Empirical essays in NSW auctions

Abstract

This thesis empirically examines why an agent, a seller or a buyer, will choose to hold or attend one particular sales mechanism over another. I examine this question using housing sales data and government bond auction data from the Australian State of NSW. In Chapter one I test whether auction or private treaty achieves higher revenue for a seller of a house in NSW, using an hedonic framework of house valuations. I hypothesise that auctions perform better when the value of the object is more dispersed. My results show that auctions do achieve higher revenue, and that more unique properties benefit even more from auctions over private treaty. In Chapter two I test whether a certain type of buyer or seller can achieve a better bargaining outcome through auction and private treaty sale. I use a regression model that explains the bargaining outcome with reference to a combination of property-related characteristics. I find that certain seller types do achieve better bargaining outcomes for private treaty sales, but there is no significant advantage for auction sales. Chapters three through five test whether discriminatory price government bond auctions achieve higher revenue than uniform-price auctions in NSW. I use a resampling method to estimate bidder valuations. I find that on average, discriminatory-price auctions achieve lower revenue than a counterfactual auction. But I then extend the analysis to include secondary market prices in the estimation of bidder valuations. When I account for the presence of secondary markets, I find that the original result is reversed and revenue raised using discriminatory-price auctions is higher than could be achieved in a counterfactual auction.