Prediction Market Liquidity: Order Book vs AMM vs Parimutuel

Prediction market liquidity runs on one of three market structures that determine your fee model, capital exposure, and affiliate revenue for the life of the platform. A central limit order book matches traders peer-to-peer and earns a pure trading fee, the model Kalshi and Polymarket run. An automated market maker using a logarithmic market scoring rule (LMSR) guarantees a counterparty in thin markets at the cost of a capital subsidy. A parimutuel pool takes no price risk at all by splitting all stakes among winners.

The verdict: order books win at scale, AMMs solve the cold start, and parimutuel pools eliminate operator risk but give up continuous pricing. This guide compares all three on price certainty, capital subsidy, slippage, and downstream liquidity economics, and shows how the choice flows through to implied probability accuracy and affiliate payouts under CFTC oversight.

The Three Models Compared

Each market structure determines how risk and capital are distributed across the trader, the operator, and the pool. The order book pushes risk onto traders and earns fees; the AMM puts operator capital at risk to manufacture liquidity; the parimutuel pool holds no risk and pays out only what the pool collected. The table below is the decision map an operator should start from.

Central Limit Order Book: Peer-to-Peer Matching

A central limit order book is an exchange mechanism that matches buy and sell orders by price-time priority, so traders set the prices and the operator never takes a position. The price where a contract trades is its implied probability: a share changing hands at $0.62 means the market prices the outcome at a 62 percent chance. This is the model both Kalshi and Polymarket run at scale.

The matching engine pairs a buyer of an outcome share at $0.62 with a seller willing to trade at $0.62, and the operator collects a fee on the matched volume. The order books that Kalshi and Polymarket run offer tight spreads, transparent depth, and a fee model that grows with open interest rather than with operator risk.

The weakness is the cold start. A brand-new market has no resting orders, so the first trader sees an empty or very wide book and slippage on any meaningful size is severe. Order books reward concentration of volume, which means a long tail of niche markets will stay illiquid unless the operator seeds them. The practical answer is a market maker, in-house or contracted, who posts continuous two-sided quotes until natural order flow can sustain the book on its own.

Automated Market Maker and LMSR: Manufacturing Liquidity

An automated market maker is an algorithm that always quotes a two-sided price from a formula, removing the need to match two human counterparties. The logarithmic market scoring rule (LMSR), used historically by Augur and Gnosis, lets the operator set a liquidity parameter that bounds the maximum subsidy the operator can lose, in exchange for guaranteeing that every trader gets a fill. An AMM can seed a market with as little as a few hundred dollars of subsidy.

The AMM is the cleanest solution to the cold-start problem, because liquidity exists from the first second the market opens, and its quoted price still reads as an implied probability the same way an order book does.

The tradeoff is capital and adverse selection. The operator subsidizes the market: well-informed traders can extract value from the AMM, and that loss is the price of bootstrapping. LMSR makes the maximum loss explicit and bounded, which is its key operational virtue, but it still means committing capital that an order book never requires. For long-tail markets where natural two-sided flow will never materialize - obscure events, niche outcomes - an AMM or a hybrid that blends AMM quotes with an order book is often the only way to offer a market at all.

Parimutuel Pools: Zero Operator Risk

A parimutuel pool is a betting structure that takes every stake on every outcome, removes a rake, and splits the remaining pool among the winners, so the operator carries no price risk whatsoever. There is no order matching and no subsidy: payouts are simply whatever the pool collected, divided by winning stakes. This is the model of horse-racing tote boards, well suited to single events with a fixed betting window.

A parimutuel market guarantees a balanced book by construction, which is its central appeal to a risk-averse operator.

The cost is the trading experience. There is no continuous, live price a trader can lock in - final odds are only known when the pool closes, so a trader who stakes early does not know their effective price. That uncertainty makes parimutuel pools a poor fit for the always-on, exchange-style trading that order books deliver, but a strong fit for discrete events where simplicity and zero operator risk matter more than continuous pricing. Many operators use parimutuel for specific event types and an order book or AMM for the rest.

Bootstrapping Liquidity in Thin Markets

Operators must seed every new prediction market, because each structure starts thin and demands a different bootstrapping sequence. With an order book you contract or staff a market maker to post two-sided quotes on headline markets and accept that the long tail stays thin until volume concentrates. With an AMM you set the liquidity parameter high enough that spreads are tradeable but low enough that the subsidy is affordable. With parimutuel you do nothing extra, because the pool is balanced by design. Work the steps below in order.

1. Pick the structure first, because order book, AMM, and parimutuel each demand a different seeding approach.
2. For an order book, dedicate a market maker to marquee markets and let the long tail stay thin or seed it selectively.
3. For an AMM / LMSR, tune the liquidity parameter to balance tradeable spreads against the bounded subsidy cost.
4. For parimutuel, skip seeding but disclose that final odds and implied probability are unknown until the pool closes.
5. For a hybrid, open with AMM quotes for instant liquidity, then transition headline markets to a pure order book as depth grows.

Downstream: Fees, Slippage, and Affiliate Revenue

The liquidity model determines the revenue base that funds your affiliate program. Trading fees on an order book scale directly with matched volume, so a deep, low-slippage book generates more fee revenue per trader and supports richer RevShare or CPA payouts to affiliates. A thin, high-slippage market drives traders away before they generate fees, starving the program.

On an order book, trading fees flow straight through to partner economics, which is why operators model fee yield per active trader before fixing RevShare or CPA terms with affiliate and referral partners.

Whichever model an operator runs, the affiliate layer needs to attribute funded, active traders and reconcile commissions against real fee revenue, not raw signups. Track360 provides affiliate tracking, commission management, and reporting for prediction-market operators, so revenue-share payouts can be tied to the fee yield each model actually produces. See the prediction markets industry hub and our software build-vs-buy guide for how the stack fits together. The market structure you run also shapes your compliance posture: a CFTC-regulated designated contract market that lists an event contract faces different jurisdiction rules than an offshore venue, and the liquidity model also governs how fast each contract reaches resolution and settlement. Established markets such as CME Group plus oracle frameworks like UMA show how mature this market-structure thinking has become.