Variation across price segments and locations: A comprehensive quantile regression analysis of the Sydney housing market

Standard house price indexes measure average movements of average houses in average locations belonging to an average price segment. Such procedures obscure a huge variety of price development patterns in housing markets across price segments and geographical areas. Unfavourable price developments may be offset by opposing movements in other sub-markets creating a false sense of security. This paper uses quantile regression techniques to reveal this kind of variation. Two novel hedonic approaches based on the time-dummy and imputation method respectively are developed to compute quality-adjusted price segment- and location-specific house price indexes. The proposed methods are applied to house sales in Sydney, Australia, between 2001 and 2014. The analysis finds a rich set of variation across sub-markets over time. Whereas the price peak in 2004 was driven by sharply increasing prices of suburban, low-priced houses, the peak in 2010 can be mainly attributed to rising prices in the inner city. From 2012 onwards, the entire market experiences large increases which are strongest in the lowest price segment. The findings clearly suggest that standard house price indexes are not enough to assess the state of a housing market and actually obscure a lot of variation within a market. The joint analysis of movements in price segments and geographical areas allows deep insights which are likewise of interest for policy makers, home owners, urban planners and investors.