Momentum and mean reversion in regional housing markets: evidence from variance ratio tests

Abstract

We study the persistence and reversion patterns of housing price growth by computing variance ratios applying Kim's (2006) Wild bootstrapping and using finnish data for the period 1987–2010. The momentum effect in housing price growth is found to be long-lasting and substantially greater in size than the eventual reversion. The results indicate that high-order autocorrelations are important concerning the long-horizon attractiveness of housing investments and that housing is a notably riskier asset in the long term than suggested by conventional portfolio analysis. The analysis further shows that the dynamics vary across regional housing markets and across dwelling types. The findings have important implications for investors, credit institutions, and policy makers.

Keywords

• Housing prices
• Momentum
• Mean reversion
• Variance ratio
• Portfolio allocation