Automated Market Maker

An automated market maker, often shortened to AMM, is an algorithmic liquidity mechanism that always quotes a buy and sell price for outcome shares in a prediction market. Because the algorithm stands ready to trade at any moment, a participant can enter or exit without waiting for a matching order on the other side. This is what lets new or thinly traded markets function before they attract a crowd of traders.

Many AMMs use a scoring rule to set prices, the common example being the logarithmic market scoring rule, or LMSR, used historically by Augur and Gnosis. The rule moves the quoted price as shares are bought or sold, so heavy buying of a Yes share pushes its price up and the implied probability with it. A liquidity parameter controls how sharply the price reacts, trading off price stability against the operator subsidy required.

An AMM contrasts with a central limit order book, where prices come from trader orders rather than an algorithm, and with a parimutuel market, which only settles after pooling all stakes. A defining feature of an LMSR-style maker is bounded loss: the operator subsidizes the market but its worst-case cost is capped by the liquidity parameter, drawn from a liquidity pool set aside for the market. This bounded subsidy is the price of bootstrapping prediction-market liquidity in early or quiet markets.

For operators and partners, the AMM choice shapes fee economics and how commission is calculated. The platform may recover the subsidy through trading fees or a spread around the algorithmic quote, and affiliate revenue share then attaches to those fees on referred trading volume. Reporting must capture both the fee and the subsidized depth so partner attribution stays accurate as markets mature.