ABSTRACT
The further condition for semi-strong efficiency is that there should be no arbitrage opportunities between times M and R. Assume that F represents fair odds, and that the public believe that the tipping column has proven profitability. Path ACD has an initial segment AC during which time there is sustained betting support, net of hedging (laying-off liabilities) on the betting exchanges, to the approximate midpoint between M and R. Odds decrease at a constant rate over the segment AC. Beyond this point in time, segment CD indicates a reversal of odds movement, with hedging on the exchanges outweighing further betting-to-win. If the situation in ACD is representative, semi-strong inefficiency exists in the sense that there are transparent arbitrage opportunities associated with the segment AC. In contrast, the initial adjustment downwards of odds following publication in ABCD, represented by the segment AB, is much more rapid after publication, followed by segment BC (further bets-to-win outweigh hedging by early bettors), and CD (hedging outweighs bets-to-win). ABCD still constitutes semi-strong inefficiency in relation to the dynamics of the market, but not to the same degree as ACD. The limiting case is where adjustment to fair odds following publication is instant, after which odds remain constant until R, represented by AF in Figure 17.1. This limiting case offers no opportunity for profitable arbitrage and fulfils both conditions for semi-strong efficiency in the scenario depicted above. The precise nature of odds adjustment to tipped horses in reality is in question, but anecdotal evidence suggests that in the case of the most popular newspaper tipping features (typically those published in the Racing Post, the leading UK horse race trade newspaper) there is rapid assimilation of this information into odds, which frequently contract substantially following publication, broadly consistent with semi-strong efficiency, but that there is often also a period later in the market when odds extend due to bettors locking in value by laying selections on the exchanges. The current author is unaware of any empirical studies that have fully addressed the dynamics of markets in relation to intermediate stages in bookmaker or betting exchange markets in relation to horse race betting markets; rather, such studies have been confined to analysis of starting price (SP)3 market data alone; or of comparative static analysis of odds at two or three points in time. For example, morning odds, opening show,4 and SP (Crafts, 1985). Smith (2003) studied the relationship between the tips of UK racing journalists and the price movements and profitability of racehorses that they select. Greater attention was paid to one tipster than the others, Pricewise of the Racing Post. The idea behind the column is to identify ‘over-lays’; that is, horses whose true chance is understated by the bookmakers’ prices – the odds are too big. The column recommends an average of between two and three bets on days featuring high-profile races, sometimes more than one in a race. Pricewise has long been portrayed in the media as a feature that leads to significant odds movements, and claims to be a notably profitable newspaper column. In relation to 4,434 horses, Smith measured odds movements from early-morning bookmaker odds to SP; the measure used was that employed by Crafts (1985).5
