Loading...

Are properties in Munich partially overvalued? An investigation of local real estate bubbles and price development within the districts

Master's Thesis 2014 99 Pages

Business economics - Miscellaneous

Excerpt

Table of contents

Abstract

Table of contents

List of abbreviations

List of figures

List of tables

List of units

List of data media

1 Introduction

2 Scope and limitations of research
2.1 Data collection techniques
2.2 Definition of a real estate bubble
2.3 Fundamental value - The stock-flow model

3 Literature Review

4 City based perspective from 1975 to
4.1 Historical development
4.1.1 Population and labour market
4.1.2 Building and housing
4.1.3 Housing and renting prices
4.1.4 Interest rate and mortgages
4.2 Characteristic ratios
4.2.1 Price-to-rent ratio
4.2.2 Price-to-income ratio
4.2.3 Housing affordability index
4.3 Empirical time series analysis
4.3.1 The time series model
4.3.2 Statistical analysis

5 District based perspective from 2000 to
5.1 District based investigation
5.1.1 Development of population in the districts
5.1.2 Development of building and housing in the districts
5.1.3 Development of housing prices in the districts
5.2 Empirical panel data analysis
5.2.1 The panel model
5.2.2 Cross section versus time series dimension
5.2.3 Statistical Analysis
5.3 Forecast

6 Conclusion, results, perspectives

Appendix
Appendix I: Imputation, forecast and figures of missing values
Appendix II: Figures time series analysis
Appendix III: Panel data analysis transformation of variables
Appendix IV: Figures panel data analysis

Bibliography

Abstract

The burst of the US- real estate bubble in the year 2008 has contributed significantly to the global financial crisis, from which the global economy is still recovering from. In the US, housing prices collapsed after years of increase, which led to credit de- faults and the collapse of the banking system. The subprime crisis, however, was only the highlight of a great new set of consequential price bubbles on international real estate means, which could be seen to swap over to Spain as well as Ireland. This poses the question whether real estate bubbles might exist in Germany. The opinions of the nation are divided. The German city investigated particularly in this article is Munich, since prices are at an all-time high and have been rising continuously during the last decades. To be more specific, this master thesis aims to identify whether there is a local real estate bubble within the German city Munich. This is done by regressing two sets of data and using a stock-flow model in order to determine the relationship between housing prices and explanatory macroeconomic variables: The first set is a time series regression on a whole city perspective within the years 1975 and 2013. The second set is a panel data analysis on district basis within the years 2000 to 2012. Results suggest that in the first set condominium prices are moderately and rowhouse prices are slightly overvalued in Munich, especially in the recent years. The results of the second set illustrate an overvaluation of condominium prices on a slightly lesser level.

Keywords: Real estate bubble, housing bubble, fundamental value, fundamen- tal house prices, stock-flow model, price development, Munich

List of abbreviations

Abbildung in dieser Leseprobe nicht enthalten

List of figures

Figure 1: Fundamental value based perception of a real estate bubble

Figure 2: Population development within Munich, from 1975 to 2013

Figure 3: Residential building- and flat stock in comparison to annual permits in Munich, from 1975 to 2013*

Figure 4: Average condominium and rent price for median flat of 75 m² and rowhouse buying price of Munich, nominal, from 1975 to 2013

Figure 5: Growth rate of condominium, rent and average rowhouse prices of Munich, real, from 1975 to 2013

Figure 6: Nominal and real average, effective interest rates for mortgage credits to private households in Germany (all time spans), from 1975 to 2013

Figure 7: Price-to-rent ratio of average rowhouse prices and condominium prices for an average flat of 75m²in Munich, from 1975 to 2013

Figure 8: Price-to income ratios for average rowhouse prices and condominium prices for an average flat of 75m² in Munich, from 2000 to 2013*

Figure 9: Housing affordability index for condominium price of an average flat of 75m² and average rowhouse prices in Munich, from 2000 to 2013*

Figure 10: Condominium price deviation from long-run equilibrium in %, from 1975 to 2013

Figure 11: Condominium price deviation from calculated fundamental value, from 1975 to 2013

Figure 12: Rowhouse price deviation from long-run equilibrium in %, from 1975 to 2013

Figure 13: Rowhouseprice deviation from calculated fundamental value, 1975 to 2013

Figure 14: Housing starts in Munich per year, from 2000 - 2012

Figure 15: Migration balance of districts in Munich per year, from 2000 to 2012

Figure 16: Effective, average interest rate for mortgage credits in Germany versus condominium price (Index 2000=100), from 2000 to 2012

Figure 17: Migration balance in 1000 inhabitants versus condominium price (Index 2000=100), from 2000 to 2012

Figure 18: Consumer price index versus condominium price (Index 2000=100), from 2000 to 2012

Figure 19: Relative deviation of condominium price to long-run-equilibrium in %, from 2000 - 2012

Figure 20: Illustration of condominium price deviation from calculated fundamental value, from 2000 - 2012

Figure 21: Condominium housing prices per m² per district of Munich, 2000 - 2012

Figure 22: Rawdata of time series gdpcap with labelled time span 1974 - 1996

Figure 23: Time series GDP with imputed and estimated values for the year 2012/ 2013, Note: 90% and 95%- confidence interval, part 1 of 2

Figure 24: Time series GDP with imputed and estimated values for the year 2012/ 2013, Note: 90% and 95%- confidence interval, part 2 of 2

Figure 25: Condominium price per m² in Munich, real, from 1975 to 2013

Figure 26: Average rowhouse price in Munich, real, from 1975 to 2013

Figure 27: Gross domestic product of Munich, from 1975 to 2013

Figure 28: Consumer price index of Germany, from 1975 to 2013

Figure 29: Effective, average interest rate for mortgage credits to private households in %, from 1975 to 2013

Figure 30: Flat stock of Munich, from 1975 to 2013

Figure 31: Residential building stock of Munich, from 1975 to 2013

Figure 32: Population segment of ages from 30 to 55 of inhabitants living in Munich in % to total population, from 1975 to 2013

Figure 33: Population density measured in inhabitants per hectare in Munich, from 2000 to 1975

Figure 34: Unemployment rate in Munich in %, from 1975 to 2013

Figure 35: Condominium price per m² on district-basis of Munich, from 2000 to 2012

Figure 36: Condominium price per m² in the districts of Munich, from 2000 to 2012

Figure 37: Permits for residential buildings on district-basis of Munich, from 2000 to 2012

Figure 38: Permits for residential buildings in the districts of Munich, from 2000 to 2012

Figure 39: Number of inhabitans subject to social insurance in Munich on district- basis in Munich, from 2000 to 2012

Figure 40: Number of inhabitans subject to social insurance in Munich in the districts in Munich, from 2000 to 2012

Figure 41: Proportion of 30 - 55 year olds in %, in comparison to total population of Munich on district-basis, from 2000 to 2012

Figure 42: Proportion of 30 - 55 year olds in %, in comparison to total population of Munich in the districts, from 2000 to 2012

Figure 43: Proportion of 30 - 55 year olds in %, in comparison 30 - 55 year olds of Munich on district-basis, from 2000 to 2012

Figure 44: Proportion of 30 - 55 year olds in %, in comparison 30 - 55 year olds of Munich in the districts, from 2000 to 2012

Figure 45: Migration balance of Munich per year on district-basis, from 2000 to 2012

Figure 46: Migration balance of Munich per year in the districts, from 2000 to 2012

Figure 47: Migration balance of foreigners in Munich per year on district-basis, from 2000 to 2012

Figure 48: Migration balance of foreigners in Munich per year in the districts, from 2000 to 2012

Figure 49: Migration together in % of corresponding annual migration in total in Munich on district-basis, from 2000 to 2012

Figure 50: Migration together in % of corresponding annual migration in total in Munich in the districts, from 2000 to 2012

Figure 51: Migration together in % to total population in Munich on district-basis, from 2000 to 2012

Figure 52: Migration together in % to total population in Munich in the districts, from 2000 to 2012

Figure 53: Population density in hectare of Munich on district basis, from 2000 to 2012

Figure 54: Population density in hectare of Munich in the districts, from 2000 to 2012

Figure 55: One person households to total households in Munich in % on district- basis, from 2000 to 2012

Figure 56: One person households to total households in Munich in % in the districts, from 2000 to 2012

Figure 57: Average number of people per household in Munich on district-basis, from 2000 to 2012

Figure 58: Average number of people per household in Munich in the districts, from 2000 to 2012

Figure 59: Foreigners to total population of Munich in % on district-basis, from 2000 to 2012

Figure 60: Foreigners to total population of Munich in % in the districts, from 2000 to 2012

Figure 61: Housing starts in Munich in total p.a. on district-basis, from 2000 to 2012

Figure 62: Housing starts in Munich in total p.a. in the districts, from 2000 to 2012

List of tables

Table 1: Variable definitions and descriptive statistics: Whole city perspective

Table 2: Full model OLS for condominium per m² (1) and average rowhouse price (2), first time use, real

Table 3: Model according to variable selection: OLS for condominium per m² (1) and average rowhouse price (2), first time use, real

Table 4: Devlopment of population according to districts from 2000 - 2012

Table 5: Comparison of one person households to population density per hectare, from 2000 to 2012

Table 6: Development of real condominium housing price per m² (first time use) within the districts of Munich, from 2000 to 2012*

Table 7: Variable definitions and descriptive statistics: District perspective

Table 8: FE- (1) versus RE-Model (2)

List of units

Squaremeter 1 m² = 1 meter x 1 meter

Hectare 1 ha = 10.000 m², or 100 meter x 100 meter

List of data media

Abbildung in dieser Leseprobe nicht enthalten

1 Introduction

The burst of the US-real estate bubble and the corresponding global financial crisis has started people all over the world wondering, whether a real estate bubble exists in their own country.1 After neglecting the existence of bubbles over the last decades, the burst of bubbles in the neighboring European countries Ireland and Spain in combination with recent increases of housing prices in metropolitan areas in Germany has brought this complacence to an end.2 Now there are more credible concerns than ever before that a real estate bubble might occur in Germany.

In order to attenuate the financial markets and to preserve liquidity, central banks all over the world including in the European Union (EU), and hence Germany, have lowered interest rates of capital and money markets. This pretentious, yet necessary move influences the risk aversion of all market participants.3 Even if the basic condi- tions for residential real estate seem to be rather positive again, prices in the city of Munich are at an historical high. Particularly since the financial crisis, a substantial recovery can be recognized, by assessing rising job occupation and at the same time falling unemployment.4 But does this recovery justify prices as at an all-time high for real estate in Munich? History shows that Munich never had a balanced housing market since the end of the Second World War.5

To any individual, a purchased condominium or house is generally one of the great- est investments of for their whole lifetime, if not the greatest.6 Hence, the importance of real estate’s role it is not to be disregarded, neither for the economy, nor the indi- vidual. Real estate can, compared to other factors of production, be understood as the basis-factor of all human existence. Real estate is a non-substitutional as well as nec- essary good. Furthermore, Germany’s real estate market acts as the greatest national market contrasted with the rest of the EU.7

Even though, a lot of effort has been put towards fixing the housing situation, Mu- nich has evolved to be the city with the highest rental prices as well as the highest ground prices in the country.8 This might not come as a surprise, since Munich is one of the most attractive cities, when it comes to real estate.9 But this is not the only reason why prices have skyrocketed into unprecedented heights, especially within the last few years. The uncertainty of the future development of the EU, along with the scarce choice of lucrative investment opportunities due to a historically low rate of interest, are possibly the most vital indicators for a possible start of a bubble. Apart from the loose monetary policy, also the intensifying shortfalls in housing supply and rising annual migration illustrate reasons for rising house prices.10 This is the reason why investors from all over the world see the real estate market of Germany and also Munich as a decent investment and therefore must expect future rising prices.11

The aim of this master thesis is hence, to examine the residential property market in one of Germany’s real estate hot-spots: Munich. A closer view will be taken at the development of the historical housing prices for condominiums as well as rowhouses in between the years 1975 and 2013. Furthermore the 25 districts of Munich will be scanned and evaluated in order to determine whether a local real estate bubble might exist in the city of Munich.

The research questions in this master thesis are:

1. Is there a real estate bubble in Munich as a whole?
2. Is there a real estate bubble within the 25 official districts of Munich?

The following Chapter 2 of this master thesis first of all lists, describes and justifies the data acquired in subsection 2.1. It then compiles previous definitions of housing bubbles in other literature, in order to develop a sound definition of a housing bubble in subsection 2.2. Afterwards a stock-flow model based on the article of Kajuth, F. et al. (2013) is presented, which will later, be used to calculate the fundamental value of the housing prices of Munich in subsection 2.3.

Chapter 3 will provide a literature review of the most important published infor- mation on this topic, summarizing the content as relevant to this thesis. In chapter 4, a city-wide perspective of Munich ranging from the years 1975 -2013 is presented. In subsection 4.1, the most important historical factors driving the de- velopment of Munich’s housing market are highlighted. These factors include the demographic developments, the building and housing development and housing pur- chase- as well as rental prices. Subsection 4.2 calculates characteristic ratios for the interpretation of the housing market, such as the price-to-rent ratio, the price to in- come ratio and the affordability index. This is followed by a concluding time series regression for condominium as well as rowhouse prices in subsection 4.3, where a possible local real estate bubble within Munich is investigated. This is done by pre- senting the time series model, followed by the actual analysis.

In chapter 5, a district-based perspective is presented which examines all 25 official districts of Munich, covering the years 2000 to 2012. First, this master thesis will take a look at the districts themselves in subsection 5.1. This view includes a closer look at the development of the population, the building and housing and the prices of condominiums in within each district. This is followed by an empirical panel analysis of all the district data gathered in subsection 5.2. The panel model will be presented first, followed by a cross section versus time series dimension, and then the statistical analysis. Finally, subsection 5.3 gives insight how the future will most likely be in terms of pricing, housing and demographic changes.

Finally, chapter 6 will summarize the core findings of this thesis, followed by a conclusion and outlook over the market.

2 Scope and limitations of research

2.1 Data collection techniques

In order to conduct an appropriate analysis and to contextualize this investigation, past results of existing academic literature were summarized and compared. The arti- cles found not only cover the topic of foreign housing bubbles, but also the German development of pricing bubbles and possible ways to measure those. Since this topic is still rather emergent, articles do not range very far into the past. A systematic data- base search with various search runs was conducted, in order to find relevant sources. Furthermore, especially in terms of the local data for Munich, several state agencies along with companies were contacted. This was not only helpful, but necessary, since data in the real estate market place is rarely complete and recorded properly.

Data for buildings and housing, population, economy and labor market of Munich was supplied by the “Statistisches Amt München” and the “Bayrisches Landesamt für Statistik und Datenverarbeitung”. This was indispensable to the ability to complete the data otherwise collected. Data for housing and renting prices in Munich have kindly been granted by the “Bulwiengesa AG” as well as the “Sozialreferat München” and the “Immobilienverband Süd”. Data for interest rates and the consumer price index have been supplied by the “Deutsche Bundesbank”. Finally the “Referat für Stadtplanung und Bauordnung” has contributed to this thesis in terms of providing future building plans for the city of Munich.

On the basis of this data, it was decided to split this thesis into two parts: A whole city perspective and a district perspective of Munich. Amongst other causes, that was due to the district data only being available from the year 2000 and onwards. Hence it was chosen to enlarge the horizon of investigation back to 1975, which was only possible on a whole city basis. Furthermore, whole city data is the most recent, so that it was possible to acquire data from the year immediately past, i.e. 2013. Unfor- tunately, this was not possible for the district perspective, as the most recent data only ranges up to the year 2012.

Overall, efforts were made to estimate as little data as possible, to avoid bias and therefore to produce the most accurate results.

2.2 Definition of a real estate bubble

Since the comprehension of a real estate bubble is not quite clear, the definitions of leading articles have been compiled in order to give the reader a conception as well as an overview on the main aspects of pricing bubbles. This step is necessary in order to be able to answer the question whether a local real estate bubble exists within Munich. Here the article of Kajuth, F. et al. (2013) as well as the article of the Deutsche Bundesbank (2013a) monthly report of October 2013 is emphasized.

Even though, due to the latest developments in international real estate markets, the topic of bubbles might appear to be relatively young to some, history shows that pric- ing bubbles are not a recent phenomenon at all. One of the preeminent examples would be the Dutch tulip pricing bubble, which took place between the years 1634 - 1637. During that time, tulip bulps became enormously popular and, due to a lack of supply, prices increased dramatically. The bubble finally burst, and the price of tulip bulps collapsed.12 Garber (1990, p. 37) calls this bubble “one of the most famous” bubbles.

The literature identifies essentially two different kinds of pricing bubbles: The ra- tional and the irrational bubble.13 But the terms “rational” and “irrational” do not account to describe the bubble itself; they rather describe the actions of market par- ticipants in terms of behavior and expectation. If investors choose to make rational decisions, and hence a bubble develops, the result is a rational bubble. In this case the rationality of market participants does not imply that the price of an asset has to be equal to its fundamental value.14 In theory in an efficient market, the rational equi- librium price of any asset of today is the equal to the future expected sum of dis- counted cashflows.15 But since it is uncertain how future cashflows develop, in case of a rational bubble, investors are willing to pay higher prices for assets, because, according to their rational expectation, a higher price increase in the future is pro- jected. Articles refer to rational bubbles as stochastic, information based and rational speculative bubbles.16

Correspondingly, if investors behave in an irrational manner, and a bubble evolves, it is an irrational bubble. Being that way, due to the success of some investors, which attracts public attention, other investors start to enter the marketplace and start to bid up prices. This means that prices deviate strongly from their fundamental values, but market participants think that the overvalued price still reflects its fundamental value, and continue to bid up prices. Several articles also refer to this phenomenon as herdlike behavior, behavioral based-, irrational speculative bubble.17

But does any bubble evolve in the same way as another, and can a real estate bubble distinguished from other pricing bubbles? When there is a high increase in property pricing, how can one tell whether this is due to a lack of supply or a high increase in demand and whether or not this bubble is unsustainable? One of the first authors to provide a general definition of housing bubbles was Stigliz (1990, p.13):

If the reason the price is high today is only because investors believe that the selling price will be high tomorrow — when “ fundamental ” factors do not seem to justify such a price — then a bubble exists

Figure 1: Fundamental value based perception of a real estate bubble

Abbildung in dieser Leseprobe nicht enthalten

Source 1: Own illustration, based on Rombach, T. (2011, p.47)

As it can be seen in Figure 1, the difference of the actual market value (MV) sub- tracted from the fundamental value (FV), the unjustified value, could be the sign for a real estate bubble. In t0 both FV and MV are equal to each other, meaning that the price for real estate is on a justified fundamental basis. Even though the MV rises, in t1 the FV unhooks itself from actual MV. This could be due to a loose monetary poli- cy and resultant an increased speculation. In t2 the MV has reached a totally unjusti- fied price level as opposed to its true fundamental value. This high increase might be a result of ongoing speculation on ever-growing prices; more and more people expect an ongoing increase of real estate prices in the future. At this point just one piece of credible bad news could be enough for the unjustifiable increased FV to collapse like a house of cards.18 Investors are now trying to sell their dearly purchased real estate, but they need to be fast, as prices react immediately and decrease quickly. This movement will lead to t3, where the MV now equals the FV again, meaning that the fundamental value is represented by the actual market value.

But what exactly are fundamental factors of real estate? A bubble might be indicated, if increases in prices of real estate cannot be explained through a change of the fundamental value. Hence, the price is not justified by a change in the economic variables, which determine the fundamental value price.19

Nevertheless real estate bubbles occur more rarely as opposed to bubbles in other market segments, such as the stock market.20 Moreover, no pricing bubble develops and behaves exactly the same as any other. Finally, it is to be noted that even strong increases in housing prices over several years do not necessarily have to result in a bubble. Hence, it is often difficult to recognize the formation of a bubble in time.21 In order to provide the reader of this master thesis with a possible approach to recog- nizing the development of housing bubbles, a closer look is warranted at the articles of Katjuth, F. et al. (2013) and the Deutsche Bundesbank (2013a) in the next chapter. The influence as well as change in macroeconomic variables on housing prices will be empirically examined, the fundamental value will be calculated, and the deviation to actual prices will be examined in order to determine whether a bubble in Munich might exist.

2.3 Fundamental value - The stock-flow model

The following definition of the stock-flow model is closely related to the derivation of Kajuth, F. et al. (2013) as well as the monthly report of the Deutsche Bundesbank (2013a).

The stock-flow model describes the housing market by using an equation of motion of the dwelling stock, a demand function for flats and an adjusted equation for hous- ing prices as well as real estate investments, whenever demand deviates from the existing supply.

The development of the dwelling stock st therefore performs as

Abbildung in dieser Leseprobe nicht enthalten

(1)

The housing supply at the point of time t is equal to the housing supply at the point of time t-1 adjusted by the depreciation δ plus the investment bt − 1 at the point of time t − 1.

The demand of housing xt assumes, that it is correlated positively with the income yt and negatively with the rent mt as well as other factors zt (demographic as well as labor market factors).

(2)

Abbildung in dieser Leseprobe nicht enthalten

The current income can be seen as to measure the capacity of mortgages or the mar- ket cycle. Furthermore, the current income can influence affordability as well as the loan to value. Next to the lending value of a house, banks often use measures of af- fordability in order to evaluate loan to values for mortgages. Hence, affordability can be calculated by comparing the repayments of loan and the fixed part of current re- turns.

The single-period real rents mt are related to the real housing prices because of the asset pricing conditions for housing:

Abbildung in dieser Leseprobe nicht enthalten

(3)

The predefined return of the rent of a housing unit rt is equal to the quotient of the expected return of housing prices at the next point of time t+1, Etpt+1 plus the rent mt at the point of time t minus the current housing price pt over pt. Through rearranging algebraically according to pt,

(4)

Abbildung in dieser Leseprobe nicht enthalten

as well as forward iterations

(5)

Abbildung in dieser Leseprobe nicht enthalten

and the recognition of the condition of transversality22 the equation delivers

housing prices as sum of expected discounted rent returns in future.

(6)

Abbildung in dieser Leseprobe nicht enthalten

Furthermore, another two assumptions are:

1. Rent is a fixed return and grows sustainably at a rate of[Abbildung in dieser Leseprobe nicht enthalten]

Abbildung in dieser Leseprobe nicht enthalten

(7)

2. Future rent payments are discounted with a long-term median rate of [Abbildung in dieser Leseprobe nicht enthalten]

Abbildung in dieser Leseprobe nicht enthalten

(8)

If those assumptions are incorporated into the previous equation and this equation rearranged to [Abbildung in dieser Leseprobe nicht enthalten] as well as using the sum formula under the condition [Abbildung in dieser Leseprobe nicht enthalten] one receives

Abbildung in dieser Leseprobe nicht enthalten

If one takes the logarithm of this equation and the logarithm laws are applied[Abbildung in dieser Leseprobe nicht enthalten] and puts this result into the equation for the housing demand with the law of equilibrium,

[Abbildung in dieser Leseprobe nicht enthalten](10)

meaning that demand is equal to supply, one receives the inverted curve of housing demand with short term inelastic supply:

[Abbildung in dieser Leseprobe nicht enthalten](11)

With[Abbildung in dieser Leseprobe nicht enthalten]. Henceforth the price [Abbildung in dieser Leseprobe nicht enthalten]can be described through observable economic factors und can deviate from the current housing price, for instance be- cause of shocks. Considered, that the current housing prices converge back to a level that is merely determined by economic factors [Abbildung in dieser Leseprobe nicht enthalten] can be viewed as the fundamental value.

With the stock-flow model defined and the calculated fundamental value determined, the deviation of actual housing prices can be measured. A positive deviation (FV<MV) will therefore represent an overvaluation and correspondingly a negative deviation (MV<FV) will represent an undervaluation of real estate. However the question arises what exact point of an overvaluation is necessary in order to fulfill the definition of a housing bubble. Unfortunately there is no through answer to re- spond to that question adequately. This is also due to housing bubbles only being able to be identified ex post, meaning after they have burst.23 Nevertheless a more than minor overvaluation of real estate represents the most important basis of a po- tential housing bubble. Therefore the greater the overvaluation, the greater the risk of a housing bubble is.

3 Literature Review

In the following part, a short insight will be given, how leading articles in Germany have treated, understood and identified the topic of real estate bubbles. Apart from a short summary, the main quintessences and results of each article will be presented. The articles presented in the following are listed according to importance to this mas- ter thesis.

With regards to the topic of bubbles, Kajuth, F. et al (2013) and the monthly report of the Bundesbank (2013a, October) use a stock-flow model for the housing mar- ket in order to estimate the relationship between housing prices and explanatory mac- roeconomic variables in Germany. In order to do so, Kajuth, F. et al (2013) use a regional panel dataset for 402 administrative districts of Germany for the years 2004 to 2012 and a subset of 93 towns as well as cities for the years 1996 to 2012. The report of the Deutsche Bundesbank (2013a) uses the same regional panel dataset, but only for the years 2004 to 2010. For both articles, the data for housing prices was supplied by a private firm named Bulwiengesa AG. In order to estimate the funda- mental value of housing prices for single family homes as well as apartments, sepa- rate equations were regressed. The explanatory variables used are the housing stock per capita, income per capita, rate of unemployment, real long-term interest rates on mortgages, a demographic measure, the population density as well as growth expec- tations on housing prices. The monthly report of the Deutsche Bundesbank (2013a) uses the same variables, but also includes the population aged between 30 to 55, as it is common in this age cohort to buy real estate. Furthermore, the growth expectations of real GDP were included as variable, instead of growth expectations of housing prices, as in the approach of the article of Kajuth, F. et al. (2013). Both articles use the regression residuals as a measure for deviations of actual house prices.

In result, Kajuth, F. et al. (2013) find that the single family housing prices moved along the estimated fundamental value, but the prices of apartments significantly exceeded their estimated fundamental value in recent years. For Germany as a whole, Kajuth, F. et al. (2013) determine a moderate overvaluation of five to ten percent in terms of apartments. In contrast the article of the monthly report of the Deutsche Bundesbank (2013a) discovered price increases of housing prices in metropolitan areas of up to 20%, however a substantial overvaluation for Germany as a whole was empirically not detected. Both of the articles, conclude that the pricing-pressure in terms of housing is likely to continue in the short term, as supply will not be enough to satisfy the numbers of migration and therefore increased demand.

The DIW (2012b) presents a study, which analyzes the prices of new flats for rent and prices for condominiums available for sale in Berlin to create an index. Rent contracts existent at the time of the study are not being observed. In contrast to Kajuth, F. et al. (2013) and the report of Deutsche Bundesbank (2013a), the underly- ing data is acquired from well-known internet portals such as www.immobilienscout24.de, www.immowelt.de as well as www.immonet.de. Since the real estate offers in these portals compromise great detail in data, the DIW was able to create a qualitative adjusted index. Compared to Kajuth, F. et al. (2013) and the report of the Bundesbank (2013a), the time span of observed data comprised only June 2011 to March 2012. The DIW produced a regression, which is taking the quali- ty of flats into account. Hence position (ground floor to 21. Floor), building age (1900 to after 2000) and type of condominium (apartment to rowhouse) are all thor- oughly investigated. In addition, variables such as space in m², number of rooms, available kitchenette/-celler/-garden/-elevator and parking space are observed in the equation. Altogether 37 variables are being investigated. Hence, in contrast to Kajuth, F. et al. (2013) and the report of the Bundesbank (2013a), this regression includes a fairly comprehensive suite of variables.

In conclusion, the results reflected the expectations beforehand. There is no sign of a housing bubble within Berlin, even though there were thorough increases in housing prices in the past years. Those increases were derived to the stagnation of prices sev- eral years before the spectated period of time. The study is able to analyze single districts of Berlin, as well as the whole city and regions. New data can easily be inte- grated into the existing model, without a large interval of time. Another advantage of this model lies in the great variety of detail, hence observed variables.

In the final report of the DIW (2011), an extensive analysis of bubbles for chosen OECD countries such as Australia, Canada, France, Germany, Italy, Japan, Nether- lands, Portugal, Spain, Sweden, UK, USA and Switzerland is conducted. Hence, there are several countries being investigated in contrasted to the article of Kajuth, F. et al. (2013), the report of the Bundesbank (2013a) and the study of the DIW (2012b). The observation spans for the countries range from 1969 to the end of 2009, however in Germany the span of observation ranges from 1991 to the end of 2009. The emphasis of this study lies on real estate markets as well as stock markets. The criteria of pricing-bubbles as well as their chronology for all the markets are listed. The equation for real estate prices uses variables such as: real GDP per capita (as a proxy for available income), population, real-interest, and lastly, the degree of urban- ization (proportion of urban population as opposed to the whole population). Except for the former variable, the other variables mentioned were also used by Kajuth, F. et al. (2013) and the report of the Bundesbank (2013a). In the final part, the report’s aim is to develop an early warning system using econometric systems. The two mod- els that point out is, on one hand, a signal approach and on the other hand, a probit and logit approach, which is also a different approach than used in the articles of Kajuth, F. et al. (2013) and the report of the Bundesbank (2013a). A panel analysis is used for all estimations, which again corresponds to the other articles.

In conclusion, the article states that a chronology of bubbles is necessary beforehand, in order to develop a substantial warning system. It points out that the probability for a speculative pricing bubble increases, in case there is an expansive monetary policy as well as facilitated loans. The study’s constructed warning system does in fact detect a real estate bubble in Germany in between the years 1992 - 1994. This bubble is derived to the reunion of Germany, whereby an extensive capital investment in residential as well as commercial estates took place.

The paper of the Deutsche Hypovereinsbank (2012) also analyzes the housing mar- ket of Germany. The span of observation ranges from 1995 to 2011 and the devel- opment of flat rents/ condominiums and row houses are the focus, similarly to Kajuth, F. et al. (2013) and the monthly report of the Bundesbank (2013a). However, in contrast to the other articles, also the future demographic development is also highlighted. The emphasis lies in the first part of the study: the inventory of residen- tial housing is summed up and analyzed. Especially the differences between Eastern and Western Germany are accentuated. Correspondingly the development of relevant influential factors on supply and demand is examined. In the second part of the study, a potential housing bubble in Germany is discussed by making use of and in- terpreting common ratios such as the price-to-rent and price-to-income ratio as well as by calculating the affordability index.

The study concludes that the outlooks of the residential housing markets are generally positive. Moreover, the study determines that there are no signs for a speculative housing bubble in Germany as a whole; however tendencies towards a bubble, especially in regional prospering cities like Munich or Berlin, cannot be suspended. Hence, as opposed to Kajuth, F. et al. (2013), the report of the Bundesbank (2013a) as well as the study of the DIW (2012b), there was no empirical approach of investigating whether there is a bubble in Germany.

Apart from other factors that challenged the financial system in the year 2012, the Financial Stability Review of the German Bundesbank (2012) analyzes the German Banking systems, with its low interest rates as well as the German housing market, which is, compared to the previously mentioned articles and papers, not an approach to measure bubbles. The emphasis lies on mortgage credit lending and debt sustainability of German households. Not only the residential market is observed, but also the commercial real estate market, where the Deutsche Bundesbank sees a moderately higher threat in contrast to residential real estate.

The article concludes that the amount of mortgage credits will keep on rising in the future. However, the Deutsche Bundesbank does not see a threat for a housing bubble developing. This certainty can be attributed to the soundness of the German policy based lending, where most mortgage credits have a fixed interest rate for at least 5 years and have an equity ratio. Yet the Deutsche Bundesbank moderately warns about price-overheating in regional sub-segment-markets.

Apart from other factors that challenged the financial system in the year 2013, the Financial Stability Review of the German Bundesbank (2013b) analyzes the fi- nancial stability of Germany due to the European debt-crisis and also investigates the default-probabilities of mortgage credits in the German Market, which is similar to the financial stability review of the German Bundesbank (2012). The review comes to the conclusion that there is still a risk attributable to the Euro- pean Debt crisis regarding the financial stability in Germany. The countries in the EU might have reconciliated their balance of current accounts, and are headed into the right direction, but the national debts keep on rising and therefore expose the weakest point in the system. However, in the worst case of mortgage credit defaults, the German Banks would be able to sustain the damage caused. Nevertheless, this result does not take the recent years 2012 and 2013 into account, and therefore can- not make a concrete statement. Furthermore, as opposed to the financial stability review of 2012, the Deutsche Bundesbank this time puts more emphasis on warning about a potential price overheating, again especially in congested areas of Germany.

4 City based perspective from 1975 to 2013

In this part of the master thesis the Bavarian capital Munich will be analyzed as a whole in between the years 1975 and 2013.

In 4.1 the historical development will be highlighted and this part will aim at giving a better understanding of developments in Munich between 1975 and 2013 of how population in subchapter 4.1.1, building and housing in subchapter 4.1.2, housing prices and renting prices in subchapter 4.1.3 and lastly how the interest rate for mort- gages to private households in subchapter 4.1.4. Furthermore, outliers will be ex- plained using older as well as current literature from Munich as well as the rest of Germany.

In 4.2 characteristic ratios such as the price-to-rent ratio in subchapter 4.2.1, the price-to-income ratio in subchapter 4.2.2 and the housing affordability index in sub- chapter 4.2.3 will be calculated and interpreted briefly for the city Munich. The for- mer two indicators unfortunately were only able to be calculated for the years 2000 to 2013, because data did not reach back further. Nevertheless whole city data is be- ing used.

In subsection 4.3 an empirical time series investigation will be conducted. In order to do so, first the variables and the model are presented in subchapter 4.3.1. In a second step condominium prices as well as rowhouse prices will be regressed and funda- mental values of both will be calculated on the basis of the calculation of Kajuth, F. et al. (2013) as well as the monthly report of the Deutsche Bundesbank (2013a) in subchapter 4.3.2. To conclude, the relationship between the fundamental value and the actual prices for condominiums and rowhouses will be investigated, analyzed and interpreted.

4.1 Historical development

4.1.1 Population and labour market

Figure 2: Population development within Munich, from 1975 to 2013

Abbildung in dieser Leseprobe nicht enthalten

Source 2: Own illustration, data from Statistisches Amt M ü nchen

Looking at Figure 2, it is perceptible that the total population of Munich has experi- enced a 150.000 head increase within the investigated period of the years 1975 - 2013, reaching a total number of inhabitants of 1.464.962 in 2013. Surprisingly, the population of Germans, even if compared to the year 1975, has declined. On the oth- er hand, the share of foreigners within that time span rose about 67%, meaning that as to date 25% of all Munich inhabitants are of foreign origin. Hence the increase of the total population can only be attributed to foreign immigration. Besides, in com- parison to other metropolitan areas of Germany, Munich has a leading role in the substantial growth of its population.24 In the year 1989, the city experienced a net- migration of 57.000 people, which can be explained by the fall of the wall in Berlin. However, in comparison to the net migration to West-Berlin in 1989, 44% of people have settled in Munich, since 130.845 have migrated from the former East to West- Berlin.25 But the increase of foreigners in the year 1990 cannot solely be attributed to the opening of the borders of East Berlin. In 1990 the Yugoslavia crisis evolved to be one of the most significant determinants explaining the increase in foreign residents, with a peak migration of 78.8% in the year 1992 as opposed to the year 1990.26 The decrease of both Germans and foreigners in Munich in the year 2000 is due to a change in the register of residents. Previously, inhabitants entitled to residence were counted, regardless of, whether Munich was their primary residence or their second- ary residence. From the 01.01.2000 only inhabitants of Munich were counted, whose primary residence was registered in Munich.27 Hence 67.320 prior inhabitants of Munich were not counted anymore.

The factor that contributes most towards the high rates of population growth would most certainly be the migration. 2010 was the first year, where migration balances started to pick up the pace, until the previous year, 2013. Especially in the recent years 2010 until 2013, firm figures in terms of migration could be discovered. Hence, only within those 4 years there was a balance of 88.519 people that migrated to Mu- nich. For comparison, within the years 2000 - 2009 the balance of migration was 99.919 people. However, with a share of 87.88%, foreign migration accounts for most of the increase of migration in the years 2010 to 2013.28

But not only migration contributed towards the high growth rates of the population of Munich. The steeper increase in 2005 can also be attributed to the German popula- tion of Munich increasing again, after having remained relatively static for the previ- ous five years. A live birth surplus of over 13.196 of 83.3% German origin in the year 2005 can be seen as a reason.29 While the rest of Germany worries about reces- sive rates of births, the trend of the ever rising birth surpluses in Munich keeps on rising. In 2012, the sixth year in a row, the city had an even greater birth surplus as opposed to the preceding years with 15.092 new live childbirths.30

Compared to other cities of Germany, Munich has, especially in recent years, shown the greatest growth of inhabitants. This fact demonstrates the popularity of the Bavarian capital. But such rapid development and growth also comes along with distinct challenges in terms of building.31

4.1.2 Building and housing

Figure 3: Residential building- and flat stock in comparison to annual permits in

Munich, from 1975 to 2013 *

Abbildung in dieser Leseprobe nicht enthalten

Source 3: Own illustration, data from Statistisches Amt M ü nchen

In Figure 3 one can clearly see that both stocks of residential buildings as well as the stock of flats have risen within the years 1975 to 2013.

Remarkably, prior to the year 1975, more flats were built every year as opposed to any year investigated in the period 1975 to 2013. Throughout the years 1952 and 1974 the mean average increase of stock added up to 13.087 flats per year. The peak of newly built flats occurred in the year 1972, where, due to the Olympic Games in Munich, 21.286 new flats were added to the stock.32 At the same time Munich expe- rienced a negative growth of population. Hence, Munich officials saw an excess in supply of housing, which was the reason for reforming the city land development plan (STEPL), where only a restricted number of new flats were approved, as there were flats withdrawn from the stock, for instance due to age. This can be seen in fig- ure 3, where there is a sudden drop in permitted flats in between the years 1975 and 1979. Around 1980 the new demand of housing was due to the former baby boomers now starting to be independent and looking for their own flats and houses. The de- mand was urgent, so that Munich’s officials developed the housing provision pro- gram (“Wohnungsraumbeschaffungsprogramm”), where it was determined that 7000 new flats should be built every year.33 This is reflected in the peak of 2.050 residen- tial permits in the year 1980. Furthermore 1983, another STEPL was set forth, in which, in the long-term, more flats were permitted to be built.34 This resulted in the highest number of permitted flats of 8.885 permits in the year 1986. Another peak can be found in the year 1990 with 7947 permits, which is most likely due to the reu- nification of Germany and the accompanied migration to Munich. Additional peaks of permits for flats can be found in the years 1994 with 7.615 followed by 2005 with 8.574 permits. The last peak in the investigated period of time is 2011 with 8417 permits. On the contrary, even though the housing provision program was developed, Munich’s housing starts did not reach the number of 7000 again, in any year investi- gated in the time span 1975 to 2013.35 Surprisingly enough, already in 1993 the GEWOS report forecasted the collapse of the Munich housing market to take place in the year 2000, due not only to the housing development progressing in an inappro- priate mean, but also the previous restrictions made by the officials of Munich.36

Overall, starting at 546.499 units, the stock of flats in Munich rose 40.52% to a total of 767.940* and starting at 100.451 units, the stock of residential buildings in Munich rose 35.71% to a total of 137.355 buildings within the time span of 1975 to 2013. Comparing the year 2010 to 2013*, an astonishing increase of 51.4% in terms of finished flats per year can be discovered.37 Especially the ongoing rising numbers of migration will not be able to be satisfied by the current building activity.38 The recent building activity can hence, for the most part, be viewed as catch-up effect for the low building activity of the previous years.39 This however cannot exclude the possible danger of a housing bubble within Munich.

4.1.3 Housing and renting prices

Figure 4: Average condominium and rent price for median flat of 75 m ² and row-

Abbildung in dieser Leseprobe nicht enthalten

Source 4: Own illustration, data from Bulwiengesa AG

By examining 25 metropolitan cities in Germany, the DIW (2012a, p. 4.) determined the median flat to be 75m² in size and with three rooms. Henceforth, when necessary, this also accounts for the results in this master thesis. Since average nominal prices of condominiums as well as rent were supplied by €/m², an average flat of 75 m² has been assumed. When investigating Figure 4, it is evident that all prices increased within the time span of 1975 to 2013. When comparing solely the year 1975 to 2013, a nominal increase of 258.57% for condominium prices, a nominal increase of 204.35% for rent prices and a nominal overall increase of 361.19% for average row- house prices can be discovered. The increase of average rowhouse prices being greater than condominium prices and rent prices might indicate a saturation of row- house prices in Munich. This suspicion might be confirmed by comparing solely the years 2000 to 2013, here condominium prices experienced an increase of 173.50%, rent prices an increase of 124.44% and average prices for rowhouses an increase of 133.96%, demonstrating that rowhouse prices did not rise as steeply in the last 13 years as opposed to prices for condominiums. Figure 5 provides a better understand- ing of pricing increases in Munich, illustrating the real growth rates.

Figure 5: Growth rate of condominium, rent and average rowhouse prices of Mu- nich, real, from 1975 to 2013

Abbildung in dieser Leseprobe nicht enthalten

Source 5: Own illustration, data from Bulwiengesa AG

In Figure 5, the real annual growth development of condominium and rent prices per m² as well as the average prices for rowhouses is being illustrated for the time span of 1975 to 2013. The consumer price index of Germany (CPI) was used to deflate the nominal values of condominium, rent and average rowhouse prices. The average rent price per m² shows the highest volatility, the average rowhouseprice per m² accounts for the lowest volatility. The peak of average rowhouse prices takes place in the year 1977 with 14.8%, for condominium prices in the year 1980 with 20.92% and for rent prices in the year 1989 with 11.09%. The current (2013) real increase in terms of pricing for condominiums was at 13.07%, for rowhouses 7.18% and for rent 2.19% in Munich. This might indicate that even though renting prices increased, they did not increase as much as they should have compared to condominium and rowhouse prices. A reason for that could be the rising amounts of welfare construction and wel- fare flatting.40 When prices of housing rise higher as opposed to prices of rent, it de- vours renting yields and real estate might become less attractive. Additionally, a pos- sible overvaluation of housing can be indicated.41 Especially in the recent years 2010 to 2013 this appears to be the very case.

[...]


1 Cf. DIW (2011), p. 4.

2 Cf. Dombret, A. et al. (2013), p.3.

3 Cf. Deutsche Bundesbank, (2012), p. 42.

4 Cf. Deutsche Hypo Markt-Analyse (2012), p. 4, 8.

5 Cf. Rutchy, O. (1988), p. 414.

6 According to the income- and consumption-sample of 2008 (“Einkommens- und Ver-

brauchsstichprobe”), the German households on average held real estate in the fair value of 236.100€, whereas the net average household income was around 27.504€ in the eastern part and

36.662 € in the western part of Germany.

7 Cf. Deutsche Hypo Markt-Analyse (2012), p. 4.

8 Cf. Rutchy, O. (1988), p. 414.

9 Cf. Bulwiengesa AG (2013), p. 2.

10 Cf. DIW (2012a), p. 13.

11 Cf. Deutsche Bundesbank (2012), p. 60.

12 Cf. Van der Veen, M. (2009), p. 3-10.

13 Cf. Löwer, A. (2003), p. 27-29.

14 Cf. Blanchard, O. et al. (1982), p. 1.

15 Cf. Holtemöller, O. (2010), p. 559.

16 Cf. Löwer, A. (2003), p. 29-30.

17 Cf. Shiller, R. (2000), p. 148f., Langsing, K. (2008), p. 26-28.

18 Cf. Deutsche Hypo Markt-Analyse (2012), p. 21.

19 Cf. DIW (2011), p. 5.

20 Cf. DIW (2011), p. 8.

21 Cf. Dombret, A. et al. (2013), p. 16.

22 With the forward iteration a term evolves, where the enumerator is the expected housing price at t+T and the denominator the discounting from t until T. The condition of transversality now states, that for t tending towards infinity, the discounting will be higher as opposed to the expected hous- ing price. Because of the existence of interest, the same amount would have a higher value (be- cause one would be able to invest at a risk free interest r). In case of real estate this means that the current value is determined through the current fundamental data as well as the future expected fundamental data. Hence, this specification needs to be made, in order to establish an equilibrium price (asset pricing model). Hence, if the housing price would deviate from the fundamental, the condition of transversality would be hurt.

23 Ahearne, A. et al. (2005), p. 2, Bhattacharya, U. et al. (2008), p. 1.

24 Cf. Rutchy, O. (1976), p. 79f.

25 Cf. Huss, E. (1994), p. 273.

26 Cf. Huss, E. (1994), p. 279, Breu, F. (1996), p. 359.

27 Cf. Statistisches Amt München (1999), p. 157.

28 Values based on own calculation of migration data supplied by Statistisches Amt München.

29 Cf. Kengeroglu, S. (2013), p. 16.

30 Cf. Kengeroglu, S. (2013), p. 18, 20.

31 Cf. Desch, G. (2011), p. 20.

32 Cf. Kreiling, H. (1993), p. 325, Rutchy, O. (1988), p. 414.

33 Cf. Dr. Dheus, E. (1980), p. 239.

34 Cf. Kreiling, H. (1993), p. 341.

35 Except for the year 2006, where, according to the Statistisches Amt München 11.640 belated regis- trations of housing starts in Munich led to the total of 15.908 housing starts.

36 Cf. Süddeutsche Zeitung no. 170, (27.07.1993), p. 27. See Appendix I.II: Estimated value of 2013, all other values are genuine.

37 Own calculation based on housing data supplied by Statistisches Amt München.

38 Cf. Deutsche Bundesbank (2013a), p. 30.

39 Cf. Bürkle, T. (2013), p. B5.

40 Cf. Dombret, A. et al. (2013), p. 10.

41 Cf. Deutsche Bundesbank (2012), p. 61.

Details

Pages
99
Year
2014
ISBN (eBook)
9783656702818
ISBN (Book)
9783656703945
File size
3.4 MB
Language
English
Catalog Number
v277647
Institution / College
University of Regensburg – Lehrstuhl für Immobilienfinanzierung (International Real Estate Business School, University Ratisborn)
Grade
1,3
Tags
munich

Author

Share

Previous

Title: Are properties in Munich partially overvalued? An investigation of local real estate bubbles and price development within the districts