This paper examines the selected issues related to the economy of Denmark: divergence in house prices, house prices in Denmark's cities, macroprudential policies, and product market reform and firm productivity. Recent house price developments in Denmark have been characterized by a growing divergence between different parts of the country, with big cities experiencing much more rapid price increases than other parts. House price booms and busts in Denmark, like in many other countries, are a big-city phenomenon. Macroprudential policies can help contain risks for households, the financial system, and the broader economy, but they should be carefully calibrated to avoid an undue drag on growth.

Abstract

This paper examines the selected issues related to the economy of Denmark: divergence in house prices, house prices in Denmark's cities, macroprudential policies, and product market reform and firm productivity. Recent house price developments in Denmark have been characterized by a growing divergence between different parts of the country, with big cities experiencing much more rapid price increases than other parts. House price booms and busts in Denmark, like in many other countries, are a big-city phenomenon. Macroprudential policies can help contain risks for households, the financial system, and the broader economy, but they should be carefully calibrated to avoid an undue drag on growth.

Macroprudential Policies in Denmark1

Despite the moderate recovery, house prices have risen strongly. Given high household debt and some specific features of Danish mortgages (in particular, a high share of variable rates and interest only loans), financial stability concerns could soon be emerging if rapid house price growth continues. Macroprudential policies can help contain risks for households, the financial system, and the broader economy, but they should be carefully calibrated to avoid an undue drag on growth. In this paper, we use a general equilibrium framework to study the impact of different macroprudential tools. Our tentative results suggest that macroprudential policies generally have limited impact on the real economy, while the effect on household debt can be substantial.

A. Introduction

1. The Danish economy is gradually recovering from the global financial crisis. Denmark was hard hit by the crisis, which coincided with the puncture of a local housing bubble during which average house prices fell by 20 percent between 2007 and 2009. As a result, residential investment has stagnated and a significant share of households is still re-building their balance sheets. In this environment, growth has been relatively weak in recent years though a modest recovery has been underway since 2014.

A03ufig1

House Prices

(Index, 1992Q1 = 100)

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A003

Sources: Danmarks Nationalbank.

2. Yet, house prices have risen strongly in recent years and household debt remains at a high level. The continuing decline of interest rates, demographic factors, relatively healthy household disposable income growth, and—quite possibly—self-fulfilling price dynamics have driven up house prices again. Indeed, the price of owner-occupied apartments has almost reached its precrisis peak in nominal terms, though the price for one-family houses still remains well below its peak. Household debt has also been gradually increasing in nominal terms. And although the household debt-to-income ratio declined about 60 percentage points over 2008–14, it has started to rise again in 2015Q4.

A03ufig2

Lending to Households for Housing Purposes

(bn DKK)

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A003

Sources: Danmarks Nationalbank.

3. Rising house prices and high household debt are creating financial stability concerns. Banks and mortgage institutions have substantial direct exposures to the housing market. Their total lending to households for housing purposes amounts to approximately 90 percent of GDP (just under DKr 1,750 billion). Moreover, the change in housing wealth has a considerable impact on private consumption, as indicated by Denmarks’s own experience. The last crisis also underscored the positive correlation between household leverage and sensitivity to shocks (Andersson, 2014). The rapid increase in house prices is again opening a positive valuation gap, especially in major cities, increasing the risk of another correction in the housing market further down the road. Importantly, credit booms and high level of household debt are associated with a higher probability of crises, deeper recession, and slower recovery (Schularick and Taylor, 2012; and Jorda and others, 2013) When this risk materializes, the impact on the macroeconomy could be even higher than the previous crisis since the recovery has been a relatively weak one.

A03ufig3

Households’ Debt-to-Disposable Income

(percent)

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A003

Sources: Riksbank.
A03ufig4

Relationship Between Debt-to-Income Ratio and Consumption Growth between 2007 and 2012 1/

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A003

Sources: Riksbank.1/ Adjusted consumption growth has been calculated following Floden (2014).

4. In view of the emerging risks, macroprudential policies need to play a primary role in safeguarding financial stability. Monetary policy is fully dedicated to maintaining the peg with the euro (which has the side effect of a very low imported interest rate that likely adds fuel to house price increases), while fiscal policy adjustment should be calibrated to manage the broader economic cycle (which does not call for any sharp tightening at present). This leaves macroprudential policy as the main instrument to preserve financial stability, especially over the short run. A more stable housing market (i.e., less volatile house prices and household debt) would also contribute to smoother economic cycles, supporting other policies.

5. However, macroprudential policies need to be carefully calibrated to avoid putting a drag on growth. Any macroprudential measures that effectively impact credit and house price growth would inevitability also affect the macroeconomy. As the Danish economy is still recovering from the last crisis, new measures must be carefully calibrated to balance the benefits of reduced financial stability risks against any negative short-run impact on growth.

6. In this paper, we analyze the impact of macroprudential policies in a general equilibrium framework. This allows us to assess the impact of these policies not only on household debt, but also on broader macroeconomic variables, such as consumption.

7. The paper is organized as follows: Section B briefly describes the Danish housing market. Section C discusses the model and its calibration. Section D presents the model simulations. Finally, section E concludes with a discussion of policy implications.

B. The Danish Housing Market—an Update

8. Prices for both single family houses and owner-occupied flats have been increasing. In mid-2015, prices for single family houses had risen by about 6 percent in real terms compared to a year earlier and by some 12 percent when compared to the market trough in 2010. However, prices have been rising faster in selected segments of the market, e.g., for Copenhagen, the increases since last year and since the trough were 13 and 43, respectively, while they were 12 and 33 percent for owner-occupied flats. Moreover, house price expectation indicators suggest the increasing trend in house price is most likely to continue.

A03ufig5

House Price Expectations

(Index)

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A003

Sources: Nykredit, Greens Analyseinstitut for dagbladet Borsen, and fund staff calculations
A03ufig6

Distribution of Loan-to-value Ratios for Indebtedness Households in 2013

(Percent)

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A003

Sources: Danmarks NationalBank

9. Sizable pockets of vulnerable households remain. The loan-to-value (LTV) ratio for the aggregate mortgage stock has fallen by about 5 percentage points from its peak, to about 55 percent in 2015. However, over 30 percent of indebted households have a LTV ratio above 100 percent. Similarly, at 12 percent, the share of households with a debt-to-income (DTI, defined as aggregate household debt as a share of income) ratio above 400 percent is high. Moreover, the share of families with both high LTV and DTI ratios remains above the level in 2007 suggesting a significant share of households are vulnerable to shocks.2 Moreover, the LTV cap, currently at 95 percent, suggests that highly indebted households could borrow more as house prices increase.

10. Variable rate mortgages account for a majority of the existing mortgage loans. The share of variable rate mortgages increased by 38 percentage points before reaching a peak of 66 percent at end-2014. More recently, however, households have been increasingly opting for fixed rate loans. Besides locking in historically-low interest rates, one reason for this trend is the increasing differentiation of administration margins, which has made fixed rate mortgages more attractive. Still, 75 percent of the families with high debt have variable rate mortgages, making them particularly vulnerable to an increase in interest rates.

A03ufig7

Distribution of Debt-to-income Ratios for Indebtedness Households in 2013

(Percent)

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A003

Sources: Danmarks Nationalbank
A03ufig8

Share of Families with High Debt and Their Debt level 1/

(Percent)

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A003

Sources: Danmarks Nationalbank.1/ High debt households is defined as families whose total debt excees the value of their home and constitutes at least four times their annual gross income.

11. The share of interest only mortgages has increased significantly. Only a decade ago, almost all of the mortgage-credit loans had amortization plans. Since then, however, a greater number of households have opted for deferred amortization loans with variable interest rates, including with a view to optimize the benefits of interest tax deductibility. In 2015H1, variable rate loans with deferred amortization accounted for 42 percent of mortgage lending to households. This ratio increases to about 55 percent if fixed rate deferred amortization mortgages are also included. In order to maintain a good lending standard in the low interest rate environment, mortgage banks require the borrower to be able to service a 30-year fixed rate loan with amortization when granting deferred amortization loans. However, in the current environment, this means that a mortgage bank will effectively grant a variable rate, interest-only loan once a borrower is able to service a loan with a fixed rate of 3 percent.

A03ufig9

Share of Fixed and Variable Mortgages

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A003

Sources: Danmarks Nationalbank.
A03ufig10

Household Mortgage Debt

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A003

Sources: Danmarks Nationalbank.

12. The very low policy rate has to great extent been passed through to mortgage borrowing costs. At the beginning of the year, mortgage bond yields fell below zero for bonds with a maturity of up to three years. Despite mortgage banks having raised their administration margins, borrowing costs have fallen to the lowest level since 2003, with the cost of a 10-year fixed rate mortgage at only 3 percent in mid-2015.

13. Tax policy is also providing incentives for debt financed house purchases. Housing taxes remain largely unchanged over time as the property tax has been frozen in nominal terms since 2002. Moreover, there is a cap on year-on-year increases in land tax in the form of a regulation ratio. As a result, the effective tax rate varies across different regions in Denmark, with areas experiencing higher house price increases sees the lower increase in property taxes. Moreover, mortgage interest payments are tax deductable, currently at 33.6 percent.3 This provides an incentive to households to finance a house purchase with debt and allows them to borrow more by reducing the effective borrowing cost.

A03ufig11

Mortgage Banks’ Average Interest Rates by Interest Rate Fixation 1/

(Percent)

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A003

Sources: Danmarks Nationalbank.1/ Including administration margins on new lending to households.

14. The authorities have recently taken welcome steps to strengthen mortgage lending standards. A minimum down payment requirement of five percent for new mortgages was introduced. In addition, the authorities have issued guidelines—“Seven Best Practices”—for banks in areas with rapid house price increases to strengthen their risk management practices. Moreover, the FSA has introduced the supervisory diamond for mortgage-credit institutions, which was introduced at the end of 2014, and is being phased in over a number of years.

A03ufig12

DFSA Supervisory Diamond for Mortgage-Credit Institutions

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A003

Source: Danish FSA.1/ Lending segments are the following: private homeowners, rental property, agriculture and other corporate.2/ Applies only to lending to private homeowners and rental property.3/ Applies only to private homeowners.

C. The Modeling Approach and Calibration

15. To analyze the effectiveness of macroprudential policies, while considering also the impact on the real economy, we use a DSGE model based on Chen and Columba (2016). The model features a small open economy framework with financial frictions, an imperfectly competitive banking sector, and a selection of macroprudential tools. To adapt the original model to the Danish policy framework, we have modified the monetary policy rule so that the target of the central bank is to maintain nominal exchange rate stability. In addition, we have introduced a property tax into the model. The model can be used to analyze a range of possible instruments including loan-to-value caps, amortization requirements, tax deductibility of mortgage interest payments, property taxes, and mortgage risk weights.

16. The model features a monopolistic banking sector and a range of macroprudential instruments. The modeled world consists of two economies: home and foreign. The home economy is populated by two types of households—patient and impatient—and entrepreneurs. Households consume, work and accumulate housing (the supply of which is fixed), while entrepreneurs produce a homogenous intermediate good using physical capital bought from capital-goods producers and labor supplied by households. Agents (households and entrepreneurs) have different degrees of time preference (“patience”) reflected in different discount factors for their future utility. The heterogeneity in agents’ discount factors provides a simple way to generate financial flows in equilibrium: patient households (“savers”) purchase a positive amount of saving assets (deposits at domestic banks and foreign bonds) and do not borrow, while impatient households (“borrowers”) and entrepreneurs borrow from the domestic banking system. Impatient households can borrow up to a fraction, at 95 percent, of the value of new housing acquisitions each period. We include an amortization requirement by assuming that the impatient households must repay a fixed fraction of the loan principal each period. We defer reader to Chen and Columba (2016) for a more detailed discussion of the model.

A03ufig13

Model Structure

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A003

17. We calibrate the model to resemble the Danish economy, partly using estimated parameters from the literature. Since the literature on estimated DSGE models for Denmark is limited, however, we rely in particular on Pedersen and Ravn (2013). For example, we set the LTV cap for new mortgages at 95 percent, and an average amortization period of mortgages to 40 years. We set the minimum bank risk weighted capital ratio to 12 percent, matching the capital requirement for Danish banks in 2019. For price and wage adjustment costs, we calibrate the adjustment factors such that they are in the range suggested by Pedersen and Ravn (2013). The calibrated model closely resembles key features of the Danish economy. For instance, the model implies consumption and investment shares of about 50 and 20 percent of GDP, respectively. The steady state debt as a share of the value of the housing stock (i.e., LTV ratio for the aggregate economy) is about 60 percent.

D. Policy Exercises

18. We use the framework to study the effectiveness of macroprudential policies and assess their impact on the real economy. We focus on three macroprudential instruments: (i) tax deductibility of mortgage interest payments; (ii) amortization requirements (or limits on deferred amortization loans); and (iii) loan-to-value caps. Given that the model has more than 100 variables, we focus the discussion on the key ones.

19. Specifically, we compare the steady states of the economy after changing the combination of macroprudential policy instruments. We describe the transition paths of the key variables from one steady state to another. For example, a permanent reduction in tax deductibility of mortgage interest payment would tighten borrower’s liquidity conditions and change the relative price between consumption goods and housing investment (making consumption goods cheaper). Such a change would affect the borrower’s behavior, which would in turn interact with the savers, entrepreneurs, banks and other agents in the model, until a new equilibrium is reached.

Tax Deductibility of Mortgage Interest Payments

20. A reduction in tax deductibility has complicated, but eventually negligible effects on consumption. As discussed above, a reduction in the tax deductibility of mortgage interest increases the cost of servicing mortgage debt, thereby tightening a household’s budget constraint—a negative income effect. Moreover, such a reduction would make debt-financed house purchases more costly relative to goods consumption—a substitution effect. The two effects have opposing effects on household consumption. The former suggests a reduction in tax deductibility lowers household’s consumption, while the substitution effect implies that households consume more as the relative price change has made consumption goods cheaper. The text figure illustrates the impact of a 5 percentage points reduction of tax deductibility (i.e., a reduction to 28.6 percent):

  • Short run. The income and substitution effects lower borrowers’ demand for housing, with house prices falling by 0.5 percent in the near term. However, borrowers’ consumption increases in the short run due to the substitution effect. Moreover, the model assumes the government redistributes the savings raised from the reduction in tax deductibility to both savers and borrowers. This helps to alleviate some of the impact from the negative income shock to the borrowers. Savers would increase their housing investment as house prices decline, but they also benefit from the higher transfers leading to higher consumption. Overall, consumption, output, and inflation rise slightly over the short run.

  • New steady state. Borrowers’ debt falls by about 2 percent, with a similar change in the debt-to-income (DTI) ratio. Their consumption increases by 0.1 percent, but housing investment falls by 2 percent. Savers’ consumption also increases in this case by 0.05 percent, driven by higher transfers as explained above. The latter would be sufficient to offset the impact of the decline in bank profits such that a decline in saver’s consumption is avoided. How the government redistributes the “savings” from the reduction in tax deductibility is important, and it can be shown that savers’ consumption would decline if the government does not fully redistribute the “savings” from reducing tax deductibility.

A03ufig14

A 5 Percentage Points Reduction in Tax Deductibility

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A003

Note: Figure depicts maximum impacts on household mortgage debt, DTI and consumption (Cons) following a permanent reduction in mortgage tax deductibility. And changes in the three variables in the new steady state (tax deductibility = 28.6 percent) compared with the baseline (tax deductibility = 33.6 percent).

Loan-to-Value Cap

21. A cap on the LTV ratio constrains how much households can borrow and has a substantial impact on debt levels. House prices would fall as borrowers’ housing demand is reduced by their more limited ability to borrow. Falling house prices implies a lower value of housing collateral, thus it amplifies the effect of a tightening of the LTV cap. In what follows, we consider a scenario in which the LTV cap is reduced by 5 percentage points to 90 percent.

  • Short run. Borrowers’ consumption falls sharply by 5 percent immediately after the shock. Savers’ consumption decreases as well for two years after the shock, although by less. This is partly driven by a decline in the deposit rates as the lower stock of mortgages reduces bank profitability. Altogether, the tightening of the LTV cap has a contractionary short-term effect on the economy, lowering aggregate consumption by 1.5 percent one year after the shock.

  • New steady state. Borrowers’ mortgage debt declines by 7 percent, in part because house prices fall by 2 percent. The relatively modest decline in house prices reflects to some extent the significant price elasticity of savers’ demand for housing compensating for the decline in borrowers’ demand (as supply relative to population is assumed to be fixed). Notably, borrowers’ consumption of goods will be permanently higher by about 0.2 percent in the new steady state, as their debt service burden is lower, partly offsetting the decline in their consumption of housing services. But savers’ consumption would decline during the transition, and be 0.4 percent lower in the new steady state. This result is driven by bank profits falling by about 2 percent as banks cut back on mortgage lending. Lower bank profits also imply that banks deleverage, cutting loans to firms by about 0.8 percent implying lower investment and production. As a result output will be about 0.2 percent lower in the new steady state.

A03ufig15

A 5 Percentage Point Reduction in LTV

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A003

Note: Figure depicts maximum impacts on household mortgage debt, DTI and consumption (Cons) following a permanent reduction in LTV cap. And changes in the three variables in the new steady state (LTV= 90 percent) compared with the baseline (LTV= 95 percent).

Amortization Requirement

22. An amortization requirement would have the strongest impact on debt. With the introduction of amortization requirements, a portion of the mortgage principal must be repaid each period, in an amount set by the amortization plan in the loan contract. This then sets the maturity of the loan. We examine the effects of reducing the maturity of new loans from 40 years to 35 years.

  • Over the short run. The model suggests that borrower’s consumption would fall by a very small amount (i.e., by 0.2 percent 2.5 years after the shock). Overall, a tightening of the amortization requirement has almost no impact on growth.

  • In the new steady state. Borrowers’ debt will fall by about 15 percent, with households’ DTI ratio falling by the similar amount, and house prices down by 0.5 percent. In addition, borrowers’ housing stock will be about 1 percent lower in the new steady state. Borrowers’ consumption will be about 1.5 percent higher permanently. Savers’ consumption will decline by 0.4 percent. This again reflects a lower bank profits (by over 5 percent) leading to a decline in credit to firms of more than 1.5 percent. As a result, output is lowered by about 0.4 percent.

A03ufig16

A Reduction in Required Amortization by Five Years

Citation: IMF Staff Country Reports 2016, 185; 10.5089/9781498319805.002.A003

Note: Figure depicts maximum impacts on household mortgage debt, DTI and consumption (Cons) following a permanent reduction in amortization requirement. And changes in the three variables in the new steady state (amortization = 35 years) compared with the baseline (amortization = 40 years).

E. Conclusion

23. This paper provides some tentative estimates of the potential effects of tightening macroprudential policies in Denmark. The results are intended to provide food for thought for policymakers when thinking about macroprudential policy formulation. A few results stand out. First, the overall impact from macroprudential policies on the real economy appears limited, while the effect on debt levels can be significant. Second, out of the various modeled policies, amortization requirements seem to have the strongest impact on household debt, suggesting the importance of limiting the growth of deferred amortization mortgages. The results are indicative and the usual caveats apply. Mainly, the paper provides a basic framework for further research, including possible welfare analysis that enables the policymakers to understand better the trade-offs between household’s ability to borrow and consume housing, and their vulnerability to financial shocks. For example, Chen and Columba (2016) show on the basis of a similar framework for Sweden that a combination of macroprudential tools can achieve higher welfare than relying on a single measure.

References

  • Chen, Jiaqian, and Francesco Columba, 2016, “Macroprudential and Monetary Policy Interactions in a DSGE Model for Sweden,” IMF Working Paper WP/16/74 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Pedersen, Jesper and Ravn, Soren Hove, 2013, “What Drives the Business Cycle in a Small Open Economy? Evidence from an Estimated DSGE Model of the Danish Economy,” Danmarks Nationalbank Working Paper No. 88.

    • Search Google Scholar
    • Export Citation
1

Prepared by Jiaqian Chen.

2

High debt households are defined as families whose total debt exceed the value of their home and constitutes at least four times their annual gross income.

3

The tax value of capital income exceeding DKr 50,000 per person is gradually reduced by 1 percentage point a year until 2019 to a rate of 25.6 percent.

Denmark: Selected Issues
Author: International Monetary Fund. European Dept.