Home, sweet home or is it - always?

The question of whether the housing market is efficient or not is posed by an increasing number of economists, policymakers, current homeowners and prospective homebuyers. This article tests the efficiency hypothesis on data from the Norwegian housing market in its capital, Oslo. We employ the Case-Shiller time-persistence-test on a repeated-sales model of a house price index and returns to housing. Our data cover the period 1991-2002 and comprise 20 752 transactions of same-object-repeated-sales. We explain why certain features, sometimes suppressed in earlier tests, of the data set are of importance in efficiency tests, and argue that ours is particularly well-suited for the purpose. We demonstrate that the repeated-sales house price index contains inertia and time-persistence. In addition, we investigate how the price history of returns; which consist of capital gains, dividends, and interest payments; can be exploited to predict future returns. Both the house price index and housing returns contain forecastable elements, so we reject the null hypothesis of martingale processes, a finding that is indicative of Case-Shiller inefficiency. This discovery is supplemented with an exploration of trading and timing rules by examinations of intra-market and inter-market returns. We show that the housing market consistently yield higher return at lower risk than does the stock market over the period, which is inconsistent with inter-market efficiency.