The project team compiled a new annual city-level dataset for 21 global cities in 11 OECD countries over the last 150 years. In many cases we hand-collected housing market data from diverse primary sources such as statistical agencies, tax records, newspapers, real estate agents or other types of archival data. We constructed price indices, rent indices and total return series for all the cities in our sample. In one specific case, Germany, we even digitized and harmonized the universe of real estate transaction data for 19 major German cities. Building on this data collection, we constructed localized house price indices for individual cities and for Germany as a whole. The data and indices have been made freely available on www.greix.de and have quickly become the standard reference for information on the German housing market.
Using the new long-run, cross-country housing data, we studied local dispersion in prices and returns on real estate. Our results indicate that in the long run total returns on real estate have been higher outside the large city centers. We explain this surprising result as a compensation for the risks that investors face in remote locations. We also identified the key sources of risk facing housing investors outside the large agglomerations. In smaller cities and rural areas, the correlation of asset returns with income risk is higher and lower liquidity means that price uncertainty is higher. Investors are compensated for taking on these risks by up to 1% per annum. Overall, our research established, for the first time, the presence of substantial spatial differences in returns on housing assets that are linked to economic fundamentals. The findings in this work package have advanced our understanding of the long-run financial implications of investments in housing assets across space.
In a second work package, we have studied the role of residential real estate for the evolution of wealth inequality. Housing assets are the most important asset for households outside the very top of the wealth distribution. The long-run returns on these assets therefore have important implications for the growth and distribution of household wealth. Our analysis showed that in particular for middle-class households in Europe, capital gains on real estate were the most important factor for wealth growth in the past half century. Savings flows, by contrast, have been a more important driver of wealth accumulation in the top-10. The bottom half of the wealth distribution has barely seen any wealth growth since the 1970s. An important insight is that trends in wealth inequality depend to an important degree on price developments in asset markets. In particular, wealth growth of middle-class households in particular has been closely linked to valuation gains in housing markets. By contrast, the bottom half of the population does not own financial assets and has fallen behind substantially with regard to wealth growth in the past three decades of valuation gains across markets. Measured by the gap between wealth levels at the 25th and 75th or 90th percentile, wealth inequality in Europe has grown substantially in almost all countries in Europe since the 1980s. Separate work on the U.S. has confirmed that owing to systematic differences in the portfolio structure between middle-class households (mostly invested in real estate) and households at the top of the distribution (with substantial business wealth), the relative evolution of property prices and stock prices is a first order determinant of changes in the wealth distribution over time. On a global level, the boom in real estate markets brought substantial gains in middle-class wealth that counterbalanced the growing concentration of savings accumulation at the top of the distribution.
The third work package looked at the cyclical swings in the housing market and the implications for financial stability, as well as the policy tools to deal with credit-fueled housing booms. The key insight of the first project in this work area was that historical attempts to use monetary policy, i.e. higher interest rates, to deflate credit-driven housing booms, have typically increased rather than decreased financial stability risks. Put differently, discretionary “leaning against the wind” policies are more likely to trigger crises than prevent them. These insights on the negative effects on interest rate policy to deal with housing booms, have major ramifications for the debate on the appropriate policy response to financial bubbles. Surprise interventions by central banks into booming asset markets can make the situation worse, which calls for policies to prevent the emergence of credit-fueled housing booms in the first place. A second project investigated the causal role of a loose monetary policy stance in triggering credit booms and housing bubbles. Combining long-run estimates of r-star (the equilibrium interest rate) with the actual policy stance of central banks revealed that “loose” monetary policy leads to the build-up of systemic financial risks. In a third project, we constructed a novel historical database of cross-country housing finance deregulation episodes. This allows us, for the first time, to show that deregulation policies have been relevant drivers of cyclical swings in credit and housing markets by expanding credit supply. The findings highlight the consequences of facilitating access to credit for home purchases both with respect to price trends in housing markets and the financial stability implications of large expansions of housing loans.
