On the Efficiency and Equity of Betting Markets

Confident in the belief that vital resource allocation decisions depend on well functioning capital markets, economists over the last two decades have spent a good deal of their time, and not a little of the computer's, studying the behaviour of stock market prices. Their main aim has been to discover the extent to which capital markets, and in particular markets in equity shares, are "efficient". Perhaps surprisingly to the layman, their answer has usually been "very efficient", and despite the amount of effort devoted to attempts at refutation, Fama was able to conclude that "the evidence in support of the efficient market's model is extensive and (somewhat uniquely in economics) contradictory evidence is sparse" (Fama, 1970, p. 416). The definition of "efficiency" being used is not, however, the usual one and reflects the fact that stock markets are being examined as information markets, not as service industries. A stock market is "efficient" if its prices always "fully reflect" "available information". Almost all the empirical work proceeds on the assumption that the conditions of market equilibrium can be stated in terms of expected returns, and it is becoming conventional to talk of three subsets of information in relation to the determination of equilibrium expected returns: historical prices (and returns), other publicly available information (e.g. announcements of earnings, issues, etc.) and "inside" information (i.e. information to which particular groups or individuals have monopolistic access at relevant points of time). Defining "available information" as each of these in turn pro-duces "weak", "semi-strong" and "strong" tests of efficiency. If prices always fully reflect historical prices in the sense that price changes approximate a random walk, the market is said to be "weakly efficient". The bulk of empirical work has been done in this area. While some drift is usually found and some filters (mechanical trading rules) have been discovered that produce profits in excess of a naive buy-and-hold policy, it is generally agreed that these have little practical significance given the transaction costs (brokerage) needed to implement them. If prices fully reflect public announcements as soon as they are made, the market is said to be "semi-strongly efficient". Only a small, though rapidly growing, volume of work has been done on this, particularly on dividend announcements and bonus issues.