Decomposing the Term Structure of Housing Risk : Implications of Market Segmentation and Liquidity

This paper studies the empirical properties of the term structure of total housing risk in Sweden using a particularly rich dataset of housing transactions. We contrast recent studies by showing that, even after ten years, 75% of the total variation in housing returns will be attributable to the idiosyncratic shock. This insight first becomes visible when incorporating the autocorrelation present in the systematic real estate returns. In addition, exploiting unique variables at hand, we show that the effect of liquidity only explains a small portion of the idiosyncratic risk. Furthermore, with a simple but powerful simulation, we illustrate that just relying on index movements as the best signal for total housing risk severely underestimates the riskiness of real estate equity, especially for shorter holding periods. Our conclusions are robust both across housing categories and geographical areas, where we also note cross-sectional variation in the levels of both idiosyncratic and systematic risk