ACCENTRO-IW HOUSING COST REPORT 2019
An Analysis of Rents and Owner-Occupied-Housing Costs in 401
ACCENTRO-IW HOUSING COST REPORT 2019
Third-Party Sponsored Expertises Expert Opinions
An Analysis of Rents and Owner-Occupied-Hous- ing Costs in 401 Districts
Prof. Dr. Michael Voigtländer, Pekka Sagner
Client: ACCENTRO Real Estate AG Cologne, 12 April 2019 CONTACT DETAILS
Prof. Dr. Michael Voigtländer +49 (0)221 / 4981 - 741 [email protected]
Pekka Sagner +49 (0)221 / 4981 - 881 [email protected]
Cologne Institute for Economic Research (IW) PO box 10 19 42 D-50459 Cologne
2 ACCENTRO-IW Housing Cost Report 2019
CONTENT
Summary ...... 4
1 Introduction ...... 5
2 User costs ...... 5
2.1 Notes on the Methods Used ...... 5
2.2 Findings for Germany ...... 8
2.2.1 Long-Term Owner-Occupied-Housing Costs ...... 8
2.2.2 Latest Owner-Occupied-Housing Costs and Rents ...... 9
2.2.3 Regional Evaluations ...... 11
2.2.4 Interest Rate Sensitivity on the District Level ...... 15
3 Trend in Property Prices and Property Financing ...... 16
4 Responses among Property Buyers ...... 21
5 Conclusion ...... 25
6 Bibliography ...... 26
List of Tables ...... 27
List of Figures ...... 27
Annex ...... 28
3 Cologne Institute for Economic Research
SUMMARY
The user cost approach makes it possible to compare rental costs and the expenses periodically arising for homeowners. The approach is used to determine the relative economic benefit of homeownership versus renting your home.
Homeownership remains attractive. At present, owner-occupying your home is more affordable than rent- ing in 94 percent of Germany’s districts and independent cities. The average economic benefit of home- ownership nationwide is nearly 40 percent. Even in the country’s metropolises, where condominium pric- es experienced particularly brisk growth in recent years, homeownership retains its economic benefits. Owner-occupancy in Berlin, for instance, is currently about 27 percent more affordable than renting, while being 35 percent more affordable in Hamburg. This is explained by the combination of simultaneous rental growth and the still highly favourable conditions on the market for mortgage loans. While interest rates have registered a modest increase lately, they are not expected to go up sharply any time soon. Moreover, an interest rate sensitivity analysis revealed that a large number of districts would be resilient to an inter- est rate tightening cycle. The fact that owner-occupied-housing costs are lower than the costs of renting almost everywhere also suggest that prices for freehold residential property are more likely to keep rising, defusing concerns that the housing market might be overheating.
Meanwhile, the fact that the number of first-time buyers has declined overall while their age average and— most recently—their income levels have increased indicates that few households in Germany benefit from the favourable terms of financing in the sense that they take advantage of them to acquire homes for own- er-occupancy. Capital adequacy requirements, which go up in sync with rising prices, represent one of sev- eral reasons for this. In order to enable more households to opt for homeownership and, not least, to bolster their retirement schemes by doing so, the body politic should review the entry thresholds on the market for freehold residential property and lower them for the sake of preserving the country’s financial stability.
4 Survey compiled for ACCENTRO Real Estate AG ACCENTRO-IW Housing Cost Report 2019
1 INTRODUCTION
The user-cost-of-housing approach or user cost approach, for short, is an internationally wide-spread meth- od for evaluating developments on the housing market (cf. Poterba [1984] or Himmelberg et a.. [2005]). The approach is used, on the one hand, to derive possible indications of overvaluations and thus of a spec- ulative bubble, but, on the other hand, it also permits observations regarding the advantage of home own- ership over renting. For several years now, the Accentro-IW Housing Cost Report has applied this approach to Germany as a whole and to each of its 401 administrative districts.
As with any empirical analysis, it is of the essence to keep verifying the data used and the underlying as- sumptions and to adjust them as needed. In the case of the Accentro-IW Housing Cost Report, this process prompted a number of adjustments. For one thing, we no longer use asking prices but instead rely on transaction data made available by vdpResearch (2019), a research firm focused on the German real estate market. In addition, we put the process of deriving anticipated property appreciation figures on a broader data basis in order to avoid possible distortions among the districts due to an unduly short period of time. While these adjustments compromise the comparability of the findings with previous analyses, they help to make the profiling of developments more accurate. Moreover, the repayment calculation previously used was replaced by an affordability calculation that is generally more intuitive.
In addition to these methodological adjustments, we also mapped the long-term development this time, based on OECD data. Looking back all the way to the 1970s is making it easier to put today’s situation on the housing market in perspective. The analyses moreover covered the performance of first-time buyers in the housing market and differentiated between urban and rural regions.
Despite the methodological adjustments, the overall picture remains robust: In most of the districts, home- ownership is still more attractive than renting a home even though the owner-occupied-housing costs have begun to rise because of the price trend and rebounding interest rates. However, the stats for first-time buyers show that few households actually take advantage of the opportunities in the housing market. It is high time German policymakers made an effort to lower the threshold to homeownership.
2 USER COSTS
2.1 Notes on the Methods Used
The user-cost-of-housing approach that is subsequently applied to Germany and that is also referred to as “user cost approach” was introduced by Poterba (1984) and Himmelberg et al. (2005). The approach is based on the assumed premise that households are principally indifferent to the choice between buying or renting a given property. The premise is assumed to be valid if the relative costs of either option are identi- cal. If, for example, the costs were to shift in favour of homeownership, this would boost the relative appeal of buying a home and with it demand for homes. Increased demand in the market for freehold residential property will raise the selling prices in the respective regions while rents will become more affordable, relatively speaking, restoring the balance. Over short periods of time, the residential property market is rigid, meaning that new-build construction will respond to a surge in demand for housing in a given region with considerable delay. More important yet is the fact that people are loath to relocate, so that the pace of the market adjustment is very slow. The slow response rate, including to declining demand, explains why owner-occupied-housing costs and rents may drift apart for limited periods of time.
5 Cologne Institute for Economic Research
Comparing the costs of tenants with those of property owners is no mean feat because rental costs are a flow variable whereas the purchase price represents a one-time payment. This is where the concept of own- er-occupied-housing costs comes in: The purchase price, including incidental acquisition costs, and with financing costs and lost income on the equity employed for the property purchase all taken into account, is converted into a flow variable. The variable permits a comparison between rental costs and the costs borne by owner-occupiers.
According to Schier/Voigtländer (2015), the annual owner-occupied-housing costs can be determined on the district level at any given time as follows:
SNKkt = Pkt ∙ (1 + gkt + mkt + e + n) ∙ [b ∙ iF,t + (1 - b ) ∙ iA,t ∙ (1 - τt ) + s + a -∆Pk ] The letter P represents the property purchase price in euros per square metre of dwelling floor area. For the first time, we used data obtained from vdpResearch (2019). The term in the subsequent bracket combines the incidental costs of a property acquisition: g represents the amount of the real estate transfer tax, whose rate varies from one federal state to the next, anywhere from 3.5 to 6.5 percent of the purchase price. Moreover, we assume that the property is acquired through an estate agent, with the agent’s fee m also varying from one federal state to the next and generally ranging from 3.57 to 7.14 percent of the purchase price. For the entry in the land register e and the notarial charges due n, we impute a flat rate of 1.525 per- cent. The purchase price is normally financed in the form of a mortgage loan. In recent years, the borrowed capital share or leverage b approximated 78 percent of the purchase price (Dr. Klein, 2019). As variable borrowing rate iF we assume the mean effective interest rate that German banks charge for housing loans to private households with an initial fixed interest period of more than 10 years (Deutsche Bundesbank, 2019a). In addition to the actual payments to be made for the property bought, opportunity costs are in- curred for the equity employed (on average, 22 percent of the purchase price). As opportunity interest rate, we impute the average current yields of domestic bearer bonds as iA (Deutsche Bundesbank, 2019b). The income generated from the investment on the capital market must be taxed at the rate τ. To this end, we use the average tax rate as delineated in the official finance statistics (BMF, 2018). Homeowners also shoulder annual costs in the form of maintenance and depreciation a. To cover these, we impute a flat rate of 3 per- cent (Clamor et al., 2013). They represent costs that should be considered on an opportunity basis. If a con- dominium owner skipped these expenditures, for instance by failing to execute necessary refurbishment or modernisation measures, the property would depreciate year by year. Finally, the expected long-term price growth rates for residential real estate in a given district are factored in as ΔP with a negative sign. The long-term price expectations in this context are based on the average annual price growth rate for the years 2005-2018. To defuse the significance of excessive recent price hikes, if any, we limit the maximum annual price growth rate to 3 percent.
6 Survey compiled for ACCENTRO Real Estate AG ACCENTRO-IW Housing Cost Report 2019
Table 2.1: Variables and data sources
Variable Explanation Source Purchase price in euros per sqm of vdpResearch (2019) Pkt dwelling floor area b Debt capital share Dr. Klein (2019)
i F,t Mortgage interest rate Deutsche Bundesbank (2019a)
i A,t Current yield on bearer bonds Deutsche Bundesbank (2019b)
τt Tax rate BMF (2018)
ΔPk Purchase price change F+B (2019) Source: IW Economic Institute
Example This section will illustrate the calculation of the owner-occupied-housing costs by determining them for a model city as follows. Let us say that the purchase price per square metre of dwelling floor area is 4,000 eu - ros The incidental acquisition costs break down into the real estate transfer tax (in this case 6 percent), the agent’s fee (in this case 3.57 percent), the costs of the land register entry and the notarial charges (in this case 1.525 percent), adding up to a sum total of 444 euros per square metre or about 11 percent of the pur- chase price. The leverage is 78 percent and subject to an interest rate of 1.96 percent. The equity ratio is 22 percent, while an opportunity interest rate of 2.48 percent is assumed for an investment in this amount in the capital market, which matches the average current yields that German bearer bonds generated in 2018. Such investment income is taxable, the imputed tax rate being 22.4 percent (average tax rate as delineated in the financial statistics for 2017 and adopted for 2018). The expected annual price growth rate is imputed as 2.5 percent.
The above assumptions return owner-occupied-housing costs in an amount of c. 9.08 euros per square metre of dwelling floor area and month. If these housing user costs happen to be lower than the monthly rent for a comparable dwelling, owner-occupying your home has relative economic benefits over renting, and vice versa.
Critical Discussion of the Assumptions One of the key factors influencing the determination of owner-occupied-housing costs are the exception- ally low interest rates on the mortgage loan market. Lower interest rates on the mortgage lending market reduce housing user costs because they decrease the total loan amount to be financed. Since they are cur- rently based on a very low mortgage rate level, residential user costs are much more sensitive to an interest rate hike. If, to use the above example again, mortgage rates were to rise by one percentage point, it would raise the owner-occupied-housing costs to 11.97 euros or 31.8 percent. If all other assumptions continued to apply, an increase in mortgage rates from 3 percent to 4 percent would raise the owner-occupied-hous- ing costs from 12.08 euros to 14.97 euros – an increase by 23.9 percent only.
The expected long-term price growth rates are also of key significance for the owner-occupied-housing costs. Whenever the expected annual price growth rates go up, the owner-occupied-housing costs will decline. The average price growth rates in Germany’s districts and independent cities have fluctuated sub- stantially in recent years. In order to include both current trends on the real estate market and to exclude price exaggerations in particularly strained housing markets, the expected price growth rate is determined here as a long-term price trend based on the average annual price growth rates of the years 2005 through 2018 and capped at an annual 3 percent. The negative impact of the price growth rate on owner-occu-
7 Cologne Institute for Economic Research
pied-housing costs could actually cause the rate to be overestimated in some of the districts. As a conse- quence, the economic benefit could be underestimated in these districts while homeownership is more advantageous than apparent. While the expected price changes by no more than 3 percent (nominal) per year may seem very moderate—at least when compared to the development of recent years—they need to be contextualised with the historical price trend. Nationwide, average German selling prices have only increased by an annual 0.23 percent (in real money terms) since 1970 (cf. Abbildung 3.1).
2.2 Findings for Germany
2.2.1 Long-Term Owner-Occupied-Housing Costs
Before studying the latest findings of the user cost approach below, let us have a look at the long-term trend of rental costs and owner-occupied-housing costs, respectively.
The OECD provides data on the growth of rents and prices in Germany since 1970. The data are based on figures provided by Deutsche Bundesbank. However, the OECD only offers indices, no absolute figures. They are therefore unsuitable for deriving actual owner-occupied-housing costs. The idea in the present context, however, is only to outline the long-term trends of owner-occupied-housing costs relative to rental costs. To do so, we proceed as follows: In a first step, the index of house prices is linked in a multiplicative relation with the long-term interest rate, which is also provided by the OECD. The long-term interest rate reflects the returns on long-term government securities, whose interest rates tend to be slightly lower than long-term mortgage rates; still, their performance is comparable. The changes in interest rates tend to be parallel, subject to minor differences in the respective levels.
By linking house prices and interest rates, you get the simplest version of owner-occupied-housing costs (DiPasquale/Wheaton, 1992). To be able to compare this calculated owner-occupied-housing cost index with the rent index, the second step references the central assumption of the user cost approach: In the long term, owner-occupied-housing costs and tenant costs are the same. It is therefore assumed that the costs between 1970 and 2018 are identical on average. This means that the differences in costs between the two types of use across the entire period return a zero sum. The index of owner-occupied-housing costs is shifted exactly as much as needed to satisfy this premise. For the outcome of the calculation, see Figure 2.1.
The figure shows that the economic benefit of homeownership has clearly shifted. Between 1970 through the end of the 1990s, tenants were better off than homeowners. This phase was characterised by very high interest rates and occasional price hikes that compromised the appeal of residential property relative to rents. But since the Zero Years, homeownership has been the cheaper option because the interest rate development more than offset the price trend. Another factor was that rents rose much faster during the 1990s, partly due to strong demand in the early 1990s in the wake of Germany’s reunification and due to immigration, but to some extent also because of the abolition of the limited-profit housing status in the late 1980s. Only recently have owner-occupied-housing costs started to perk up again, after the gap between renting and buying had steadily widened for years. Due to the index scores used and the assump- tions made (equivalence of rental costs and owner-occupied-housing costs), one should not read too much into the analysis: Particularly the differences between the two cost types are strongly dependent on the un- derlying assumptions. Some may find fault with the use of interest rates on long-term government securi- ties because their interest rates have fallen so much faster, particularly in recent years, than mortgage rates after the ECB’s interventions (cf. Demary/Voigtländer [2018]). Nonetheless, the figure is very revealing. On the one hand, it confirms prior analyses, such as the Expert Commission on Housing Policy (1995), which found that renting generally offers an economic advantage. On the other hand, however, it also shows that
8 Survey compiled for ACCENTRO Real Estate AG ACCENTRO-IW Housing Cost Report 2019
the appeal of real estate in general has increased significantly as a result of declining interest rates over several decades. Historic multipliers, meaning price-to-rent ratios, are therefore useful only to a certain extent since the drop in interest rates implies higher multipliers. Looking at the slow adjustment processes, the benefits of homeownership could stay in place for a long time to come. The section below will now take a closer look at the latest developments.
Figure 2.1: Index1) of long-term rent rates and owner-occupied-housing costs
Rent index Owner-occupied-housing costs
140
120
100
80
60
40
20
0 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018
1) Index 1997 = 100, nominal prices
Source: OECD (2019); IW Economic Institute
2.2.2 Latest Owner-Occupied-Housing Costs and Rents
The average owner-occupied-housing costs in Germany are close to 5.58 euros per square metre of dwell- ing floor area and month. These contrast with average monthly rental costs of 9.24 euros for acompa- rable flat rented on new tenancy. Thus, the costs of owner-occupancy are almost 40 percent lower than the rental costs. By 2018, owner-occupying a home was more affordable even when compared to passing rents, the average rental costs on unexpired leases being 6.72 euros and therefore 17 percent higher than owner-occupied-housing costs. Figure 2.2 illustrates the development of owner-occupied-housing costs and rents between 2010 and 2018. No extended time series are available for the transaction data provided by vdpResearch (2019). At the start of the current decade, owner-occupied-housing costs and rental costs were still on the same level. But since 2012, owner-occupied-housing costs have undercut even passing rents. One of the drivers of owner-occupied-housing costs is the interest rate development. Annual interest
9 Cologne Institute for Economic Research
on mortgage loans with maturities of over 10 years was around 4 percent as recently as 2010 and 2011 (Deutsche Bundesbank, 2019a). During the period under review, owner-occupied-housing costs bottomed out in 2016. Up to then, interest rate had kept tumbling until they hit 1.8 percent. In 2017 and 2018, the owner-occupied-housing costs began to nudge up again in sync with the interest rate development. Lately, the growth of owner-occupied-housing costs has outpaced both passing rent and new tenancy rents, imply- ing an emerging convergence of rental costs and owner-occupied-housing costs. Nevertheless, the relative attractiveness of owning your home versus renting it remains high. Nationwide, a comparison of average passing rents and average owner-occupied-housing costs also shows that even terminating an unexpired lease to buy a home is generally associable with economic benefits.
Figure 2.2: Trend in owner-occupied-housing costs and rents1) Population-weighted German2) average, in euros per square metre of dwelling floor area and month
Owner-occupied-housing cost New tenancy rents Passing rents
10
9
8
7
6
5
4
3
2 2010 2011 2012 2013 2014 2015 2016 2017 2018
1) Passing rents (F+B, 2019) refer to a dwelling of standard fit-out specifications and state of repair. Rents on new leases (vd- pResearch, 2019) and selling price are based on transaction data and refer either to first-sale prices or to resale prices of fully refurbished flats in good locations and with good interior specification.
2) With no population data for 2018 available yet, the population weightings of 2017 were adopted for 2018, too. To do the census reset of 2011 justice, retrograde calculation was used for 2010 as defined by the BBSR Federal Institute for Research on Building, Urban Affairs and Spatial Development (2018).
Source: IW Economic Institute based on data by vdpResearch (2019); F+B (2019)
10 Survey compiled for ACCENTRO Real Estate AG ACCENTRO-IW Housing Cost Report 2019
2.2.3 Regional Evaluations
The situation in Germany’s metropolises, the so-called “Big Seven” cities, more or less matches the na- tional average, which is explained not least by the fact that the sheer number of residents living in these cities definitively influences the country’s population-weighted average. In all of the metropolises, own- er-occupied-housing costs undercut the rental costs of a new tenancy. In Berlin (27 percent), Hamburg (35 percent) and Munich (38 percent), the advantage of owner-occupied-housing costs falls short of the population-weighted nationwide average. Inversely, the advantages of homeownership actually exceed the German average in Düsseldorf (54 percent), Frankfurt am Main (50 percent), Cologne (54 percent) and Stuttgart (44 percent).
Yet the major German cities seem to be in a process of converging owner-occupied-housing costs and rental costs. The recent surge in purchase prices relative to rental growth, combined with slightly higher interest rates for mortgage loans, have led to an increasing equalisation of owner-occupied-housing costs and new tenancy rents, especially in Berlin. Here, owner-occupied-housing costs went up by 19 percent annually over the past two years. The economic benefits of owner-occupancy are now drastically diminished.
Figure 2.3 shows moreover that the owner-occupied-housing costs exceed passing rents in Berlin, Ham- burg and Munich. This means that in these cities, living in a rented flat with standard fit-out specification on an average passing rent tends to be cheaper than moving into a refurbished condominium with a good fit-out specification in a good location. As said before, no passing rent stats are available for directly com- parable assets. Still, drawing the comparison makes sense insofar as it will be an attractive proposition for many prospective property buyers to buy a home with superior fit-out specifications. In Berlin, own- er-occupied-housing costs have lately exceeded passing rents by 21 percent, the latter rising noticeably slower because the city has a very large rental housing market where tenants make up around 82 percent of the total (Henger et al., 2019). Owner-occupied-housing costs similarly top passing rents by 10 percent in Hamburg and by 3 percent in Munich. Still, it is safe to say even for these cities: If you plan to move and have the choice of two comparable flats, one to let, one for sale, owner-occupancy will be more affordable.
11 Cologne Institute for Economic Research
Figure 2.3: Owner-occupied-housing costs and rents in Germany’s metropolises In euros per square metre of dwelling floor area and month