Three essays on the performance of housing markets in Australia
thesisposted on 28.03.2022, 19:32 authored by Haresh G. Pardasani
This dissertation presents three empirical studies on the performance of housing markets in Australia. The evidence presented in this dissertation addresses a number of outstanding issues in the extant literature. Specifically, this dissertation: examines the returns derived by home owners undertaking government-sanctioned home improvements; assesses persistence in the components of housing returns, both rental and capital; and evaluates the impact of regulation introduced to reduce information asymmetry by penalizing real estate agents who underquote list prices during marketing campaigns. The first study examines the realized returns of homeowners who undertake significant home improvements compared with those who do not. Using a proprietary data set of repeat home sales, this study finds that, overall, the cost-adjusted return to households who improve their homes is 2.4% lower vis-a-vis owners who make no alterations following a property purchase. The results are robust across multiple model specifications and additional tests, including sample selection bias.The findings are consistent with the consumption view that households undertake home improvements for hedonistic consumption purposes and consequently overcapitalize. Further analysis, which identifies speculators or flippers conditioned on the holding term, demonstrates that homeowners who buy-improve-sell within two years achieve higher returns (around 5.4%), particularly for works including extensions and alterations. This finding is consistent with speculative investment behaviour. Non-speculators, on the contrary, are consumption-motivated and are either unaware of, or undeterred by, the possibility that they may be overcapitalizing on home improvements for their distinct utility. The second study examines the persistence in inflation-adjusted rental and capital index housing returns across multiple geographic demarcations including national, capital city and regional areas for both houses and units across Australia. Applying univariate variance ratio tests and non-parametric independent runs tests on both monthly and quarterly index returns between 2005 to 2017, this study provides empirical evidence of persistence in the rental and capital components of housing market appreciation in Australia. Overall, this study finds higher persistence in the capital component compared to the rental component of housing market returns. The study also finds higher persistence in greater capital cities (for example, Greater Sydney) relative to the regional areas. Analysis of smaller regions (i.e. SA3) through to the national level of aggregation reports that the proportion of regions with persistent returns systematically increases. This result suggest that, on average, the smaller regions with persistence must have at least a strong enough degree of persistence to be able to influence it in index returns at a higher level of aggregation. By property type, this study finds higher persistence in returns for houses than units with respect to the capital component of housing market returns, whereas the converse is true for the rental component of housing market returns. Additionally, the study also finds that, overall, there is more persistence at a monthly frequency than a quarterly frequency. The final study investigates the impact of the introduction of regulation to stamp out underquoting practices by estate agents. Underquoting occurs when a property is listed for sale at a price that is lower than the likely selling price for that property. Despite opposition from the real estate industry, which believed that the regulation was unnecessary, two of eight state governments passed reforms on October 22, 2015, and November 02, 2016, which subsequently came into effect on January 01, 2016, and May 01, 2017, in New South Wales and Victoria, respectively, providing a natural experiment. Utilizing the population of listings and home sales data from 2014 to 2018, classified into pre-enactment, enactment, and post-commencement periods, this study provides evidence of the widespread prevalence of the underquoting practice in the states, and a reduction in underquoting practices just after the law was enacted. The empirical findings for New South Wales show that, for 'Auction' sales, there is a 4.1% and 5.7% reduction in underquoting during the enactment period and after the commencement of the law, respectively. Meanwhile for 'Private Treaty' sales, the study finds a 1.2% and 3.8% reduction in underquoting, respectively. Comparison of the change in underquoting ratio of listings in Victoria relative to other states shows that, in the case of 'Auction' sales, there is a 2.3% and 7.3% decrease in underquoting during the enactment period and post-commencement period, respectively. However, in 'Private Treaty' sales, the study finds the decrease in the underquoting practice is evident only after the law was effective, with a 2.7% reduction in underquoting, while there is a marginal increase in underquoting of 0.5% during the enactment period. These findings are robust to location fixed effects at the local geographic area level and controlled for market index estimates, various property and listing characteristics, and other factors - spatial or temporal -- summary.