Testing Market Efficiency in a Fixed Odds Betting Market

This paper tests the hypothesis of market efficiency for the fixed odds betting market of Swedish trotting head-to-head matches. The hypothesis is carried out by a Wald test within a logistic regression model. Data support rejection of semi-strong efficiency at the 5 percent level of significance, while the weak form efficiency cannot be rejected. Moreover, evaluation of the simple strategy to bet on those horses where, conditional on the estimated model, the expected profit is positive results in a profit of 7.8 percent per bet.