Home prices and fundamentals: solving the mystery for the G-7 by accounting for nonlinearities

Abstract

Purpose: The purpose of this study is to determine whether house prices and income share a stable, stationary relationship in the G-7 countries. This stable relationship has been clearly implied by theory but has been difficult to uncover empirically in previous studies. Design/methodology/approach: The analysis entails using nonlinear tests for a stationary relationship between home prices and per-capita income for the G-7 countries, whereas most previous papers on the topic have used linear methods. Findings: When the standard linear ADF test is used, no stationary relationship for home prices and income is found for any of the G-7 countries. When the more powerful (but still linear) Ng–Perron test is used, the USA, but no other G-7 country, exhibits a stable relationship between the two variables. When the nonlinear Enders–Granger test is used, stationarity between home prices and income is found for five of the remaining six G-7 states. Practical implications: Previous research has shown that as house prices have risen far above the income, especially over bubble periods, income has done a poor job in predicting home values. The findings show that income has a clear long-run stationary relationship with home values. This implies income could be helpful in providing home price forecasts. Originality/value: Where previous studies have failed to find a long-run relationship between home prices and income while using linear methods, results in this paper show this theoretical asset–pricing relationship holds once the adjustment process is allowed to exhibit nonlinearity.