HOUSING MARKET DEVELOPMENTS

Transcription

1 EUROPEAN SEMESTER THEMATIC FACTSHEET HOUSING MARKET DEVELOPMENTS 1. INTRODUCTION Housing market developments in the years preceding the financial crisis, and notably the accompanying rise in household indebtedness in a number of Member States, have contributed to the build-up of macroeconomic imbalances. Indeed, before 28 house prices were characterised by a long and unprecedentedly strong expansion in most Member States. This expansion was accompanied by large increases in construction volume and in credit. Since then, the correction experienced in EU Member States has been very uneven. In some Member States, notably in Ireland and Romania, real house prices decreased by more than 5% in the few years following the crisis. In contrast, in Belgium, Luxemburg and Sweden, prices have barely adjusted or have continued to grow. As the economic recovery is gaining ground, demand factors have picked up. Accordingly, house prices have accelerated in most Member States in 216. This development calls for renewed attention to the potentially destabilising role that large swings in housing markets can have and to the potential accelerating role that the credit market can play. The remainder of this factsheet reviews potential vulnerabilities in the housing market and policy actions taken in the Member States to address them. It first looks at developments in house prices and volumes to assess potential risks, notably linked with credit. Social aspects are also reviewed. In the second section, policy priorities are discussed, focusing notably on macro-prudential policy, tax policy and on measures to tackle supply constraints. Developments in the credit market are an important qualifier for the assessment of vulnerabilities related to the housing market. This factsheet should therefore be read in conjunction with the one on the banking sector. Tax policy can also have a significant impact on housing developments and should be taken into account. Finally, as housing policy has bearings on social protection and cohesion, the factsheet on this topic should also be considered. 2. CHALLENGES 2.1. House price dynamics A number of Member States are experiencing high or rising housingmarket-related vulnerabilities. In a context where interest rates have reach an historical low and where growth is strengthening in most Member States, demand pressures on housing markets are building up. House price increases picked up in most Member States (Figure 1). In Hungary and Sweden, the cumulative growth of house prices in 215 and 216 exceeded 2%. In Greece, Italy and Finland, real house prices continued to record negative growth in 216. For Greece and Italy, this means the ninth consecutive year of house price contraction. In order to assess the risks of housing price corrections, a number of valuation techniques can be used to complement the analysis of financial and housing indicators. These rely on the analysis of developments in the ratio Page 1

2 between house prices and rents (price-torent ratio) or households' gross disposable income per capita (price-to-income ratio). Statistical analysis is also used to disentangle the contributions of the various house price drivers. As house price time-series are only available in index, the absolute values for the price-torent and price-to-income ratio are not informative. In particular, they cannot be compared across countries. As a consequence, the analysis focuses on the deviation of the price-to-income and price-torent ratios from their long-term average. This gap provides a useful benchmark to gauge house price developments. The price-toincome ratio provides an indication of the efforts required by the average household to purchase a house. A level which is far above the long-term average signals that affordability has reduced. This can affect demand and bring house prices back to a more affordable level. Similarly, in equilibrium and for a given cost of capital, agents should be indifferent to the choice between owning and renting a house. Rent and house prices should therefore move together. The ratios' deviation from their long-term value can be interpreted as deviation from the equilibrium on the housing market. This would result in a change in the demand for rental and owneroccupied apartments, bringing the price-torent ratio back to its long-term level. Figure 1 House price growth % change in deflated House % CAGR Prices BE BG CZ DK DE EE n.a IE EL ES FR HR IT CY n.a. -4. LV LT LU HU n.a MT NL AT PL n.a PT RO n.a SI n.a SK n.a FI SE UK Source : Eurostat Page 2

3 SE LU BE UK AT FR FI DK CY CZ IT NL DE EL EE ES SK SI IE HU PT MT HR PL LT LV RO BG SE LU BE UK AT FR FI DK CY CZ IT NL DE EL EE ES SK SI IE HU PT MT HR PL LT LV RO BG Figure 2 Price-to-disposable-income ratio, Source: Eurostat, European Central Bank (ECB), Organisation for Economic Development (OECD), Bank of International Settlements (BIS), European Commission calculations. Note: Gap compared to the long-term average over The price-to-income and price-to-rent ratios confirm that the adjustment in house prices since 28 has been very high in most Member States. In most of them, the current level of these ratios is well below the 28 level and also below the long-term average (Figure 2 and 3). This suggests that the correction potential is limited for these countries and that further appreciation in real house prices can be expected. Meanwhile, a number of countries show ratios that are more than 1% above their long-term average. These include Belgium, France, Luxembourg, Austria, Sweden and the United Kingdom. For them, the countries that experienced only a small adjustment in housing prices since the financial crisis, the valuation ratios suggest that sizeable adjustment can be expected in the future. Figure 3 - Price-to-rent ratio, Source: Eurostat, ECB, OECD, BIS, European Commission calculations. Note: Gap compared to the long-term average over Page 3

4 Deflated HP growth, 216 (%) In order to complement the analysis through valuation ratios, statistical analysis can be conducted to analyse if the developments in real house prices are in line with what macroeconomic drivers would suggests 1. Indeed, in some cases, the strong growth in house prices may be linked to demographic and supply factors, to a pick-up in income or to the particularly low interest rates. In other cases, these usual fundamental drivers do not suffice to explain the rise in real house prices. This suggests that house prices could be overvalued. Combining the information from this fundamental analysis with the analysis through ratios, a valuation gap is estimated. This makes it possible to map countries based on the risk of correction in real housing prices and on the recent developments (Figure 4). The mapping shows that in most Member States, real house prices are within 1% of the level suggested by the various valuation metrics. Looking to countries where prices have increased fast recently, two groups can be delineated. In the first group, including Luxembourg, Austria, Sweden and the United Kingdom, prices are increasing at a fast pace although the valuation gaps are sizeable. These countries are thus those where price dynamics need to be monitored most closely as they cumulate risks of overvaluation and dynamic price developments. In a second group, including notably Bulgaria, the Czech Republic, Latvia, Hungary and Slovakia, house prices have picked up recently and are increasing fast. Still, real house prices in these Member States remain below the peak values recorded in 28 and valuation metrics do not show significant overvaluation. In these cases, the increase in prices could give rise to speculative bubbles which may be difficult to deflate. Among countries with large valuation gaps, Belgium and France stand out. In both countries, the correction in prices and in valuation metrics following the 28 crisis was relatively limited. However, housing price developments since then have remained muted. As a consequence, while risk have not deflated significantly since the crisis, the sluggish growth in house prices, if sustained through the economic recovery, could lead to a soft return towards fundamental values. Figure 4 Valuation gap and real growth in house prices, HU RO LT IE NL BG CZ SK PT DE ES MT EE SI LV DK AT UK LU SE 2 PL HR IT CY FI FR BE -2 EL Estimated valuation gap, 216 (%) Source: European Commission, ECB, OECD, BIS, European Commission analysis. Note: The valuation gap was calculated based on the price-to-rent ratio, the price-to-income ratio and a statistical model for fundamental drivers of house prices. 1 For details on the methodology, see Philiponnet N. and A. Turrini (217), 'Assessing house price developments in the EU', Discussion paper, European Economy. Page 4

5 2.2. Volume developments Volume indicators for housing markets are useful for quantifying developments in construction activity, and should be seen as a complement to house price developments. In particular, developments in the housing market are affected by how the housing supply reacts to price and demand pressures. Supply constraints, which induce a slow response of volumes to prices, can result in longer episodes of booms and busts, with higher price volatility. On the other hand, a more responsive housing supply means that hikes in house prices translate into higher levels of construction activity. While rising supply has a mitigating effect on price developments, it also means that the macroeconomic impact of a correction in house prices accompanied by a drop in volume will be much more severe. Residential construction is a volume indicator that measures investment in residential buildings in a given period. On average in the EU, dwelling investment represented 4.7% of GDP in 216 2, markedly below the 6.% peak reached in 27 but also below the pre-boom levels. Investment in dwellings increased in 216 across the Member States, with only a few exceptions. This reflects the general improvement in macroeconomic conditions and shows that, as a share of GDP, dwelling investment remained close to its 215 level. In some Member States (e.g. Spain), the drop in housing investment reflects the overinvestment of a few years ago; in others, this is related to impaired credit supply and demand, and regulatory bottlenecks. Germany, Austria, Sweden and the United Kingdom are the only Member States where investment in dwellings as a share of GDP was higher in 216 than before the crisis. Beyond the actual investment, the number of building permits awarded within a year is an early indicator of residential investment. This figure gives an indication of the number of development projects under way. Building permits have rebounded in 216, with strong growth recorded in some countries, notably those where the construction sector suffered most following the crisis (Figure 5). The Netherlands and Romania are the only Member States in which issuance of new building permits is still decreasing. At the other extreme, the number of building permits is increasing fast in some Member States, notably in Hungary and Malta. Despite the recent growth, construction activity in 216 was generally well below its pre-crisis level. Germany, Lithuania, Poland, Slovakia, and Sweden are notable exceptions. In these Member States, the number of building permits issued is above the average recorded for the 2-27 period and continues to increase at an annual rate of more than 1%. Figure 5 Number of building permits, in % change, NL RO EL IT SI FI BG CZ UK DK EE AT ES PT IE CY LV HRHU FR BE MTDE SK PL LT SE Contracting <1% growth p.a. >1% growth p.a. Level relative to 2-7 average Latest annual growth rate Source: European Commission. Note: Data for Luxembourg is not included. 2 Source: European Commission. Page 5

6 While building permits focus on new buildings, the number of residential property transactions relates to both new and existing buildings. As is the case for building permits, the number of overall transactions has decreased dramatically after the crisis and now appears to be picking up again. Still, this process remains very slow and both the number and the total values of transactions remain well below pre-crisis levels (Figure 6). Figure 6 House transactions, Value (EUR bn.) Number (') Source: ECB Note: Data for the EU-28 where available. When information for a specific country is missing, values are grossed up to the EU-28 average to avoid that the aggregate is impacted by country composition Impact of credit developments Besides the potential valuation gap, the potential risks stemming from housing price developments are notably linked to the level of banks' exposure to mortgages and the indebtedness of households. Developments in the housing market can significantly contribute to vulnerabilities in the financial sector. This includes banks' growing reliance on mortgages, persistently high leverage and weak lending standards (e.g. elevated loan-tovalue ratios, long loan amortisation maturities, low risk-weights on banks' balance sheets for real estate exposure, etc.). In countries where household indebtedness is high (e.g. Denmark, the Netherlands and the United Kingdom), challenges linked to households' financing capacity may also be a risk for banks 3. While house prices and construction appear to be picking up in most Member States, developments in mortgages are more subdued. In a number of Member States, households still face significant deleveraging needs. Notably, household debt exceeds gross disposable income in almost half of Member States. The flow of mortgages is still negative in Ireland, Spain, Portugal and the United Kingdom. In some cases, notably in Hungary and Croatia, the lingering vulnerabilities in the banking sector weigh on mortgages. In contrast, some Member States are experiencing very strong growth in mortgages, suggesting risks of a creditinduced housing boom in some of them. In 216, annual mortgage growth exceeded 1% in Romania and Slovakia and was above 5% in Belgium, the Czech 3 Source: European Commission. Page 6

7 Republic, Estonia, Lithuania, Luxembourg and Malta Social impact of housing developments Developments in housing prices have a deep impact on the affordability of housing, with significant social consequences, in particular in growing urban areas where demand outnumbers supply. This is demonstrated by high levels of homelessness, housing cost overburden, long social housing waiting lists, and difficult access for adults to enter the housing market. While the crisis already had a negative impact on new constructions and welfare provisions, the rise of poverty and migration further stretched demand. An affordable and good-quality home is crucial to a person's well-being and social participation. Access to affordable housing is becoming increasingly difficult in the EU. The crisis halted the construction of new social housing in many Member States and even led owners to sell existing stocks 5. In many countries, there is an acute lack of adequate and affordable social and private rental housing. Homeless people's access to housing is limited as allocation mechanisms for social housing rarely prioritise the homeless. Europeans consider that housing is too expensive. Less than a third (3%) of the people surveyed said that it is easy to find decent housing at a reasonable price in their area. overburden rate 7 captures the running costs associated with the use of housing, and not only its prices. It thus provides a more comprehensive assessment of housing conditions and potential vulnerabilities related to impoverishment and housing exclusion. In 215, the housing cost overburden rate was highest in Greece, at 4.9% of the population. This was considerably higher than in any other Member State. Bulgaria, Denmark, Germany and Romania had the nexthighest ratios, all at around 15%. In contrast, less than 5% of the population in Ireland (4.6%), Cyprus (3.9%), Malta (1.1%) and Finland (4.9%) lived in a household facing housing cost overburden. On average, 11.3% of the EU population was overburdened by housing costs, but for those at risk of poverty the figure was 39.3% and for the overall population it stood at 5.4%. In most countries, the housing cost overburden was significantly higher for tenants paying market rent (27% in the EU on average), than for owners with a mortgage or loan (6.7% in the EU on average). The rising house prices constitute an increasing obstacle to housing access 6. On average, roughly two in three (65%) 4 Source: European Central Bank /the-state-of-housing-in-the-eu-215 Housing Europe is the European Federation of Public, Cooperative & Social Housing. It exists since 1988 and is a network of 45 national and regional federations gathering 43. housing providers in 24 countries. Together they manage over 26 million homes, about 11 % of existing dwellings in Europe. 6 /publicopinion/archives/ebs/ebs_355_en.pdf 7 The proportion of households for whom total housing costs represent more than 4 % of disposable household income, net of housing allowances. Total housing costs consist of mortgage interest payments (for owners) or rental payment (for tenants) as well as other costs associated with the actual use of the accommodation, including the costs of utilities (water, electricity, gas and heating), insurance, repair and maintenance, taxes, service and charges. Eurostat: EU- SILC (survey ilc_lvho7a). Page 7

8 Figure 7 Housing cost overburden rate by poverty status 215 Source: Eurostat Households struggling with mortgage and rent arrears, high utility bills and overindebtedness and which are not receiving any support may in the end face repossession and eviction. For the period, a sharp rise in evictions has been registered in Bulgaria, Ireland, Spain, Cyprus, Latvia, Netherlands and the United Kingdom POLICY RESPONSES IN THE MEMBER STATES Public policies can fuel or mitigate the build-up of vulnerabilities in the housing sector. In a number of Member States, fiscal measures (e.g. deductibility of mortgage interest payments) reduce the marginal cost of acquiring housing, which - together with low interest rates and expectations of future house price increases - can increase the potential for speculative property investments by households and increases in household leverage. Furthermore, supply-side measures (e.g. to boost the construction of new properties amid high demand and rising property prices) can help improve the responsiveness of the house supply to potential price increases, limiting the risk of prices spiralling up. Given the strong 8 FEANTSA is the European Federation of National Organisations Working with the Homeless the-second-overview-of-housing-exclusion-ineurope-217. link between housing and credit, macroprudential policies, which seek to ensure financial stability, can have a strong impact on housing market developments Macro-prudential policies Several Member States have actively implemented macro-prudential measures to address vulnerabilities stemming from the real estate sector. Their primary objective is to dampen the inherent pro-cyclical dynamics between property lending and housing prices but they also strengthen the resilience of the banking and household sectors to financial shocks. Macro-prudential measures can take the form of bank-based measures or borrower-based measures. Measures targeting banks typically aim to ensure appropriate capital requirements and impose a capital buffer to this end. Such buffers can target economy-wide systemic risk but also be directly related to exposures to real estate. Other macro-prudential measures may, for instance, aim to align the risk-weights for specific items on banks' balance sheets - such as mortgage loans - with risk profiles. Specific capital requirement add-ons for mortgage portfolios have been introduced in a number of Member States (e.g. Belgium, Estonia, Austria, Slovakia and Sweden) with the aim of increasing the banking sector's resilience amid heightened real-estate-related vulnerabilities. Finland will introduce an institution-specific risk-weight floor for mortgages on houses on its territory as of Page 8

9 January 218. In Denmark, in March 217 the Systemic Risk Council recommended that variable rate housing loans in the Copenhagen and Aarhus regions be limited or have deferred amortisation if the borrower's total debt exceeds 4 per cent of income before tax. The evidence to date suggests that while bank-based measures have strengthened financial sector resilience in a number of Member States, increased capital requirements have been insufficient to stem soaring housing prices in some countries (e.g. Denmark, Luxembourg and Sweden) 9. In addition to bank-based instruments, national authorities have implemented national lending restrictions targeting borrowers. Among these, the most frequently used are limits to loan-to-value, loan-to-income or debt-to-income, and debt-service-to-income 1 ratios (Figure 7). Other instruments include loan maturity restrictions and requirements for amortisation. Borrower-based instruments directly target credit standards at their origin. Especially if implemented in a welldesigned package of mutually supporting actions, they have been empirically proven to be effective in restricting risky lending practices across a wide number of jurisdictions. Apart from reducing households' vulnerability to property-pricerelated shocks, they can also increase bank resilience. Their design is flexible and allows for parameters to be adjusted in order to influence housing and credit market conditions. The complementarity of borrower- and bank-based tools is particularly pertinent during the upswing of credit cycles. In times of upswing, bankbased tools could become less effective as capital ratios increase due to high bank profitability and buoyant asset prices. In such circumstances, measures which directly target lending standards at origin can reduce banks' incentives to engage in riskier (high-loan-to-value/high loan-toincome) lending. An in-depth comprehensive evaluation of the effectiveness of these measures across the EU is challenging due to considerable data gaps and also given the still fairly limited experience in applying macroprudential tools. Encouragingly, studies from some Member States which implemented macro-prudential measures following the recent financial crisis tend to confirm their effectiveness in mitigating real-estaterelated financial stability risks 11. However, in a number of Member States with real estate risks, authorities have faced difficulties in implementing macro-prudential instruments in a timely manner. This is due either to constraints in their own national legal frameworks or to institutional and governance arrangements for macroprudential policy. For example, in November 214 the Swedish supervision authority (FSA) announced a draft regulation on amortisation for new loans. The Administrative Court of Appeal in Jönköping issued an opinion that the FSA does not have the legal base to impose compulsory amortisation. The legislative initiative was then transferred to the government and had to pass through additional court reviews before it was deemed constitutional, paving the way for its implementation in the course of 216. In Belgium, the National Bank of Belgium (NBB) is the national competent and macro-prudential authority but only the federal government can activate loan-tovalue, loan-to-income and debt-service-toincome limits. In the spring of 217, the NBB proposed a buffer to cover riskweights for mortgage portfolios secured against residential property to follow-up on an earlier similar measure which had expired. However, so far the government has not taken up this proposal. Risks related to the housing market are followed up not only at national level, but also at EU level given their potentially systemic nature. The European Systemic Risk Board (ESRB) is the EU body in 9 European Systemic Risk Board (November 216), Vulnerabilities in the EU residential real estate sector. 1 Debt-service-to-income gauges the financial burden to households of paying the interest and principal on loans. 11 See Commission Staff Working Document SWD (216) 377 on the impact of capital and liquidity requirements; 12/216/EN/SWD F1-EN-MAIN- PART-1.PDF. Page 9

10 charge of monitoring macro-financial risks. It has a mandate to issue warnings when significant systemic risks are identified and when it is necessary to flag them. The ESRB has undertaken systematic work on analysing vulnerabilities in the EU residential real estate sector. In November 216, this culminated in the ESRB issuing warnings for eight EU Member States: Belgium, Denmark, Luxembourg, the Netherlands, Austria, Finland, Sweden, and the United Kingdom 12. The key vulnerabilities highlighted by the ESRB assessment are of a medium-term nature. They relate in particular to: high or rising household indebtedness, notably in Belgium, Denmark, the Netherlands, Finland, Sweden and the United Kingdom; overvaluation or strong price dynamics of residential real estate, e.g. in Belgium, Denmark, Luxemburg, Austria, Sweden and the United Kingdom; loosening lending conditions or the pace of bank lending, e.g. in Austria, Belgium and Luxemburg; and/or potential spill-over effects for other countries especially in the Nordic-Baltic region if risks were to materialise e.g. in Denmark, Finland and Sweden. As regards the remaining Member States, the ESRB either did not identify a build-up of material vulnerabilities relating to the residential real estate sector, or identified such vulnerabilities but considered that the policy stance was sufficient to address them. This was the case for Estonia and Slovakia. Since the ESRB warnings were issued, country-specific monitoring under the European Semester will follow real estate developments in the countries for which real-estate-related stability risks are a concern. The analysis of real estate markets in the EU is hampered by sizeable data gaps, however. Against this background, the ESRB has issued a specific recommendation on closing real estate data gaps 13. Figure 7 Macro-prudential measures targeting real estate risks on bank balance sheets Measure Adjusted risk-weights LTV limits LTI/DTI limits DSTI/PTI limits Stress test/sensitivity test/other prudent lending standards requirements Loan maturity Loan amortisation requirements EU Member State(s) using the measure BE, HR, IE, LU, MT, RO, SI, FI, SE, UK CZ, DK, EE, IE, CY, LV, LT, LU, HU, MT, NL, PL, RO, SK, FI, SE IE, PL, UK EE, CY, LT, HU, PL, RO, SK BE, CZ, DK, IE, CY, LU, RO, SK, UK EE, LT, NL, PL, RO, SK DK, NL, SK, SE Source: ESR Note: LTI Loan-to-value; LTI Loan-to-income; DTI Debt-to-income; DSTI Debt service-toincome; PTI Price-to-income. 12 ESRB press release of 28 November 216: html/pr en.html. 13 See: dations/216/esrb_216_14.en.pdf. Page 1

11 FR UK EL DK IT BE ES PL CY NL PT LV SE FI IE RO HU SI DE SK LT BG EE CZ AT LU HR MT 3.2. Tax policy Housing-related tax policy can shape incentives on the housing market. Tax measures can notably weigh on the preference between renting and buying. Two specific types of measure are important in this respect. First, the tax system may induce a debt-bias, thus favouring home ownership. Indeed, more than one-third of Member States have a tax system that subsidises mortgage debt financing, notably through mortgage interest deductibility. Steps to reduce such measures have been taken in recent years, as they often encourage house price hikes and favour high-income households. In particular, the countries where mortgage interest deductibility remains important, i.e. Belgium, Denmark, the Netherlands, and Sweden, are all characterised by high household indebtedness and have seen a significant increase in housing prices. In 217, the Council addressed countryspecific recommendations to the Netherlands and Sweden to reduce mortgage tax deductibility so as to reduce distortions on the housing market. Low property tax is also an implicit subsidy on home ownership which may result in sub-optimal resource allocation. The implicit rent that homeowners get from their investment is generally not taxed, except in a few Member States (notably the Netherlands, Luxembourg and Slovenia). This gives housing investment comparatively favourable tax treatment. Recurrent property tax on housing is somehow akin to a tax on implicit rent. However, the amounts collected are generally small (Figure 8), on average representing less than 1% of GDP and contributing at most 8% of total government revenue in the United Kingdom. In addition, the property value identified for tax purposes is often only loosely connected to market value, which creates distortions. Taxes on transactions are more closely related to a property's market value. Such taxes favour long-term ownership of dwellings, thus contributing to the reduction of speculative bubbles. However, they may also discourage potential buyers and sellers and thus reduce mobility. Figure 8 Recurrent taxes on immovable properties, in % of GDP, Recurrent taxes on immovable properties Source: European Commission Page 11

12 3.3. Addressing supply constraints Zoning restrictions Restrictions on the supply of housing, such as zoning or construction regulations, result in higher elasticity of prices to housing demand. In countries with significant demographic pressure, this contributed to an increase in prices, with differentiated impact on the various regions. Relaxing zoning regulations in a context of accelerating housing prices can lead to resources being diverted towards the construction sector. In 217, the Council addressed country-specific recommendations to Sweden and the United Kingdom recommending that they support investment and construction, notably by revising planning rules Rental market regulations The way the rental market operates can have deep implications for housing market dynamics. Because they are an essential source of affordable housing, rental markets are usually subject to strong public intervention with, on average, one third of tenants in the EU paying a reduced or no rent. Beyond such social measures, market rents themselves can be controlled, notably through caps on annual rent increases. The degree of rent control is particularly strong in Denmark, Austria, Sweden and Germany 14. On the one hand, a higher degree of rent control protects tenants from booms and busts in the housing market. On the other hand, it may result in rents being disconnected from house prices, thus depressing rental yields when house prices increase and discouraging the supply of rental housing, which contributes to a home ownership bias. Date: Bricongne J.-C. and A. Mordonu (forthcoming), 'Rental Market in a macroeconomic context', European Economy based on a methodology described in Cuerpo C., P. Pontuch and S. Kalantaryan (214), 'Rental market regulation in the European Union', European Economy. Page 12