Uniswap's 2018 launch replaced the order book with a formula

Inherited archive entry · retained from the earlier edition. Its factual, legal and policy claims have not all been reverified for 19 September 2026. Research scope.

A launch with $30,000 and no order book

On 2 November 2018, the final day of that year's Devcon conference, Hayden Adams deployed the Uniswap contracts to the Ethereum mainnet. Uniswap Labs' own account of that launch puts the opening liquidity at roughly $30,000 across three token pairs, deposited by a single provider, enough to support trades of about $100. There was no book of buy and sell orders to match. The idea traces further back than the implementation: the post credits Alan Lu of Gnosis with first conceiving an "x*y=k" market maker on Ethereum in 2016, and Vitalik Buterin with popularising it, before Adams built what the post calls "the specific implementation of an x*y=k market maker that exists on Ethereum today."

What the formula actually guarantees

The whitepaper draft that circulated alongside the launch sets out the mechanism: a pool of two tokens is priced by the requirement that the product of their reserves stays constant. A trade that adds one token to the pool must remove the other in the amount that keeps that product fixed, and the ratio of that exchange is the quoted price. No external feed sets the number; the pool's own reserves do. The original repository shows this logic implemented in Vyper, a restricted, Python-like language chosen for the kind of auditability the era's smart-contract failures had made a selling point.

What a liquidity provider actually holds

Supplying a pool is not the same as holding two static balances. A provider's claim moves with every trade that passes through the pool, and the ratio of assets they can withdraw at any moment reflects the current price, not the price at which they deposited. The formula guarantees the invariant between reserves; it makes no promise about the value of what a provider can redeem. That distinction, and the loss it can produce relative to simply holding the two assets, is a mechanism a separate analysis in this archive treats in full.

Questions to carry forward

• Does a quoted "price" from a pool reflect where the broader market is trading, or only where this pool's own reserves happen to sit?
• What is the smallest pool size at which a large trade can move the price further than a comparable trade on an order book would?
• What does a liquidity provider need to know before assuming that trading fees will outpace the pool's rebalancing against them?

Uniswap v1 replaced negotiation with arithmetic: anyone could supply a pool, and anyone could trade against it, at a price the pool itself computed. That was a genuine mechanical innovation. It was not, on its own, a claim that the resulting price was always the right one, or that supplying liquidity was a low-risk way to earn a fee.