Is there evidence of inefficiency in the Football Betting Market?

Viraj Thakor, V.T (2016) Is there evidence of inefficiency in the Football Betting Market? [Dissertation (University of Nottingham only)]

[img] PDF (Viraj Thakor Dissertation) - Registered users only - Requires a PDF viewer such as GSview, Xpdf or Adobe Acrobat Reader
Download (1MB)

Abstract

This dissertation investigates the existence of both weak form and semi-strong form efficiency in the football betting market over the last 8 seasons (2008/2009 to 2015/2016), for the English Premier League, the German Bundesliga and the Spanish La Liga.

Previous academic literature has highlighted the concept that the clear termination point of wagered bets and the transparency of breaking news make the football betting market very conducive to testing for efficiency, in comparison to the more complex nature of financial trading markets. This, along with the recent surge in online betting in the UK, with most of this activity coming in the form of gambling on football, perhaps renders the question of efficiency in the football betting market more important than ever before.

My study finds clear evidence of weak form market inefficiency, as profitable betting strategies are uncovered, whereby bettors can generate positive returns simply through exploiting arbitrage opportunities or adopting a favourite-longshot bias in particular seasons. The weak form of efficiency states that abnormal returns cannot be achieved through the solitary use of price information and therefore my findings prove that this condition of efficiency is broken in this particular market. Regression-based tests of weak form market efficiency using a linear probability model also show that the implied probabilities of home wins do not match the true probability of a home win occurring in each of our 3 European Leagues.

In testing for the existence of semi-strong form efficiency in this market, I adopt the use an ordered probit regression model with selected independent variables that are said to have predictive strength in forecasting the outcome of a football match. I find that a number of these variables contain considerable predictive power in relation to the match outcome, and therefore I deduce that the football betting market is not semi-strong form efficient as all available public information is not incorporated into the bookmaker’s odds.

Item Type: Dissertation (University of Nottingham only)
Depositing User: Thakor, Viraj
Date Deposited: 13 Mar 2017 15:13
Last Modified: 23 Mar 2017 08:06
URI: http://eprints.nottingham.ac.uk/id/eprint/36753

Actions (Archive Staff Only)

Edit View Edit View