From boom to bust: Probabilities of turning points in house prices

This paper empirically tests whether deviations between actual house prices and fundamental prices — defined as the price that can be explained by disposable income, the housing stock, and the user cost of housing — affect the probability of turning points in house price cycles on a regional level. We apply Bry and Boschan (1971)’s algorithm to identify booms and busts in house prices for 14 municipalities in four different regions in Norway, using quarterly data over the period 2003–2021. A cointegrated vector autoregressive model is used to identify long-run drivers in local house prices. We find that while house price cycles, in many cases, are similar for municipalities within the same region, there is considerable heterogeneity in the timing, duration, and amplitude of booms and busts in real house prices across the regions. We also find substantial heterogeneity across regions in the response to long-run fundamentals, particularly in the response to a change in the user cost. The model is utilized to investigate if the real house price gap, the difference between actual and model-implied prices, affect the probability of turning points in house prices. We find that an increase in the price gap from 0%–15% increases the probability of a peak (and an ensuing downturn) in house prices by 7 percentage points.