An investigation into pricing anomalies and the influence of information processing constraints on forecasts derived from sports betting markets

This thesis is made up of three separate papers, all of which explore how decision makers discount information in simplified financial markets where participants' behaviour can be easily segmented. Overall, this thesis provides a valuable contribution to the existing knowledge of pricing anomalies in betting markets by identifying uncharted research niches and adopting novel applications. The first paper explicitly examines how legacy information affects the accuracy of probability forecasts, utilising data from a low liquidity sports betting market where the demographic of bettors is predominantly homogeneous. The findings are important as they shed new light on the ability of participants to discount older information, especially as increasing amounts of past performance information are processed. The second paper explores the influence of order effects in horse race betting markets which are generally ranked from highest to the lowest quality of contenders. Research suggests the way information is presented can impact how effectively information is processed by individuals. This paper finds evidence of order effects in relation to more favoured contenders which has serious implications for the most widely researched phenomena in betting markets, that is, the favourite-longshot bias. The third paper takes a novel twist on a commonly employed benchmark model to focus purely on market makers' ability to frame accurate opening price odds. Paper three directly tests whether public information plays a role in price movements over the duration of betting. In particular, the findings challenge the current consensus in the literature.