Can forecasters forecast successfully? Evidence from UK betting markets

This paper investigates the performance of professional racetrack forecasting services in the UK. The racetrack market is chosen for investigation because it offers a clear and well-defined set of costs and returns. As such, it provides a useful perspective from which to consider the efficiency of professional forecasting performance. Five services are selected based on their reported success in a period prior to the sample selection. The results of this study indicate that all the services showed evidence of net pre-tax profits, notably for more strongly recommended tips, but this would have required quite high stakes to cover the costs of accessing the information. Statistical tests of significance were unable, however, to confirm the evidence of profitability at conventional levels of acceptance. Taking into account the transactions costs of using these services, it is even more difficult to confirm the existence of abnormal returns.