This Selected Issues paper for Belgium evaluates whether price increases in Belgium are excessive. It assesses the household and bank balance sheets and their vulnerability to a slowdown in housing prices, and identifies differences in real estate markets between Belgium and other countries. The paper analyzes policy and institutional factors that may have contributed to housing price developments in Belgium. It identifies potential policies that are available to avert a buildup of pressures. The paper also describes the model setup and calibration of some crucial economic relations and parameters.

Abstract

This Selected Issues paper for Belgium evaluates whether price increases in Belgium are excessive. It assesses the household and bank balance sheets and their vulnerability to a slowdown in housing prices, and identifies differences in real estate markets between Belgium and other countries. The paper analyzes policy and institutional factors that may have contributed to housing price developments in Belgium. It identifies potential policies that are available to avert a buildup of pressures. The paper also describes the model setup and calibration of some crucial economic relations and parameters.

Executive Summary

The staff's analytical work associated with the 2005 Article IV Consultation of Belgium covers developments in the residential real estate market and the implications of labor and product market reforms. House prices have risen substantially in Belgium as elsewhere and residential construction has contributed significantly to growth. A slowdown in price growth is likely but will affect the economy mainly through the construction sector as wealth effects on consumption are largely absent and the financial system is robust. Labor and product market reforms could boost potential output significantly over the long run. The impact of labor market reforms is comparatively larger given the important remaining rigidities. Synchronization of reforms between product and labor markets would mitigate most transition costs of labor market reform. With Belgium a small player, increased output will easily be sold abroad, diminishing the need for international coordination of the timing of reforms.

How risky are real estate price developments in Belgium? Real estate prices have risen sharply in Belgium over the past few years, in between rates observed in the UK and France. There is clear evidence of a boom and the probability of a slowdown or reversal, at least in real terms, has increased. Aside from fundamentals and low interest rates, changes in borrowing behavior, lending practices and fiscal incentives have also contributed to price growth. Some of these factors have been temporary and interest rates appear to be beyond their historic lows. Nonetheless, with the financial system robust and no notable evidence of equity withdrawal, the macroeconomic implications of a price reversal are unlikely to be dramatic and will mostly affect the economy through the construction sector. Still, there is scope for further reducing the vulnerability of the economy by providing better education to households about the consequences of their choices; improving supervision; expanding mortgage backed security markets; removing policy distortions and improving the availability of statistical information on housing related indicators.

Macroeconomic impact of labor and product market reforms. Simulations with the Fund's Global Economic Model, (GEM) quantify the impact of reforms of labor, goods and services markets in Belgium. Overall, raising competition in these markets to the average observed in the non-euro area EU15 would boost output by 12 percentage points in the long run. With labor markets relatively more distorted in Belgium, about half of this increase would stem from labor market reform. Synchronization of reforms in labor and product markets eases transition costs to the point where there is no transitory loss in output. This is largely due to the fact that Belgium is very open and a price taker so that its reforms reduce inflation only modestly, thus raising real interest rates only moderately. In addition, output is easily sold abroad. Hence, while coordination of the timing of reforms with the rest of the euro area would further help the transition, going it alone will permit Belgium to reap the benefits with limited transition costs.

I. How Risky Are Real Estate Price Developments In Belgium?1

A. Introduction

1. As elsewhere in industrialized countries, residential real estate prices in Belgium have risen considerably in recent years. The general view is that housing prices in the 90's increased rapidly in several countries, driven by a low interest rate environment and increasingly flexible mortgage markets. Countries such as the UK, the Netherlands, Spain, Australia, Ireland, and the USA have experienced strong price increases in the late 90's and early 2000's and the recent slowdown in housing prices in the UK, the Netherlands, and Australia, and the accompanying slowing in consumption in these countries raised concerns about the impact of housing on the macro economy (Figure 1).

Figure 1.
Figure 1.

Private Consumption and Real Estate Prices

Citation: IMF Staff Country Reports 2006, 076; 10.5089/9781451803259.002.A001

Source: National authorities, WEO.

2. The purpose of this paper is to (i) evaluate whether price increases in Belgium are excessive; (ii) assess the household and bank balance sheets and their vulnerability to a slowdown in housing prices; (iii) identify differences in real estate markets between Belgium and other countries; (iv) analyze policy and institutional factors that may have contributed to housing price developments in Belgium and identify potential policies that are available to avert a build up of pressures.

B. House Price Developments in Belgium

3. House price increases in Belgium, even though strong and accelerating in 2004, have been somewhat less dynamic than in the most buoyant countries such as the Netherlands and the UK (Figure 2). Real estate cycles in countries such as the UK have been shorter and were defined by rapid acceleration followed by a rapid reversal in prices. Belgium experienced a more gradual growth in prices than in most countries. However, a feature that distinguishes prices in Belgium from other countries is the upward trend, which has lasted for two decades. After the housing trough in Belgium in 1985, which coincided with the housing trough in the Netherlands, these two countries experienced a prolonged period of moderate and consistent rises in real housing prices for about a decade. In the mid 90's however, prices in the Netherlands accelerated, whereas Belgium continued on the same moderate trajectory for another decade. The rapid price gains in the Netherlands of the late 90's cooled off in the recent years. In contrast, Belgium experienced a pick up in prices in recent years. In 2004, apartment prices on average increased by 14.5 percent and the price of medium and small sized homes increased by 6.8 percent, in nominal terms. While inflation over the same period was only 1.9 percent. The jump in apartment prices in 2004 may have been partially due to the fiscal amnesty operation in 2004, which provided incentive for households to repatriate their foreign financial assets, and were probably partly reinvested in real estate. Anecdotal evidence from various newspaper articles suggest that prices moderated in 2005. According to press reports, nominal prices for apartments and homes increased by 5.0 percent and 8.3 percent, respectively, while inflation over the same period is projected to be 2.4 percent.

Figure 2.
Figure 2.

Real Estate Price Developments

Citation: IMF Staff Country Reports 2006, 076; 10.5089/9781451803259.002.A001

Sources: Provided by national authorities and BIS.1/ BIS calculations based on national data, cumulative real growth rates (1970=0).

4. To measure the extent of price increases in the Belgian real estate sector we use the Helbling and Terrones methodology.2 Helbling and Terrones use a standard event study to analyze house prices in 14 industrialized countries from 1970 to 2002. They identify boom and bust cycles. They look at all episodes of real price increases and decreases and define a “bull market” as a market with increasing and a “bear market” as a market with decreasing prices. They define a “boom” as the top quartile of price changes in all bull markets and a “bust” as the bottom quartile of price changes in all bear markets. For example, as a result, a drop of 14 percent or more in real housing prices is defined as a bust. Price increases are measured as the peak-to peak increases or, alternatively, as the cumulative eight-quarter pre-peak increases in inflation-adjusted housing prices. The rationale for comparing peak-to-peak housing price increases is to get a measure of above trend growth rather than capture a possible correction of an earlier bust.

5. Using Helbling and Terrones' methodology, price increases in Belgium have been considerable (Table 1). Belgium experienced its last peak in the housing market in 1979 and last trough in 1985. If we use 2004 as a cut-off value, then the cumulative inflation adjusted housing price increase in Belgium since the previous peak has been 44 percent. This increase is larger than the 42 percent increase in the Netherlands, 41 percent in France, and 34 percent in the United States. However, it is less than the increase in the UK, Spain, and Australia, at 50 percent, 53 percent, and 61 percent, respectively. Further, the run up in prices in Belgium is larger than in previous housing price cycles. In the sample of Helbling and Terrones, average peak-to-peak increases during previous housing price booms have been 32.7 percent.

Table 1.

Inflation-Adjusted House Prices Using Helbling and Terrones as a Benchmark

(Percent cumulative change as of end-2004)

article image

Helbling and Terrones's sample (Average for 14 industrialized countries, 1970–2002) (Percent cumulative change up to peak)

article image
Sources: Author's calculations; Helbling (2005); and BIS paper 21.

Since prices in the Netherlands slowed down recently, the cumulative two year increase is calculated for 1998–2000 NB: Inflation adjusted real estate prices: BIS calculations using national data.

6. Comparing cumulative price increases from trough-to-peak, Belgium again has experienced considerable price growth. With an 86 percent increase since its last trough, Belgium leaves all other countries except the Netherlands behind. This increase is also a multiple of the historic sample of Helbling and Terrones, where the cumulative increase from trough-to-peak in all bull markets on average has been 11.3 percent and lasted for about 3 years.

7. However, a mitigating factor is that these price increases happened over a long time period. The housing cycle in Belgium has lasted for 19 years, which is comparable to the Netherlands (19 years) and Australia (12 years), but is significantly longer than other countries in the sample. For that reason a comparison of real estate prices in Belgium to the Netherlands, given also similar supply restrictions in both countries may be the most relevant comparison.

8. Price development in rentals in Belgium have been significantly lagging housing price increases. If housing prices are increasing more than rental property prices, this can potentially be a cause of concern, since distressed households would have difficulty to rent out property to cover mortgage payments. In Belgium, rental prices are controlled in the sense that during the length of a renting contract, rents can only be adapted for inflation. According to the HIPC index, the actual rent component has since 1995 increased on average by 2 percent per year in nominal terms, which is significantly below the house price inflation over the same period (Figure 3).

Figure 3.
Figure 3.

Belgium: Real Estate Prices and Construction Activity

Citation: IMF Staff Country Reports 2006, 076; 10.5089/9781451803259.002.A001

Sources: Provided by Belgian authorities and IMF.1/ Preliminary 2005 data, collected from a news report, not directly comparable to previous years.

C. What are the Drivers of Real Estate Prices in Belgium?

9. Changes in economic fundamentals explain to a large extent the prolonged increases in housing prices in Belgium. There is a vast empirical literature estimating the determinants of house prices3. Table 2 shows the regression results of one such study, for 18 industrialized countries, including Belgium, for the time period 1971–20034. According to these estimates, population growth, credit expansion, interest rates, and economic growth are the main determinants of real house prices. Further, house prices increases are serially correlated and show significant mean-reversion, when housing affordability deteriorates.

Table 2.

What Determines House Prices in Industrial Countries?

(Summary of empirical results, 1971–2003) (18 industrial countries)

article image
Source: IMF, WEO, September 2004.Note: The symbol * denotes significance at the 1 percent level. Significance is based on robust standard errors.

10. An analysis of the regression results suggests that demographics, particularly changes in interest rates have been the main drivers of price changes in Belgium (Table 3). Population growth increased from 0.16 percent per year to 0.38 percent per year from the 1970s to the 2000s. However, some of the increase in population is counterbalanced by a significant drop in growth. Belgium's real per capita GDP growth in the 1970s was 3.1 percent and dropped to 1.3 percent in the 2000s. However, the most important factor which has determined housing prices in Belgium, as in other countries, has been the dramatic drop in interest rates. As can be seen in the regression a one percent drop in interest rates can explain half a percentage increase in housing prices. This also indicates that in case of a interest rate increase housing prices would respond sensitively. Overall, using these regressions, over the time period of 1997 to 2003, only 8 percent of Belgium house price increases were not explained by fundamentals.

Table 3.

Developments in the Determinants of House Prices in Belgium

(In percent per year, average over decade, unless noted otherwise)

article image
Source: Staff calculations, WEO.

Ratio of Real house prices to (per capita) real income.

11. Despite scarcity of data to track investor behavior in Belgium, speculation doesn't seem to be a large concern. Housing transaction costs are very high compared to other countries, likely deterring speculation. Further, speculation in housing is less likely in a country where the majority of property is owner-occupied, as is the case in Belgium (Table 4). Nonetheless, in the UK where owner occupation is high speculation has been more of a problem.

Table 4.

Owner Occupation and Mortgage Debt

article image
Sources: European Mortgage Federation (EMF); Ahearne (2005); Belgian authorities; and Scanlon and Whitehead (2004b).

12. In the United Kingdom, which experienced a run up of residential property prices, speculation may have been one of the causes of rapid price increases. If investors are buying property with the sole purpose of reselling it for a profit (“flipping” property,) this can overheat markets. It is important to monitor investor behavior, since, if future price gain expectations prove to be wrong, investors are more likely to sell their investment than owner- occupiers. In the UK, the volume of “buy-to-let” lending as a percentage of total mortgage lending has increased from 3 percent of total lending in 1999 to around 7 percent in 2003. This ratio started to decline in the second half of 2004, coinciding with slowing appreciation of U.K. house prices.5

D. The Role of Mortgage Credit and the Institutional and Policy Environment

13. Developments in mortgage credit markets, the institutional environment, and policies influencing the real estate market play a role in price developments. First, if the Belgium market is vulnerable to a slowdown or a reversal of prices in real estate markets, one would expect to see some evidence of these vulnerabilities in the economy, either evidence of excessive risk taking in households and banks, or a strong dependence of economic growth on the housing market. Second, if the risk taking behavior of households or banks has changed it would be important to understand the drivers of this change. Institutional or policy changes could have influenced such behavior.

Has the risk preference of households and banks changed?

14. Belgian households while conservative compared to other countries, have been borrowing more at riskier terms. The average size of a mortgage loan for the purchase of an existing house has doubled since 1995 and reached almost €110,000 in 2004, a 9.6 percent increase relative to 2003. Further, households have been switching to flexible rate mortgages rapidly. The market share of flexible rate mortgages increased from 6 percent of total new mortgage loans in the period 1997–2001 to slightly more than 50 percent in 2004. This is a rapid shift in the risk profile and preferences of households and significantly increases the vulnerability of households' debt service levels to interest rate increases. In comparison, in the UK, a market where consumers are more familiar with flexible mortgage instruments6, 83 percent of total new mortgage loans in 1998 were variable rate loans and this ratio came down to 64 percent in 2004 (Figure 4).

Figure 4.
Figure 4.

Belgium: Key Mortgage Statistics

Citation: IMF Staff Country Reports 2006, 076; 10.5089/9781451803259.002.A001

Sources: Provided by national authorities and BIS.1/ Mortgage loans after deduction of deposits related to mortgage loans; consolidated basis.

15. However, several factors mitigate the risk exposure of Belgian households. First, flexible mortgage contracts in Belgium are significantly more conservative than in other countries, as a result of legal provisions that limit the frequency and the extent of rate revisions, and the possibility, of so called “accordion loans” which allow to extend the period of reimbursement to retain an unchanged monthly burden. Legal requirements are that rates may not be revised more than once a year, not by more than 1 percent per year on average in the first three years of the loan, and that a limit on the cumulated revision must be included in the loan contract with an upward limit not higher than the downward limit. The use of a cumulative cap of 3 percent is a standard practice. Further, Belgian households have less exposure to mortgage debt compared to other countries. For example, mortgage debt to GDP in 2004, in the Netherlands, the UK, and Australia was 105 percent, 75 percent, 74 percent, respectively, compared to 30 percent in Belgium and 26 percent in France (Table 4.). Mortgage debt to GDP rose to 32 percent, in Belgium, in 2005.

16. Nonetheless, in case of a sharp increase in interest rates the burden to households could increase significantly. For example, one can calculate the impact of a rate increase in the theoretical but standard case of a variable rate mortgage with a cumulative cap of 1, 2, and 3 percent respectively on the up- or downward adjustment that can take place in the first, second and subsequent years of a loan. If one assumes, the worst case scenario, an upward revision of 1 percent in the first three years of the loan, capped at a cumulative 3 percent for the rest of the loan, the debt burden would increase by 27 percent from the fourth year onward. Given the actual levels of the 20-year fixed rate and variable rate with a 3 percent cap, such a rate revision would bring the monthly debt burden of the variable rate loan from 8 percent under (in the first year) to 17 percent above (from the fourth year onward) the burden of the fixed rate loan. Further, since the tax advantage of having a mortgage loan is front loaded the burden in later years may increase even further.

17. While households have been switching to riskier borrowing practices the exposure of Belgium banks to mortgage related risks has been increasing rapidly. Mortgage loans in the Belgium banking system (including nondomestic loans) grew from €80 billion in 2000 to €147 billions as of September 2005. Further, the average loan-to-value ratio of new mortgage loans, a good indicator of the aggregate risk banks are facing, increased to 89 percent in 2004. In addition, it is important to note that this ratio would jump significantly if housing prices were to fall. As a comparison, in the UK and USA average LTV ratios of new loans are currently 75 and 80 percent respectively. Under the affordability test normally applied by Belgian banks, borrowers may devote no more than 33 percent of their income to housing costs. Loan-to-value ratios are not typically used; there are no regulations limiting the LTV ratio; it is relatively easy for the borrower to get more by adding a personal loan to the mortgage, thus borrowing over 100 percent of the assessed value.7

18. The increase in mortgage loans was partly driven by increased competition among banks to attract customers (Figure 5). Similar to developments in other European countries, Belgium experienced a significant loan profile shift resulting from reduced recourse to bank credit by corporations. As a consequence, the percentage of loans to the corporate sector declined from 47.5 to 37.8 percent from 1999 to September 2005. Conversely, the percentage of loans to households increased from 45.3 to 52.6 over the same period. The pressure on banks due to the loss of business customers also becomes evident in the latest Survey of Bank Lending Standards. According to this survey, banks in the United States and Europe reported a significant easing of credit standards over the last two years. Banks listed concerns about competition from other sources of business credit as their primary reason for easing standards.

Figure 5.
Figure 5.

Belgium: Bank Lending

Citation: IMF Staff Country Reports 2006, 076; 10.5089/9781451803259.002.A001

Source: IMF, Bank Lending Survey.

19. Heightened competition can also be seen in the decline in banks' commercial margin on mortgage loans and an easing of lenders' credit standards for housing loans.8 From an average of 1.5 percent in 2003, the margin relative to a one-year government bond declined to 1.0 percent in early 2005. Further, according to the Belgian results of the Euro system's Bank Lending Survey, an easing of lenders' credit standards for housing loans took place.

20. Still, risks to the Belgium banking system are limited compared to other countries, because of the relative low level of mortgage loans in the aggregate and banks' strong financial position. Residential mortgage debt, extended by Belgian banks to residents and nonresidents, as a percent of GDP increased from 32 percent of GDP in 1999 to 50 percent in 2005, but is still lower than in other countries. Further, Belgium banks have a strong financial footing and a lower exposure to mortgage risks than other international banks (Table 5). In some cases, property lending exceeds both the market capitalization and several times its equity capital of some banks. Nonetheless, the conventional market view is that most of these banks have ample capital to withstand the risks they are facing. In addition, results of stress tests on the Belgium banking, conducted in the context of the 2005 FSAP, indicate considerable resilience against shocks, even if the system undergoes a combined macroeconomic shock.

Table 5.

Selected Large Mortgage Lenders

article image
Sources: Aherarne and others, (2005); and individual bank annual reports; staff estimates.

Mortgage loans based on staff estimates.

Moody's Bank Financial Strength Rating.

21. Further, risks to banks are increasingly shifted to investors due to the increased utilization of various instruments such as mortgage backed securities. However, this market is still embryonic in Belgium, which relies on savings deposits for funding. Currently, mortgage bonds are the second most important funding instrument for mortgage lenders in Europe after retail deposits; and according to the European Mortgage Federation (EMF), mortgage bonds fund 15 percent of the European Capital Market. In case of mortgage bonds, the issuing institution still carries the credit risk associated with the mortgage; the advantage of mortgage backed securities is that this risk is transferred to the investor. According to the EMF, for 2004, an overall growth of up to 25 percent is expected in the European covered bond market, due to the increased issuance of structured covered bonds and the emergence of covered bond legislation in Portugal and Belgium. However, in this highly dynamic environment, the recent Global Financial Stability Report9 warns, that the recent increased risks in mortgage markets may lead to a correction of tight spreads in mortgage-backed securities markets.

What has been the impact of institutional and policy changes on risk taking behavior and prices?

22. While it is hard to asses the direct impact of fiscal policy on real estate prices, recently introduced tax and regulatory changes may have fueled housing prices and encouraged increased risk taking. Some of these changes such as the reduction in transaction costs reduce inefficiency and are welcome. However, some may have encouraged holding high levels of mortgage debt, and therefore led to higher LTV values.

23. Recent changes in regulation shifted some of the risks from banks and households to the government, at no cost to the households and banks. Since the late 1990s, anyone making use of subsidies for house purchases has also received government mortgage insurance, which insures the lender against non-payment of the mortgage. In addition, the Flemish government introduced in 1998 an insurance scheme for all others, with non-subsidized mortgages. This insurance works for any household in case of income loss. When concluding a mortgage, a household can get a free insurance from the Flemish government. The provider of that insurance is a large public insurance company that won the contract from the government and the cost of the insurance is paid by the Flemish government. From the moment a mortgage is concluded, the insurance lasts for 10 years. If during that period a household becomes unemployed involuntarily, it can get, after a waiting period of six months, a monthly contribution to the repayment of the mortgage, for a maximum of three years. This contribution does not cover the whole sum and cannot exceed 496 euros per month, in addition the household still needs to pay a minimum of 248 euros.10 While it is unlikely that this change has led to a drastic shift in risk taking among households, it does reduce the overall risk of the mortgage portfolio for banks.

24. These changes may have contributed somewhat to the attractiveness of mortgage loans, despite very high mortgage enforcement costs in Belgium. The usual time required for the distribution of the proceeds to creditors in Belgium is 18 months and the administrative costs amount to approximately 19 percent of the property value (Table 6). In comparison, the time required in the Netherlands is 6 months and the administrative costs are approximately 3 percent. Whereas in the UK, the time required is 8–12 months and the administrative costs are 2.6–7 percent.

Table 6.

Time Required and Cost of Mortgage Enforcement Procedures

article image
Sources: OECD Economic Outlook, No:75; The Economist; Ahearne and others (2005).

Total time from right of execution (in the countries where the mortgage must be given executory power by a court) to the distribution of proceeds to creditors.

Costs usually include both fixed and variable components; calculated for a property value of €100,000; does not include lost interest during the procedure.

25. There were some welcome reductions in transaction costs, which may have contributed to the attractiveness of real estate in recent years. Housing related transaction costs are significantly higher in Belgium than in other European countries (Table 6.). A household pays approximately 18 percent of the house price in transaction costs: 12.5 percent in registration tax, a 2.5 percent notary fee, and a 3 percent real estate broker fee. However, 18 percent is an upper limit. Not all real estate transactions are brokered by real estate agents and registration tax is lower for modest houses (6 percent). In 2002, the region of Flanders reduced registration taxes to 10 percent (5 percent for modest houses) and introduced the portability of transaction taxes and a tax exemption on the first 12,500 Euro of the acquisition price. In the Brussels region, a tax exemption on the first €45,000 or €60,000 of the acquisition price was introduced in 2003. These changes, may explain some of the increases in house prices for apartments and small houses, particularly in the Flanders region, since young households can enter the market by buying an apartment or a small house to resell it later, rather than wait until they have sufficient finances to buy a house.

26. The tax amnesty of 2004 may explain a significant part of the run up in prices in 2004. Prices of apartments did not evolve in the same way in different regions of the country. While the nominal rate of increase stabilized at a high level in Brussels (around 15 percent), it increased to respectively 14 percent and 12 percent in Flanders an Wallonia (from around 8 percent in 2003). The highest jump in prices was recorded in the coastal area, where prices rose by 17 percent compared to only 4 percent in 2003. The high rate of increase in the coastal area points to the fact that a considerable part of the country-wide increase in the price of apartments is due to an increased demand for second residences. This likely may be a one-off development in response to the fiscal amnesty operation in 2004, since households most likely have been reinvesting some of the repatriated capital in real estate.

27. Further, while several European countries reduced interest deductibility of mortgage loans, recent changes in Belgium increased the attractiveness of debt and may have contributed to the high loan-to-value ratios. For example, in the UK the tax deductibility of interest on mortgage loans was progressively decreased during the 1990s and finally eliminated in 2000. In Denmark, the rate of tax deductibility was reduced in 1998. In Belgium, mortgage interest payments are tax deductible and the new fiscal regime for mortgage related expenses, which entered into force in 2005, frontloads the fiscal advantage of a mortgage loan to the first 10 years of the loan. The new legislation introduced a unique amount—€2,480 per year in the first 10 years and €1,860 afterwards—that can be deducted from the borrower's taxable income. These fiscal advantages significantly reduce the effective interest rate that has to be paid on mortgage loans. In case of a loan of €100,000 for instance, with a maturity of 20 years and a yearly variable interest rate of 3.3 percent, the effective rate in the first year of the loan falls to 1.3 percent, if the amount of €2,480 can be deducted by two persons, which both have a marginal tax rate of 40 percent. In case of a loan with a fixed interest rate of 4.7 percent, the effective rate would amount to 2.7 percent in the first year.11

Would a slowdown in the housing market impact economic growth?

28. With interest rates rising, a slowdown or a reversal in real estate prices is likely. The unprecedented low-interest-rate environment, the development of new mortgage products, increased price competition among banks to attract mortgage loans, and refinancing activity (Table 7) has been freeing cash flow for households and been allowing them to afford large amounts of mortgage debt. Historically, in industrialized countries, if a boom market was followed by a bust, real housing prices contracted for 25 to 30 percent from peak to trough, lasting on average 3–5 years.12 Even in the absence of a bust, if a boom market was followed by a bear market, it led to a contraction in real housing prices of 6 percent on average, lasting for about 1 year. A slowdown or a reversal in prices of these magnitudes can affect the economy through a number of channels: (i) wealth-savings-consumption, (ii) residential construction, and (iii) purchases of consumer durables associated with housing.

Table 7.

Belgium: Breakdown of New Mortgage Loans

(millions of euros)

article image
Source: Union Professionnelle du Crédit - UPC.

29. The economic slowdown that follows a housing price bust can be considerable. Historically, even though housing price busts have been associated with more modest price decline than equity price busts, the effect of a housing price bust on output has been twice as large than an equity price bust. Evidence from event studies on industrial countries suggests that, on average, the output level three years after the beginning of a bust was about 8 percent below the level that would have prevailed in its absence.13 Moreover, the slowdown after a housing price bust lasted on average about twice as long as a stock market bust.

30. Historically, housing prices in all countries have been cyclical, mainly driven by interest rate fluctuations.14 Further, housing price booms have on average lasted for only about 4 years. The fact that, since the last downturn in prices in Belgium in the mid-80s, price increases have been consistently strong, implies that price pressures may be building up. In the sample of Helbling and Terrones, the implied probability of a housing price boom being followed by a bust is about 40 percent. The current price increases in Belgium have surpassed previous cycles in housing prices. A large part of that increase, but not all of it can be attributed to changes in fundamentals, suggesting that price increases may slow down as appears to be happening in 2005.

31. In the late 1970's Belgium experienced a significant run up in real estate prices followed by a severe price correction, starting in 1979. The housing market reached its trough in 1985, falling by over 50 percent in real terms. During this time Belgium experienced a significant slowdown in growth, and a severe contraction in fixed investment (Figure 3). Residential real estate investment fell from €9.7 billions in 1980 to €4.6 billions in 1983 and did not recover to its pre-crisis levels until 1990. During the crisis, the contribution of residential real estate construction to GDP growth remained negative for 3 consecutive years and reached -2.4 percent at its worst (1981).

32. While the situation today is not as in the 1970's, a slow down in the housing market in response to higher interest rates would nonetheless adversely impact macroeconomic growth. The shock in the late 1970s did not originate in the housing market; but rather was reflected in it. Further, residential construction in Belgium now accounts for approximately 4 percent of GDP, whereas in 1980 it was 6 percent of GDP.

E. Conclusions

33. Fundamentals, in particular low interest rates, a shift of bank lending toward the mortgage market, and changes in fiscal incentives have contributed to buoyant residential property price developments in Belgium. On standard analysis, price increases are considerable, raising the probability of a slowdown or even a temporary reversal, at least in real terms. A slowing or a reversal in real estate price increases as a consequence of interest rate increases has the potential to weaken demand through a variety of channels but its impact is difficult to quantify.

34. However, price developments in Belgium have been more modest compared to other countries and lasted for a longer time period and overall risks seem to be confined. The pricing cycle in Belgium has showed similarities with the Netherlands, where prices also have been growing for almost two decades. However, while prices in Netherlands accelerated rapidly in the mid 90's Belgian real estate prices continued on the same trajectory during that time period. More recently, real estate prices in the Netherlands slowed down, while prices in Belgium have accelerated, at least through 2004. However, vulnerability is limited as mortgage debt of Belgian households is comparatively low.

35. Nonetheless, it will be important for policy makers and supervisors to be vigilant of price developments in real estate markets. In absence of monetary instruments, fiscal policy and supervision may become important tools to help reduce pressures. Timing of fiscal polices may have had an impact on recent price developments and it would be prudent to factor in price impacts of policy changes while markets are buoyant. Further, supervisors should consider following lending practices of banks closely to preempt a possible build up of vulnerabilities.

36. The development of new mortgage products in Belgium and the increased accessibility of loans are welcome from an efficiency perspective, but associated risks need to be well managed. Banks overall seem on a strong financial footing, however, a change in interest rates is likely to impact economic growth. Particularly, households with variable rate loans and the marginal households that were just able to afford a house using the front loaded fiscal benefits, may experience a cash flow squeeze when interest rates go up.

37. Some of the recently introduced policies encouraged households and banks to take risks. Interest deductibility and frontloading the fiscal advantage of a mortgage debt have encouraged households to leverage themselves. Further, government financed insurance schemes for mortgage debt have shifted some of the risk from banks to the government, and thus the taxpayer. This is likely to have adversely affected the incentives for banks to scrutinize risk contracts. Despite these various new policies, housing affordability has deteriorated.

38. Drawing on experience from other countries, the vulnerability of the economy and financial system to developments in real estate markets can be reduced along a number of dimensions:

  • Education: It is essential that households are adequately informed and educated about the risks of real estate investment and the potential future costs of mortgage loans. With the previous real estate bust a memory of the distant past it is likely that households are not aware of the possibility of a downturn in housing prices. Further, it is possible that households are excessively concentrated on their current budget and the affordability of a loan in the first year, rather than a long term view on the mortgage product that they purchase.

  • Supervision: Mortgage lenders should not only concentrate on what a household needs to pay as an effective mortgage payment in the beginning of a loan, but also calculate the estimated future burden of a loan—once interest rates go up and once the frontloaded fiscal advantages wear off. Banks could be asked to monitor LTV ratios of new mortgages more closely and eventually a ceiling on these ratios could be introduced.

  • Market efficiency: Risks to banks should be reduced by developing complete mortgage markets rather than shifting risks from banks to the government. Even though a politically sensitive topic, the improvement of mortgage enforcement could further reduce the costs and risk associated with mortgage lending.

  • Policy distortions and contingency plans: Going forward, the policy discussion should concentrate on reducing transaction costs, rather than distorting prices by giving incentives to households to acquire debt and leverage themselves excessively. In absence of an independent monetary policy, Belgium could prepare a contingency plan, using various tax policy changes to avoid price booms and busts. The experience of countries such as Denmark, which reduced interest deductibility of mortgage loans, suggests that housing markets respond sensitively to policy changes. Timing of policy changes also matters. From this perspective, some of the recently introduced policy may have been ill-timed, fueling housing prices when they were already increasing rapidly.

  • Information: to enhance market functioning and supervise mortgage related risks more effectively, the collection of housing related statistics could be improved. In particular detailed data on the stock of mortgages, such as the percentage of flexible loans in the stock should be collected. Data on investor behavior, in particular turnover of housing and buy-to-let statistics should be collected systematically. Further, more detail on housing related statistics such as rental prices, LTV ratios (by income segment), and bank exposure could be made available more frequently and in timely manner.

References

  • Ahearne, Alan G. and others, (2005), “House Prices and Monetary Policy: A Cross-Country Study,” International Finance Discussion Papers No. 841, Board of Governors of the Federal Reserve System.

    • Search Google Scholar
    • Export Citation
  • CML Research (2004), “International trends in housing tenure and mortgage finance,” Council of Mortgage Lenders.

  • GFSR (2005), Global Financial Stability Report, June 2005, IMF.

  • Helbling, Thomas, (2005), “Housing price bubbles a tale based on Booms and Busts”, in BIS Paper No. 21, “Real Estate Indicators and Financial Stability”, BIS.

    • Search Google Scholar
    • Export Citation
  • Lamont, Owen and Jeremy Stein, 1999, “Leverage and House Price Dynamics in U.S. cities.” Rand Journal of Economics, Vol. 30 (Autumn), pp. 498 –514

    • Search Google Scholar
    • Export Citation
  • National Bank of Belgium (2005), Financial Stability Review.

  • OECD Economic Outlook (2004), “Housing Markets, Wealth, and the Business Cycle,” OECD Economic Outlook. No 75.

  • IMF, (2005), “United States—Selected Issues,” IMF Country Report No. 05/87

  • IMF (2005), WEO, September 2005, Chapter I.

  • IMF (2004), WEO, September 2004, Chapter II.

  • IMF (2003), WEO, April 2003, Chapter II.

  • Scanlon, Kathleen and Christine Whitehead, (2004a). “Housing tenure and mortgage systems: How the UK compares,” Council of Mortgage lenders, Winter 2004

    • Search Google Scholar
    • Export Citation
  • Scanlon, Kathleen and Christine, (2004b). “International trends in housing tenure and mortgage finance,” Council of Mortgage lenders.

    • Search Google Scholar
    • Export Citation
1

Prepared by Sibel Yelten

3

WEO, September 2004, Box 2.1 and Lamont and Stein (1999)

4

WEO, September 2004, Box 2.1

6

According to the completeness index of the European Mortgage Federation the UK receives a perfect score (100 percent) for information and advice on mortgage related products. Belgium is not included in the study, however the same index is 70 and 80 in France and Netherlands, respectively.

7

K. Scanlon and C. Whitehead (2004a) and anecdotal information provided by banks.

8

National Bank of Belgium (2005), Financial Stability Review.

9

Global Financial Stability Report (September 2005), IMF.

11

Calculations by Belgium authorities.

12

See also World Economic Outlook, September 2005, Chapter I; World Economic Outlook, April 2003, Chapter II; and BIS paper number 21 (2005), for details.

14

In fact, housing prices in general are found to be pro-cyclical, Ahearne, and others (2005).

References

  • Alessandria, G., Delacroix, A. 2004, Trade and the (Dis) Incentive to Reform Labor Markets: The Case of Reform in the European Union, FRB of Philadelphia WP 04–18.

    • Search Google Scholar
    • Export Citation
  • Avonds, L., 2005, Een vergelijkende analyse van de Input-Outputtabellen van 1995 en 2000, Federal Planning Bureau, Working paper 4–05.

    • Search Google Scholar
    • Export Citation
  • Bayoumi, T. et. alii., 2004, GEM: A New International Macroeconomic Model, IMF Occasional Paper 239.

  • Bayoumi, T., Laxton, D. and Pesenti, P. 2004, Benefits and Spillovers of Greater Competition in Europe: A Macroeconomic Assessment, ECB Working Paper 341.

    • Search Google Scholar
    • Export Citation
  • Biatour, B, Fiers, J. Gilis, S. Kegels, C. Thiery, F. 2005, European R&D Strategy: Impact and Feasibility Study for Belgium, Federal Planning Bureau Working Paper 3–05.

    • Search Google Scholar
    • Export Citation
  • Bean, C.R., 1998, The Interaction of Aggregate-demand Policies and Labour Market Reform, Swedish Economic Policy Review, pp. 353 –88.

    • Search Google Scholar
    • Export Citation
  • Blanchard, O., and Giavazzi, F. 2003, Macroeconomic Effects of Regulation and Deregulation in Goods and Labor Markets, The Quarterly Journal of Economics, pp. 879 –907.

    • Search Google Scholar
    • Export Citation
  • Bowman, D., 2003, Market Power and Inflation, Board of Governors of the FRS, International Finance Discussion Papers 783.

  • Clark, A.E., and Senik, C. 2004, The (Unexpected) Structure of ‘Rents’ on the French and British Labour Markets, IZA Discussion Paper 1438.

    • Search Google Scholar
    • Export Citation
  • Conway, P, Janod, V. and Nicoletti, G. 2005, Product Market Regulation in OECD Countries: 1998 to 2003, OECD, Economics Department Working Paper 419.

    • Search Google Scholar
    • Export Citation
  • Crepon, B., Desplatz, R. and Mairesse, J. 2002, Price-Cost Margins and Rent Sharing: Evidence from a Panel of French Manufacturing Firms, mimeo.

    • Search Google Scholar
    • Export Citation
  • Dobbelaere, S, 2005, Joint Estimation of Price-Cost Margins and Union Bargaining Power for Belgian Manufacturing, IZA Discussion Paper 1466.

    • Search Google Scholar
    • Export Citation
  • Duval, R. and Elmeskov, J. 2005, The Effects of EMU on Structural Reforms in Labour and Product Markets, paper prepared for the conference on What effects is EMU having on the euro area and its member countries?” organized by the ECB.

    • Search Google Scholar
    • Export Citation
  • Estevão, M., 2005, Product Market Regulation and the Benefits of Wage Moderation, IMF Working Paper 05/191.

  • Gali, J., Gertler, M. and Lopez-Salido, J. D. 2002, Markups, Gaps, and the Welfare Costs of Business Fluctuations, NBER WP 8850.

  • Hunt, B., 2004, Denmark Selected Issues, IMF.

  • Jean, S., and Nicoletti, G. 2002, Product Market Regulation and Wage Premia In Europe and North America: An Empirical Investigation, OECD, Economics Department Working Paper 419.

    • Search Google Scholar
    • Export Citation
  • Kahn, H., 2004, Price-setting Behaviour, Competition, and Markup Shocks in the New Keynesian Model, Bank of England Working Paper 240.

    • Search Google Scholar
    • Export Citation
  • Konings, J., Cayseele, P. Van and Warzynski, F. 2001, The Dynamics of Industrial Markups in Two Small Open Economies: Does National Competition Policy Matter?, International Journal of Industrial Organization, pp 841 –59.

    • Search Google Scholar
    • Export Citation
  • Nicoletti, G., Bassanini, A. Ernst, E. Jean, S. Santiago, P. and Swaim, P. 2001, Product and Labor Markets Interaction in OECD Countries, OECD, Economics Department Working Paper 312.

    • Search Google Scholar
    • Export Citation
  • Nicoletti, G., and Scarpetta, S. 2003, Regulation, Productivity, and Growth, World Bank, Policy Research Working Paper 2944.

  • Oliveira Martins, J., Scarpetta, S. and Pilat, D. 1996, Markup Ratios in Manufacturing Industries -Estimates for 14 OECD Countries, OECD, Economics Department Working Paper 162.

    • Search Google Scholar
    • Export Citation
  • Oliveira Martins, J., and Scarpetta, S. 1999, The Levels And Cyclical Behaviour Of Markups Across Countries And Market Structures, OECD, Economics Department Working Paper 213.

    • Search Google Scholar
    • Export Citation
  • Orphanides, A., 2003, Historical Monetary Analysis and the Taylor Rule. Journal of Monetary Economics, pp. 983 –1022.

  • Saint-Paul, G., 2004, Did European Labour Markets Become More Competitive in the 1990s? Evidence From Estimated Worker Rents, CEPR Discussion Paper 4327.

    • Search Google Scholar
    • Export Citation
  • Schule, W., 2005, Estimating the Macroeconomic Effects of Higher Competition in Labor and Product Markets, in IMF-France-Selected Issues (forthcoming)

    • Search Google Scholar
    • Export Citation
  • Zhou, J., 2005, Employment Effects of Reductions in Labor Taxes in a Wage-Bargaining Model, IMF Country Report 05/76.

15

Prepared by Luc Everaert and Werner Schule.

16

Bayoumi and others (2004) estimate spillovers from the euro area to the rest of the world, which provide a benchmark for the degree of underestimation of spillover effects due to the reduced-openness assumption.

17

Trade flows are based on the UN COMTRADE statistics. For the sake of simplicity, the commodities sector was excluded from this version of GEM.

18

This estimate is approximate. It was guided by the observation that according to input-output tables about 45 percent of final domestic demand consists of imports (Avonds, 2005, Table 47).

19

These items are excluded since they do not represent a true claim on resources by the public sector.

20

See Bayoumi and others (2004) for the calibration details. This paper follows in their footsteps.

21

Estimates of sacrifice ratios, the cumulated output costs of reducing inflation permanently by 1 percentage point, are often higher than these values, but they might reflect slow learning by central banks during the transition from a high to a low inflation environment.

22

For an extensive discussion of the sensitivity of results with respect to these parameters see Bayoumi and others (2004).

23

Households have good access to credit both through consumer and mortgage credit and have built up large net asset positions from which they can draw to bridge temporary liquidity constraints.

24

Tax rates on capital income are fixed at 10 percent. More sophisticated scenarios can be run by changing expenditures and taxes discretionary, alleviating the burden that falls on labor taxation.

25

At this level, the cost of aging is covered by savings of the interest bill, and the public debt-to-GDP ratio is stable. In the model, a higher steady state level of debt would require somewhat higher taxes and result in lower economic efficiency.

26

See also WEO April 2005, Annex 3.3 to Chapter III.

27

This simple formula ignores adjustment costs. An elasticity of substitution of 5 translates into a markup of 1.25 (25 percent). The markup goes to zero only if all products are perfect substitutes.

28

More precisely the OECD measure covers trade and investment restriction, regulatory barriers, discriminatory procedures or ownership barriers; licensing and permits, administrative, sector-specific, and legal burdens, antitrust exemptions; and state influence measured by the size and scope of the public enterprise sector, direct controls over business, and price controls or restrictions on establishment. See Conway and others (2005).

30

Gali, Gertler, and Lopez-Salido (2002) developed a related model that explains the bulk of output gap fluctuations with price and wage markups (NBER 8850).

32

Demand reacts sluggishly because of real and nominal rigidities in the model while the gradual implementation of reforms makes sure that potential output does not jump; this time pattern is entirely plausible for Belgium and the EU as a whole.

33

More competition in the tradables sector lowers tradables prices vis-à-vis nontradables prices and therefore represents a real appreciation of the home currency.

34

Indexation is to the so-called health index, which excludes prices of motor fuels, tobacco, and alcohol.

35

Raising competition represents an asymmetric positive supply shock, when implemented in one country only.

36

Modifying some key calibration parameters does not alter the thrust of the conclusions but widens the range of outcomes. Under the alternatives considered, the effect of stand-alone reforms would range from 7 to 13.3 percent of GDP.

37

Uncertainty leads to caution, including on the side of monetary policy-makers. As a result, monetary conditions may be less than fully accommodative, even in the case of synchronized euro area-wide reform.

Belgium: Selected Issues
Author: International Monetary Fund