only with lower returns on assets but also lower funding costs, so the impact on profitability is a priori ambiguous. However, in a low-rate environment, margins tend to narrow as funding costs approach the zero lower bound. Accordingly, most studies find a strong correlation, with the short-term rate being a relatively good proxy for the interest rate conditions affecting bank margins.

7

These impacts on NIMs are slightly larger than Claessens and others (2006)’ findings for a larger set of countries.

8

Results are not reported but available upon request.

9

In particular, we (i) replace the 3-month interbank rate with the policy interest rate; (ii) include the VIX index instead of year fixed effects; (iii) use country fixed effects with errors clustered at the bank level instead of using bank fixed effects; (iv) correct for possible autocorrelation in the error structure by estimating the model using Generalized Least Squares; and (v) specify a dynamic model with the lag of NIMs as an explanatory variable. In all of these robustness tests, the coefficient on the short-term interest rate remains highly significant, with magnitude ranging from 0.03 to 0.29 (0.15 in the baseline). GDP growth and the yield curve also retain their explanatory power in most specifications, although the magnitude and precision of coefficients are relatively less consistent.

10

Robustness tests on the low rate sample comprise model specifications that include the short-term interest rate, the long-term interest rate and the term spread separately, as well as those including the short-term rate in combination with either the long-term rate or the interest rate spread. We find that the coefficients on the interest rate variables remain highly significant throughout and relatively stable compared to the baseline NIM regression.

11

With long-term yields having risen lately, these assumptions represent more a severe scenario.

1

For advanced economies that (i) have a current or projected debt-to-GDP ratio above 60 percent; or (ii) have current or projected gross financing needs-to-GDP ratio above 20 percent; or (iii) have or are seeking exceptional access to Fund resources; teams are required to use an extended set of tools to identify and assess specific risks to debt sustainability. For these “higher scrutiny” cases, teams are also required to produce a standardized summary of risks in a heat map and prepare a write-up to discuss risks, including any country-specific considerations.

Same Series

Other IMF Content

Other Publishers

  • Collapse
  • Expand
Cover IMF Staff Country Reports
Belgium: 2017 Article IV Consultation-Press Release; Staff Report; Supplementary Information
Author:
International Monetary Fund. European Dept.
Volume/Issue:
Volume 2017: Issue 069
Publisher:
International Monetary Fund
ISBN:
9781475587630
ISSN:
1934-7685
Pages:
69
DOI:
https://doi.org/10.5089/9781475587630.002
  • Figure 1.
    View in gallery
    Figure 1.

    Belgium: Macroeconomic and External Developments

  • View in gallery

    Growth Contributions of Selected Components

    (In percent year-year, 2015-16 average of quarterly data)

  • View in gallery

    Growth of Credit to the Non-Financial Private Sector

    (In percent contribution, excluding securitization)

  • Figure 2.
    View in gallery
    Figure 2.

    Scenarios for Debt and the Deficit

  • Figure 3.
    View in gallery
    Figure 3.

    Belgium: Fiscal Context

  • View in gallery

    Tax Revenue and Social Contribution, 2015

    (In percent of GDP)

  • View in gallery

    Statutory and Effective Tax Rates, 2015

    (In percent)

  • Figure 4.
    View in gallery
    Figure 4.

    Public Investment, Productivity, and Regulation in Services

  • Figure 5.
    View in gallery
    Figure 5.

    Labor Market Fragmentation

  • View in gallery

    Probability of Being Unemployed and Marginal Penalties

    (Percent)

  • View in gallery

    Household Net Financial Assets

    (In percent of GDP)

  • Figure 6.
    View in gallery
    Figure 6.

    Belgium: Challenges to Banks

  • View in gallery

    Impact on NIM by Different Bank Characteristics

    (In cumulative percentage point change by 2022)

  • View in gallery

    Impact of Stress Scenario on Excess of Assets over Liabilities

    (In percent change)

  • View in gallery

    Belgium: Components of Bank ROA

    (In percent, weighted by bank assets)

  • View in gallery

    Belgian Big-4: Components of ROA

    (Change between pre- and post-crisis, percentage points)

  • View in gallery

    Illustrative Medium-Term Impact of Low Growth and Low Rates on Bank Profitability

    (In cumul. percentage point change by 2022)

  • View in gallery

    Impact on NIM by Different Bank Characteristics

    (In cumulative percentage point change by 2022)

  • View in gallery

    Belgium: Public DSA Risk Assessment

  • View in gallery

    Belgium: Public DSA – Realism of Baseline Assumptions

  • View in gallery

    Belgium: Public Sector Debt Sustainability Analysis (DSA) - Baseline Scenario

    (In percentage of GDP unless otherwise indicated)

  • View in gallery

    Belgium: Public DSA - Composition of Public Debt and Alternative Scenarios

  • View in gallery

    Belgium: Public DSA - Stress Tests

Figure 1.

Belgium: Macroeconomic and External Developments

Growth Contributions of Selected Components

(In percent year-year, 2015-16 average of quarterly data)

Growth of Credit to the Non-Financial Private Sector

(In percent contribution, excluding securitization)
Figure 2.

Scenarios for Debt and the Deficit
Figure 3.

Belgium: Fiscal Context

Tax Revenue and Social Contribution, 2015

(In percent of GDP)

Statutory and Effective Tax Rates, 2015

(In percent)
Figure 4.

Public Investment, Productivity, and Regulation in Services
Figure 5.

Labor Market Fragmentation

Probability of Being Unemployed and Marginal Penalties

(Percent)

Household Net Financial Assets

(In percent of GDP)
Figure 6.

Belgium: Challenges to Banks

Impact on NIM by Different Bank Characteristics

(In cumulative percentage point change by 2022)

Impact of Stress Scenario on Excess of Assets over Liabilities

(In percent change)

Belgium: Components of Bank ROA

(In percent, weighted by bank assets)

Belgian Big-4: Components of ROA

(Change between pre- and post-crisis, percentage points)

Illustrative Medium-Term Impact of Low Growth and Low Rates on Bank Profitability

(In cumul. percentage point change by 2022)

Impact on NIM by Different Bank Characteristics

(In cumulative percentage point change by 2022)

Belgium: Public DSA Risk Assessment

Belgium: Public DSA – Realism of Baseline Assumptions

Belgium: Public DSA – Realism of Baseline Assumptions

Belgium: Public Sector Debt Sustainability Analysis (DSA) - Baseline Scenario

(In percentage of GDP unless otherwise indicated)

Belgium: Public DSA - Composition of Public Debt and Alternative Scenarios

Belgium: Public DSA - Stress Tests