ABSTRACT
A prediction market offers contingent contracts whose future payoff is tied with the outcome of some uncertain event of interest and attracts traders to wager on the outcome. For example, to predict how likely it is that there will be a bird flu outbreak in the United States by 2012, a prediction market can offer a contract that pays $1 if a bird flu case is confirmed in United States by the end of 2012, and $0 otherwise. A risk-neutral agent who believes that a bird flu outbreak will happen with probability α has the incentive to trade the contract and drive the price of the contract to $α. If every participant trades the contract based on his or her private information, at an equilibrium the market price of the contract can represent the consensus belief on the likelihood of a bird flu outbreak based on the pooled information. In theory, the event of interest can be thought of as a random variable X, and the payoff of the contract depends on the realized value of X. X is often a discrete or discretized random variable that has n mutually exclusive and exhaustive outcomes. Different market mechanisms can be used to facilitate trading in prediction markets. As the primary function of prediction markets is information aggregation, the following properties, among others, are desirable for market mechanisms to better achieve the information aggregation goal:
• Liquidity. Liquidity requires that market participants can find their counterparties to trade whenever they want. Participants reveal their information by trading contracts. If they cannot trade, even if they have relevant information, they cannot reveal it in the market.
