| Authors | Adams, Hayden; Zinsmeister, Noah; Salem, Moody; Keefer, River; Robinson, Dan |
|---|---|
| Year | 2021 |
| Project | Uniswap |
| License | BSL 1.1 |
| Official Source | https://uniswap.org/whitepaper-v3.pdf |
This page is an educational summary and analysis of an official whitepaper or technical paper, written for reference purposes. It is not a verbatim reproduction. CryptoGloss does not claim authorship of the original work. All intellectual property rights remain with the original author(s). The official document is linked above.
“Uniswap v3 Core” was published in March 2021 by Hayden Adams, Noah Zinsmeister, Moody Salem, River Keefer, and Dan Robinson of Uniswap Labs. It describes a fundamental redesign of the Automated Market Maker (AMM) model pioneered by Uniswap v1 and refined in v2 — replacing uniform liquidity distribution across all prices with concentrated liquidity, where LPs specify price ranges for their capital. The result: dramatically higher capital efficiency, reduced slippage for traders, and a flexible fee structure, at the cost of increased LP management complexity.
Note: Three Uniswap papers are relevant to the history: the v1 post (2018), the v2 core paper (2020), and the v3 core paper (2021). This entry covers v3 — the dominant version.
> PDF hosting: The Uniswap v3 whitepaper is available at uniswap.org/whitepaper-v3.pdf. The protocol code is under BSL 1.1 (restricting commercial forks), but the paper itself is publicly available for reference and redistribution.
Background: Uniswap v1 and v2
Uniswap v1 (2018): Introduced the constant product formula x·y = k — any two tokens can be traded against ETH, where k is invariant. LPs deposit equal value of both tokens; trades move along the curve; LP fees (0.3%) are added to the pool.
Uniswap v2 (2020): Eliminated the ETH routing requirement — any ERC-20 can be paired directly. Added TWAP (time-weighted average price) oracles, flash swaps, and optimized the core contract. Still used x·y = k over the full price range (0, ∞).
The v2 problem: Liquidity is spread uniformly from price 0 to ∞. For a stablecoin pair (USDC/USDT trading at $0.998–$1.002), only a tiny fraction of the deposited capital is ever active. The rest sits idle. Capital efficiency was extremely low for pairs with predictable price ranges.
Concentrated Liquidity: The Core Innovation
In Uniswap v3, a liquidity provider specifies a price range [p_a, p_b] for their capital. Their liquidity only activates when the current price is within their range.
Mathematical model: Within any active range, the AMM behaves like a virtual constant product (x + L/√p_b) · (y + L·√p_a) = L², where L is the liquidity (a measure of depth) and the virtual reserves extend the standard curve within the tick range.
Key consequence: An LP concentrating all capital in the range $0.99–$1.01 on a stablecoin pair is effectively providing ~100× more liquidity depth within that range than a v2 LP at the same capital level. The whitepaper demonstrates LPs can achieve up to 4,000× capital efficiency over v2 for narrow range positions.
Ticks: Discretizing the Price Space
To make concentrated LP positions manageable and gas-efficient on-chain, Uniswap v3 discretizes the continuous price space into ticks — discrete price points spaced at 1.0001^i apart (0.01% intervals). Liquidity positions are defined by their lower and upper tick boundaries.
Active tick tracking: The contract maintains a sorted data structure of all initialized ticks with net liquidity deltas. When a swap crosses a tick boundary, the contract updates the active liquidity by reading the delta at that tick. This avoids scanning all LP positions — only the touched ticks matter.
Why 1.0001^i? This geometric spacing means each tick is 1 basis point (0.01%) apart in relative terms. Stablecoin pairs can use narrow ranges with many ticks; volatile pairs need wider ranges with fewer relevant ticks.
Fee Tiers
Uniswap v2 had a single fee tier (0.30%). v3 introduced three fee tiers, later expanded:
| Fee Tier | Intended Use |
|---|---|
| 0.01% (1 bp) | Stablecoin pairs with minimal price divergence |
| 0.05% (5 bp) | Highly correlated pairs (e.g., stablecoin/stablecoin, ETH/stETH) |
| 0.30% (30 bp) | Standard pairs (ETH/USDC, WBTC/ETH) |
| 1.00% (100 bp) | Exotic or high-risk pairs |
LPs choose which fee tier to provide liquidity in. Fee tiers are separate pools — a swap routes to the pool with the best price, which may not be the deepest. This introduced competition between pools of different fee tiers.
Sections of the Whitepaper
- Introduction — Recap of v1/v2; motivation for concentrated liquidity
- Concentrated Liquidity — Formal definition; virtual reserves; relationship to constant product curve
- Architectural Changes — How v3 differs from v2 at the contract level
- Ticks and Tick Bitmap — Discretization of prices; data structures for efficient swap routing
- Position Management — How LPs create, modify, and burn positions; fee accrual within a position
- Swapping — How a swap walks through ticks; computing output and fees step by step
- Oracle — Improved TWAP oracle; observation ring buffer; geometric mean price
- Fee Computation — How fees are accumulated per unit of liquidity; fee growth tracking inside/outside a tick range
- Governance — Protocol fee parameter controlled by UNI governance (0–1/4 of LP fees)
- Conclusion — Summary; acknowledgment of increased LP complexity
Key Innovations
NFT LP Positions: Because each LP position has unique tick boundaries and fee tier, positions are no longer fungible — they are represented as ERC-721 NFTs rather than ERC-20 pool tokens. This enabled secondary markets for LP positions.
Improved TWAP Oracle: The v2 oracle accumulated price×time; v3 uses a geometric mean (logarithm sum) approach that is more robust to price manipulation and cheaper to compute. The ring buffer size is governance-controlled.
Range Orders: By providing liquidity in a narrow range above (or below) the current price, LPs can mimic limit orders. If price crosses their range, all their liquidity converts from token A to token B — equivalent to selling at that price. This was not possible in v2.
Gas Optimizations: Despite higher complexity, v3 achieved lower gas costs per swap than v2 through bitwise tick bitmap lookups and optimized storage slot packing.
Trade-offs and LPing Complexity
Concentrated liquidity dramatically increases impermanent loss risk for active LPs:
- An LP whose range goes out of bounds holds 100% of one token and earns no fees until price returns
- The optimal range depends on price volatility — requiring active management or LPs accept suboptimal capital efficiency
- This created a professional LP market and spawned an ecosystem of liquidity management protocols (Arrakis Finance, Gamma Strategies, Bunni) that auto-manage v3 positions
Research (notably the “Concentrated Liquidity” study by Topaze Blue) found that many v3 LPs were net losers compared to simply holding — the benefits accrued primarily to sophisticated market-making operations with dynamic rebalancing.
Uniswap v4 (2023 Preview)
The whitepaper for Uniswap v4 (published 2023) extends v3 with hooks — plugins that execute custom logic at swap/liquidity events (e.g., dynamic fees, TWAMM integration, on-chain limit orders). V4 also introduces a singleton contract (one contract for all pools) for major gas savings. V4 launched in early 2025.
Legacy
Uniswap v3 reshaped DeFi:
- Concentrated liquidity became the new AMM standard — Curve’s stable swap and CLMM math, TraderJoe’s liquidity book, Orca (Solana) all implemented variants
- The tick architecture became a reference implementation for AMM engineering
- NFT LP positions enabled new financial products (LP-backed loans, LP yield optimization)
- By 2024, Uniswap v3 processed more daily volume than many centralized exchanges in total
Social Media Sentiment
Last updated: 2026-04
Uniswap v3 is admired technically but has a complex reputation. Traders love the depth and low slippage on major pairs. LPs are divided — professional LPs (MEV searchers, market-making firms) profit on v3 while passive retail LPs frequently underperform. The “Uniswap is extractive toward LPs” narrative has grown louder with MEV research. The Business Source License (restricting commercial forks for 2 years) was controversial — many viewed it as Uniswap Labs compromising DeFi principles for competitive moats.
Related Terms
Research
- Adams, H., Zinsmeister, N., Salem, M., Keefer, R., & Robinson, D. (2021). Uniswap v3 Core. uniswap.org.
— Primary source. Sections 2 and 6 (concentrated liquidity math and swap logic) are the technical core.
- Adams, H. & Zinsmeister, N. (2020). Uniswap v2 Core. uniswap.org.
— The v2 whitepaper; understanding v2 is prerequisite for appreciating v3’s innovations.
- Loesch, S., Hindman, N., Richardson, M., & Welch, N. (2021). Impermanent Loss in Uniswap v3. SSRN 3872006.
— Quantitative analysis of impermanent loss dynamics specific to concentrated liquidity positions.