by Eric Girardin, Aix-Marseille University
eric.girardin@univ-amu.fr
Christelle Lecourt, Aix-Marseille University
Christelle.lecourt@univ-amu.fr
Roselyne Joyeux, Macquarie University
roselyne.joyeux@mq.edu.au
Executive summary
House prices in Europe have been very volatile over the last 30 years. After a collapse between 2007 and 2013, these prices have experienced very different movements from one country to another. One of the most striking features is the divergence between the decline in real terms of the German housing market prices and the rise of other European market prices from the mid-1990s to the global financial crisis (Ott, 2014). It is important to determine whether this growth in housing prices has not been accompanied by a similar growth in economic fundamentals. Indeed the same way as it is customary to work on the price-earnings ratio in the case of the stock market, one should similarly control for fundamentals, such as rents, through the use of the price-to-rent-ratio (PRR). An analysis based only on housing prices across countries may wrongly conclude that such prices diverge, when such divergence may vanish when controlling for the corresponding countries’ fundamentals.
In this paper we analyse whether the PRRs across the main European countries, and particularly members of the Eurozone, converge over long periods. There is an extensive literature on the convergence across the European nations in terms of common macroeconomic and financial factors, but to date, except a few recent papers (Tsai, 2018; Maynou et al. 2021), studies dedicated to the convergence of European housing markets are scarce. To this end, we investigate whether European residential markets share a single common component or whether there are convergence clubs in which members converge towards common steady states. We also analyse the dynamics of PRR fluctuations. In particular, we see to what extent Germany and the UK act as leaders in possible euro and non-euro area clubs respectively. We us of tests which eable us to detect whether there is one common factor that drives the PRR dynamics across the real estate markets or whether there are several groups with separate common factors. Moreover, the tests allow for the series to temporarily diverge from the general convergence path. Our PRR index is available for 13 European countries over a sample running from 1988Q1 to 2020Q4. This sample includes 9 euro-area countries (Belgium, Finland, France, Germany, Ireland, Italy, the Netherlands, Portugal and Spain) as well as 4 non-euro ones (Denmark, Norway, Sweden and the United Kingdom) in order to determine whether there is a Eurozone effect.
The general message stemming from our work is the presence of two major convergence clubs among European countries. European housing markets seem to have been consistently split between two poles: a group of Northwestern members (Belgium, Ireland and the Netherlands) with higher rises in the PRR than the European average and another geographically diverse group (Finland, Germany and Portugal) with lower PRR rises than average. Such patterns have not changed substantially either after the formation of the euro area or after the global financial crisis. These events only led two categories, composed respectively of Southern-European, and non-euro European countries, to alternatively belong to either of the above two groups.
More specifically we obtain two series of results. First, the convergence/divergence patterns show a redistribution of club membership from the pre- to the post-euro formation era. During the pre-euro period a first club involved Belgium, Ireland, the Netherlands and Italy (plus Denmark from outside the area), while the second club included France, Germany, Portugal, Finland, and Spain (plus Norway, Sweden and the UK from outside the area). In the euro period one club includes Belgium, France and Germany (plus Norway, and the UK from outside the area), while another club includes all the other 6 euro-area countries (plus Denmark).
Second, the dynamics of the deviation of PRRs from the European average distinguishes between three groups of countries inside the euro area. The first and second groups experienced consistently either a below-EU-average movement of their PRRs (Finland, Germany and Portugal) or an above-EU-average movement (Belgium, Ireland and the Netherlands). In contrast, the third group (France, Italy and Spain) alternated between a positive and a negative deviation from the European area average. Overall the non-euro countries’ patterns represent a flopping around of the pattern characterizing group-one countries. Instead of starting from the European average in the 1990s and deviating negatively from that average subsequently, noneuro countries deviated from the average for the first 15 years and then went back to the European average.