Ebook House Prices, Interest Rates and Macroeconomic Fluctuations: International Evidence
There is a large and growing literature documenting the nature and degree of cross-country comovement of macroeconomic variables as well as a separate and equally large literature studying comovement in stock returns and interest rates. In this paper we contribute to the effort of bridging these two strands of literature and study the inter-relationships in the degree and nature of comovement across both macroeconomic aggregates and asset returns. Included in our list of asset returns is the growth rate of real residential house prices.
Our interest in including house prices in our study is motivated by the large role that housing plays in an individual’s wealth portfolio and the role they may play in explaining macroeconomic fluctuations. Housing activities account for a large fraction of GDP and households’ expenditures in industrial countries. Moreover, housing is the main asset and mortgage debt the main liability held by households in these countries, and therefore large house price movements, by affecting households’ net wealth and their capacity to borrow and spend may have important macroeconomic implications. From a global perspective, housing is the quintessential nontraded asset yet, as we document in this paper, there is a surprising degree of synchronization in the changes in the price of this asset across industrial countries. In fact, the degree of comovement is on par with the magnitude of comovement in both financial asset returns (essentially frictionless) and macroeconomic aggregates (slowed only by trade frictions).
In this paper we provide empirical evidence on the nature and source of comovement in real house prices, real stock prices, real GDP, consumption, investment, and interest rates across 13 industrialized countries (see Appendix A). The novelty of this paper is that our econometric model explicitly allows us to study the system of variables simultaneously, identifying the international linkages between real variables and financial variables and providing evidence with implications for open economy real business cycle models, portfolio choice problems and the effects and transmission of monetary policy shocks.
To document the comovement properties of asset prices, interest rates, and macroeconomic variables we first estimate a dynamic factor model that contains a global factor capturing comovement in all variables, country specific factors that capture dynamics particular to a country, and aggregate specific factors that capture global movements specific to a variable type (e.g. global comovement in house prices not related to comovement across all variables in the system.). As already noted, an important feature of the econometric model is that all factors are estimated simultaneously, preventing confusion of the importance of each type of factor. That is, when estimating a country factor it is critical to control for comovement within the country that is caused by forces outside of the country to avoid overstating the importance of the country cycle. The second stage of our investigation uses the estimated factors in a VAR to first study lead-lag relationships in house prices across countries as well as to provide evidence on causal links between house prices, other financial variables and macroeconomic aggregates. Our results indicate a potential role for U.S. monetary policy shocks in explaining volatility in many of the variables, in particular house prices.
Related work by Chirinko et. al (2004) studies the interrelationship between stock prices, house prices, and real activity in a 13 country sample similar to ours. Their work estimates structural VARs country by country to develop systematic cross-country evidence on the importance of shocks to house prices, stock prices, real activity and monetary policy. Our work focuses on a different dimension of the data by using the dynamic factor structure to exploit commonalities across countries to better understand the sources and transmission of international shocks. In a similar spirit Dees, Di Mauro, Pesaran and Smith (2005) study the role of monetary, oil, real and equity shocks across countries in a global-VAR (GVAR) model that includes a factor structure to model linkages across countries. Econometrically, our model differs in that we estimate the factors as latent variables while Dees et. al. estimate the factors as weighted averages of observable variables. While our approach is computationally more burdensome it allows for a richer factor structure in the model as well as allowing the econometric procedure to dictate the implied ‘weights’ for each variable in constructing the unobservable factors. We also adopt different approaches to identifying shocks. While the procedures we adopt require fewer and weaker identifying assumptions, the downside is that we are unable to identify the full range of shocks that Dees et al do. Finally, our primary interest is in studying the role of house prices in the context of other macroeconomic and financial variables and hence our work emphasizes a different set of variables than in the Dees et. al. study.
Download
PDF Ebook House Prices, Interest Rates and Macroeconomic Fluctuations: International Evidence
Posted in :
