This study adds to the literature on mean aversion and mean reversion in housing prices. In contrast with the previous related literature, the persistence and reversion characteristics are studied by computing variance ratios using Kim's (2006) Wild bootstrapping and by investigating horizons up to 10 years. The variance ratios clearly indicate that housing prices do not follow random walk in any of the 15 Finnish cities included in the analysis. Instead, momentum in housing price growth is longlasting and considerable in size. Since the eventual reversion is substantially weaker than the initial mean aversion, housing is notably riskier asset in the long term than suggested by variances computed from quarterly or annual price movements. The results also show that the momentum and reversion patterns may substantially vary between regional housing markets. These differences influence the optimal housing portfolio allocation and highlight one more reason why it is complicated to use country level housing price data when analyzing the optimal portfolio allocation or housing price dynamics.
housing prices momentum mean reversion variance ratio portfolio allocation