ABSTRACT
The usefulness of nonlinear models in forecasting
financial variables and challenging market efficiency
has been extensively investigated over the past
20 years (see Mills and Markellos, 1997, inter alia).
Despite the fact that the efficiency of wagering
markets has been also widely studied, most of this
research has employed linear parametric models
(see Sauer, 1998; Vaughan Williams, 2005; Vlastakis
et al., 2007). As was the case with financial
markets, one could reasonably argue that the incon-
clusiveness of results with respect to betting market
efficiency may be due to a misspecification of the
models used. The present article extends the literature on market
efficiency by evaluating the statistical and economic