STATISTICAL AND ECONOMIC TESTS OF EFFICIENCY IN THE ENGLISH PREMIER LEAGUE SOCCER BETTING MARKET

Abstract

This thesis investigates the weak and semi-strong form efficiency of the fixed odds English Premier League soccer betting market between 2002-03 and 2007-08. Recent structural changes – including a reduction in taxes, and the rapid growth of online bookmakers – renders this market ideal for empirical efficiency analysis. Weak form evidence indicates that favourite-longshot and home ground advantage biases exist in the quoting of bookmaker odds. In order to conduct semi-strong form analysis, a number of ordered probit models are specified, incorporating fundamental variables which are widely perceived to contain predictive power with regard to the outcome of a soccer match. The Kelly betting strategy isutilised to analyse the economic significance of their predictions, for matches played in the three most recently completed seasons, 2005-06 to 2007-08. It is found that the implementation of two methodological adjustments – the avoidance of bets on away longshots, and a staggered start and finish to betting in each season – results in the generation of significantly positive returns, providing strong evidence against semi-strong form economic efficiency. Evidence presented in this thesis indicates a strong preference for a fractional Kelly strategy and supports the technique of combining forecasts, findings consistent with previous literature. Further, it is shown that a distinct improvement to the returns from any strategy can be obtained by shopping around for the best available odds.