Testing efficiency of football betting markets using adaptive drift models

Abstract

Over the years, the efficiency of financial markets has been studied extensively in literature, and the interest for investigating notions of efficiency also in the context of sports betting markets has been growing considerably. In its semi-strong form, the efficient market hypothesis holds the concept that a market is efficient if it fully reflects all publicly available information and has the ability to efficiently adapt to new information. Under efficiency, betting strategies based on this information should not be able to generate significant profits and consistent winning patterns. In this study, betting strategies informed by goal difference forecasts obtained from adaptive drift models are used to investigate the efficiency of Asian handicap betting markets for German football games. Findings based on predictions for the 2015/16 and 2016/17 Bundesliga seasons for FC Bayern Munich games suggest that efficiency and the efficient market hypothesis are supported. However, applying similar methodologies for 1. FSV Mainz 05 yields substantial winnings and indicates that inefficiency may be detected among various bookmakers.