This paper explores the real estate price determinants at four geographical levels: in the European Union as a whole, in the 28 European Union countries, in one European Union country (Portugal) and in 25 Portuguese metropolitan statistical areas (MSAs).
The authors run two time series regression models and two panel data regression models with observations of potential real estate price determinants and House Price Indices collected from Eurostat.
The results show that price determinants, such as gross domestic product (GDP), interest rates, housing starts and tourism, are statistically significant, but not in all the four geographical levels of analysis. The results also confirm the autoregressive characteristic of real estate prices, with the last period price change being the most important determinant of current period real estate price change.
Forecasting real estate prices can be made more effective by knowing that each geographical level of analysis implies different price determinants and that momentum is an important determinant in real estate returns.
To the best of the authors knowledge, this is the first study to develop and test a real estate price equilibrium model at several different geographical levels of the same political space.
