Speculative bubble detection in Australian housing markets

Housing prices have risen significantly in all Australian capital cities over the past 30 years. Strong house price growth has led to the widespread belief that speculative bubbles exist in Australian housing markets. Whilst the issue of bubbles has been discussed in Australian housing market studies for some time, a robust investigation into the existence of speculative bubbles has not been given the same level of consideration. This paper investigates whether there is any evidence of speculative bubbles in the Australian national and the eight capital city housing markets during the period 1999 to 2017. We apply the bubble detection method proposed by Shi (2017), which controls for the impact of a large set of macroeconomic factors. The results are compared with those from the bubble detection method of Phillips et al. (2015a,b, PSY). The PSY real-time detection method controls for the impact of rent. Both approaches identify a significant speculative bubble during period 2003:Q2-2005:Q4 in most of the Australia's capital cities. While the method of PSY reveals significant evidences of speculation over the 2014-2017 period, the new method finds no evidence of speculation in all capital cities except Canberra since mid-2005. By taking into consideration the impact of macroeconomic factors (such as mortgage interest rate, disposable income, employment, population, and housing supply growth) on house prices, this study provides a better control for housing market fundamentals, which leads to a more precise detection of speculative bubbles. The outcomes of the study have important policy implications by aiding the control of speculative bubbles in house prices through smoother adjustments using housing market fundamentals.