Housing price volatility and its determinants

Purpose – The purpose of this paper is to examine the housing price volatility for eight capital cities in Australia over 1987‐2007. Specifically, the volatility of Australian housing and its determinants were investigated. Design/methodology/approach – An exponential‐generalised autoregressive conditional heteoskedasticity (EGARCH) model was employed to analyse the volatility for eight capital cities in Australia. The Engle LM test was also utilised to examine the volatility clustering effects in these cities. Findings – The volatility clustering effects (ARCH effects) were found in many Australian capital cities. The importance of estimating each individual city's EGARCH model was also demonstrated in which the determinants of housing volatility vary from a city to another city. Asymmetric of the positive and negative shocks were also documented. Research limitations/implications – This study has implications for investors and policy makers in which housing investors should estimate the conditional variance (EGARCH process) of a housing market in respect to the volatility of housing series is not always constant over time. Furthermore, policy makers should also address the importance of considering the sub‐national factors in formulating the national housing policy. The analysis and results are limited by the quality of the data. Originality/value – This paper is one of the few studies in housing volatility. Additionally, it is probably the first attempt to assess the volatility spillover effects in the Australian housing market.