Housing Cycles in the Major Euro Area Countries

TL;DR: In this paper, the authors investigated the relationship between housing cycles among the four major euro area countries (Germany, France, Italy and Spain) over the sample 1980Q1-2008Q4.

Abstract: The recent burst of the house price bubble in the United States and its spillover effects on real economies worldwide has rekindled the interest in the role of housing in the business cycle. In this paper, we investigate the relationships between housing cycles among the four major euro area countries (Germany, France, Italy and Spain) over the sample 1980Q1-2008Q4. Our main findings are that GDP cycles show a high degree of comovement across these four countries, reflecting trade linkages, but much weaker ones for housing market cycles, where idiosyncratic factors play a major role. House prices are even less related than quantities across countries. We also find much stronger relationships in the common monetary policy period.

Summary

1 Introduction

• The recent burst of the house price bubble in the United States and its spill-over effects on real economies worldwide have reinforced the interest in the role of housing in the business cycle.
• Even though housing market imbalances in Anglo-Saxon countries have been stronger than in European markets, cross country heterogeneity in price and volume developments raises two issues: firstly, the interaction between housing variables and the macroeconomy within each national economy and, secondly, the role of synchronization across countries of housing variables.
• BANCO DE ESPAÑA 10 DOCUMENTO OCASIONAL N.º 1001 such as the number of building permits, the number of housing starts and employment in the construction sector.
• Specifically, for each country pair, the authors compute cross-correlation coefficients and cross-concordance indexes and carry out a lead-lag analysis of turning points.
• The authors results are consistent with previous empirical analysis supporting a broadly common GDP growth cycle among the four major euro area countries, with the German cycle presenting stronger idiosyncratic features, including those related to the reunification process.

2 Methodology

• A large number of procedures have been developed in the literature to carry out decompositions of aggregate output into a trend — which accounts for long term growth — and a cyclical component — which measures short-term deviations from this trend [see e.g. Mills (2003)].
• In turn, the Hodrick and Prescott (1997) filter does not have an oscillatory gain, but being a high-pass filter damps cyclical fluctuations with high periods and leaves short-run cycles barely untouched.
• Butterworth band-pass filters of the tangent3 can be expressed in the time domain as symmetric, two sided filters in the lag (L) and forward (F) operators given by dddd dd FFLLFL FLFLBPF )1111( )11,( 2222 22 Note that larger values of d produce sharper filters, so they provide a better approximation to the ideal filter.
• Figure 1 Estimated cycles for GDP, residential investment and nominal and real house prices (BFT filter) 3.

3 Results

• The authors describe the cyclical comovements across the four major euro area economies.
• The main focus of the analysis is the housing market and the authors distinguish between housing related real and nominal variables .
• The authors refer to synchronization as the degree to which two or more variables co-move contemporaneously and measure it for every country pair by the contemporaneous correlation coefficient and by the concordance index between their respective cycles.
• Specifically, this measure is defined as pRPR /11 where |R| denotes the determinant of the correlation matrix of the variables of interest and p denotes the number of variables.
• The authors also consider a sample for the common monetary policy period.

3.1 Correlation analysis

• First, the authors consider aggregate GDP cycles, then housing volumes construction cycles, and, finally housing price cycles.
• BANCO DE ESPAÑA 14 DOCUMENTO OCASIONAL N.º 1001 a) AGGREGATE ACTIVITY CYCLES.
• Cycles in non-residential construction investment show a moderate correlation of the French cycle with respect to the other countries (0.45 with Germany, 0.32 with Spain, 0.26 with Italy).
• These are even more country-specific (average pairwise correlation 0.10) than volume housing cycles.
• Results from cross-correlation analysis suggest that the negative contemporaneous correlation of the French cycle with Italy and Spain reflects the mildly leading nature of the former with respect to the latter.

3.2 Concordance analysis

• After discussing the results based on correlation coefficients, the authors focus on measures of concordance based on the sequence of turning points in the cycles previously .
• Once turning points have been identified, for each variable of each country i, the authors compute a binary variable (Sit), the so called reference cycle, such that Sit is equal to 1 during a descending phase of the cycle (that is, between a peak and a trough) and zero otherwise.
• As in the above correlation analysis, the authors compute the contemporaneous and the cross-concordance indexes for h such that 4,2,1,1,2,4 h , and they retain the lead (lag) that maximizes the CCI.
• Regarding residential investment cycles, concordance indexes values confirm the results from correlation analysis that synchronization is lower than for GDP cycles.
• Furthermore, Italy lags France (two quarters) and Spain (four quarters), with a high degree of concordance (0.74 and 0.71, respectively), a finding in line with results from the correlation analysis.

3.3 Convergence or divergence in the monetary union?

• The authors analyse the change in synchronization in the more recent 1999-2008 subsample, referred to as the EMU sample hereafter.
• Particularly striking is the increase in synchronization measures for GDP comovements, a finding highlighted in much of the existing empirical literature [e.g. Cabrero et al. (2004)].
• Results for house prices are mixed, depending on the measure considered.
• Candelon et al. (2009) have extended this test in order to take a small number of cycles into account.

4 Conclusions

• Recent years have seen an explosion of papers focussing on housing price cycles and, to a much lesser extent, housing volume cycles.
• The authors find that, in the four major euro area countries, GDP cycles show a high degree of comovement, most likely due to trade linkages, although idiosyncratic factors play a larger role, particularly in Germany.
• Cross country comovements are mostly contemporaneous, but developments in Spain tend to lead those in Germany, Italy and France by one or two quarters.
• In contrast, comovements are substantially weaker for housing market cycles, where country-specific or local variables, such as land availability or regulation, play a major role.

Figures

Figure 3

Figure 4

Figure 5

Figure 6

Table 1

Figure 2

Full text

HOUSING CYCLES IN THE MAJOR
EURO AREA COUNTRIES
Documentos Ocasionales
N.º 1001
L. J. Álvarez, G. Bulligan, A. Cabrero,
L. Ferrara and H. Stahl
2010

HOUSING CYCLES IN THE MAJOR EURO AREA COUNTRIES

HOUSING CYCLES IN THE MAJOR EURO AREA COUNTRIES
(*) (**)
L. J. Álvarez
BANCO DE ESPAÑA
G. Bulligan
BANCA D’ITALIA
A
. Cabrero
BANCO DE ESPAÑA
L. Ferrara
BANQUE DE FRANCE
H. Stahl
DEUTSCHE BUNDESBANK
(*) The authors would like to thank Víctor Gómez for providing his Butterworth filters software (TRACE), seminar
participants at Banca d’Italia and participants of the conference on Macroeconomics of Housing Markets organized at
the Banque de France, November 2009, as well as C. André, O. de Bandt, J. González Mínguez, T. Knetsch, L. Á. Maza
and G. Pérez-Quirós for useful comments on earlier versions of this draft. The views expressed herein are those of the
authors and do not necessarily reflect those of their institutions.
(**)
Banco de España, DG Economics, Statistics and Research; Banca d’Italia, Department of Economic outloo
k
and Monetary Policy Studies; Banque de France, Economics and International Relations Department; Deutsche
Bundesbank, Economics Department.
Documentos Ocasionales. N.º 1001
2010

The Occasional Paper Series seeks to disseminate work conducted at the Banco de España, in the
performance of its functions, that may be of general interest.
The opinions and analyses in the Occasional Paper Series are the responsibility of the authors and,
therefore, do not necessarily coincide with those of the Banco de España or the Eurosystem.
The Banco de España disseminates its main reports and most of its publications via the INTERNET at the
following website: http://www.bde.es.
Reproduction for educational and non-commercial purposes is permitted provided that the source is
acknowledged.
© BANCO DE ESPAÑA, Madrid, 2010
ISSN: 1696-2222 (print)
ISSN: 1696-2230 (on line)
Depósito legal: M. 15552-2010
Unidad de Publicaciones, Banco de España

A
bstract
The recent burst of the house price bubble in the United States and its spillover effects on
real economies worldwide has rekindled the interest in the role of housing in the business
cycle. In this paper, we investigate the relationships between housing cycles among
the four major euro area countries (Germany, France, Italy and Spain) over the sample
1980Q1-2008Q4. Our main findings are that GDP cycles show a high degree of
comovement across these four countries, reflecting trade linkages, but much weaker ones
for housing market cycles, where idiosyncratic factors play a major role. House prices are
even less related than quantities across countries. We also find much stronger relationships
in the common monetary policy period.
Keywords: Housing cycles, synchronisation measures, euro area countries.
JEL codes: E32, R21, R32.

Frequently asked questions

Q1. How is the synchronization among value added in construction variables?

Synchronization among value added in construction variables is low on average (0.19), ranging from negative between Italy and Germany (-0.25) to mildly positive between Italy and France and France and Spain (around 0.4 in both cases). 

Q2. What is the main reason why the euro area has a high degree of comovement?

The authors find that, in the four major euro area countries, GDP cycles show a high degreeof comovement, most likely due to trade linkages, although idiosyncratic factors play a larger role, particularly in Germany. 

Q3. What is the first step in the concordance analysis?

The first step in the concordance analysis is to identify peaks and troughsin the Butterworth band-pass filter cycles estimated above. 

Q4. What is the role of housing in the business cycle?

The recent burst of the house price bubble in the United States and its spill-over effects on real economies worldwide have reinforced the interest in the role of housing in the business cycle. 

Q5. What is the simplest way to determine if a series is in the same phase?

The simplest one is to compute a concordance index, which measures the fraction of time that the reference cycles of different series are in the same phase [Harding and Pagan (2002)BANCO DE ESPAÑA 17 DOCUMENTO OCASIONAL N.º 1001or Artis et al. (2004)] and which is bounded between 0 and 1. 

Q6. What is the reason for the comovement of non-residential investment?

The significant comovement of non-residential investment may be explained by the stronger interconnection of business activities as opposed to the more local nature of housing markets. 

Q7. What is the linear interpretation of the band-pass filter?

they can be given a model-based interpretation: the band-pass BFT can be obtained as the best linear estimator, in the mean squared sense, of the signal in a signal-plus-noise model, where the signal follows a particular ARIMA model. 

Q8. Why are stronger relationships seen for residential investment variables in the period with a common monetary policy?

Stronger relationships are also seen for residential investment variables in the period with a common monetary policy, probably due to convergence in mortgage interest rates. 

Q9. What is the main reason why Spain has stronger comovements across countries?

The analysis of the European Monetary Union period clearly shows stronger GDPlinkages across countries than in the whole sample, probably reflecting the increasing importance of trade flows. 

Q10. What are the main factors that influence housing demand and supply decisions?

Other factors, such as the degree of home ownership, the availability of land or the regulatory framework, can also have an important impact on housing demand and supply decisions [(Muellbauer and Murphy (2008)].