What Is an Automated Market Maker in a Prediction Market?
An automated market maker (AMM) in a prediction market is a smart contract that continuously quotes prices for outcome shares and executes trades against its own liquidity pool, allowing anyone to buy or sell without needing a counterparty. Unlike a traditional order book where buyers and sellers wait for matches, an AMM sets prices algorithmically based on the ratio of assets in its pool, enabling instant trades at any time.
How AMM pricing works in prediction markets
A prediction market AMM holds reserves of two or more outcome shares - for example, "Yes" and "No" shares for a binary market. The AMM uses a mathematical formula to determine the price of each share based on the relative sizes of these reserves. The most common formula is the logarithmic market scoring rule (LMSR), which is a variant of the constant product formula adapted for prediction markets.
The key property of an LMSR-based AMM is that it can never be drained of both outcomes. The price of a "Yes" share will approach zero as the "Yes" reserve shrinks, but never reach it, and the same applies to "No." This ensures the market always has liquidity for both sides.
How trades affect prices
When a trader buys "Yes" shares, they pay a certain amount of collateral into the AMM, and the AMM issues new "Yes" shares from its reserve. This increases the "Yes" reserve and decreases the "No" reserve, moving the price of "Yes" upward. Conversely, selling "Yes" shares returns them to the AMM, which burns them and releases collateral back to the seller, shifting the price downward.
The relationship is not linear. A small trade in a deep pool moves the price only slightly, while a large trade in a shallow pool can cause a significant price swing. This is why prediction markets with higher liquidity tend to have more stable odds.
Settlement and the AMM's Role
When a market resolves, the AMM no longer serves a pricing function. Instead, the conditional token framework allows the AMM's reserves to be redeemed for collateral at the settlement ratio. For example, if the outcome is "Yes," each "Yes" share is worth 1 USDC and each "No" share is worth 0 USDC. The AMM's "Yes" reserve becomes fully redeemable, and its "No" reserve becomes worthless. The AMM contract typically allows anyone to withdraw the remaining collateral after settlement, returning it to the liquidity pool for redistribution to liquidity providers.
Liquidity provision and impermanent loss
Liquidity providers (LPs) deposit equal value of "Yes" and "No" shares into the AMM pool. They earn fees from every trade executed through the pool. However, they face a risk known as impermanent loss, which is more accurately called divergence loss in this context.
If a market moves strongly toward one outcome, the AMM's reserves shift heavily toward the losing side. An LP who deposited when the market was 50/50 may find that their pool share is now mostly "No" shares worth very little, while the "Yes" shares they could have held directly have appreciated. The fees earned may or may not compensate for this divergence.
In practice, prediction market AMMs often attract LPs who believe the market is mispriced and want to profit from both fees and the eventual resolution. This is a different risk profile from a standard DEX AMM.
AMM vs order book: key differences
Polymarket uses a hybrid model: an off-chain order book for limit orders and an on-chain AMM for market orders and instant liquidity. Other platforms like Azuro rely entirely on AMMs.
| Feature | AMM | Order Book |
|---|---|---|
| Trade execution | Instant, always available | Requires matching counterparty |
| Price discovery | Algorithmic, based on pool ratios | From buyer and seller orders |
| Liquidity source | LP deposits | Trader orders |
| Slippage | Increases with trade size relative to pool | Depends on order book depth |
| Gas cost per trade | Typically higher due to on-chain swaps | Can be lower if using off-chain matching |
Risks specific to amm-based prediction markets
- Pool manipulation: A large trader can temporarily distort prices by making a very large trade, potentially triggering liquidations or misleading other traders. The AMM's formula makes this expensive, but not impossible.
- Front-running: Because AMM trades are public before confirmation, a validator or bot can see a pending trade and execute their own trade first to profit from the expected price movement.
- Liquidity withdrawal: If LPs withdraw their funds rapidly during a volatile period, the AMM's depth can shrink, increasing slippage for remaining traders.
- Smart contract bugs: The AMM's pricing logic and the conditional token integration are complex. A bug could allow draining the pool or mispricing shares.
How to Evaluate an AMM-Based Prediction Market
Check the pool depth for the market you want to trade. A small pool means high slippage on anything beyond a minimal trade. Look at the fee structure - some AMMs charge a flat fee per trade, others use a dynamic fee based on volatility. Also verify that the AMM contract has been audited and that the LP tokens are not subject to a rug pull by the market creator.
The AMM is not a black box. You can read the pool's reserves on-chain at any time, calculate the current price yourself using the formula, and estimate the slippage for a given trade size before you submit the transaction.
Not financial advice. wswap.site publishes market data and general information about digital assets. Crypto assets are volatile and you can lose everything you put in. Nothing here is a recommendation to buy, sell or hold, and we make no price predictions.
Prices are sourced from third parties and may be delayed or wrong. Verify anything you intend to act on against a primary source.