Comprehensive Analysis of Uniswap V3: Curve Killer, Commercial Exploration, More Suitable for Professional Market Makers

This article is an original piece by Chain Catcher, authored by Loners Liu.

In the early hours of today, Uniswap officially announced the specific details of the V3 version and updated the latest white paper. This update is also related to the directions described in the project's 2021 roadmap, which include enhancing automated market-making capabilities, continuing the development of AMM technology, and exploring scaling solutions.

Specifically, the Uniswap V3 version mainly includes five key features: non-uniform liquidity distribution, flexible fees, range orders, more advanced oracle design, and the Layer 2 version on Optimism. Below, Chain Catcher will analyze the potential impact of this V3 version on the crypto world based on these features.

1. Non-uniform Liquidity Distribution

In the Uniswap V3 version, LPs can inject liquidity within a specific price range. If the trading price is concentrated within this range, they can earn the same trading fees with less capital.

When users trade on Uniswap, they need to pay a 0.3% fee. For example, if a user trades dy USDT for dx ETH in the ETH/USDT trading pair, Uniswap will first deduct 0.3% * dy from the user's principal, then calculate the dx to be given to the user. After providing dx, the 0.3% * dy principal will be injected into the LP Token, effectively increasing its value.

After the new version, LP Tokens will be distributed in a non-uniform manner, meaning that only trades within the pre-set liquidity range will earn fee rewards. Thus, the LP Tokens for a specific price range are reduced, allowing the same amount of capital to earn more fee dividends under unchanged trading volume (i.e., unchanged LP Token appreciation).

However, this also complicates LPs' choices regarding liquidity ranges. While narrower price ranges can improve utilization, they also mean greater impermanent loss risk. If the price of one token rises beyond the range, it may lead to "selling out," and if the price falls below the range, all tokens will be converted to the relatively depreciating token, creating "holding the bag" risks.

Previously, impermanent loss was countered by accumulating the 0.3% trading fees and price corrections, but smaller liquidity ranges mean lower fault tolerance, inadvertently increasing the risk of trying to cover impermanent losses through continuous fee accumulation in volatile markets.

Therefore, this design is more suited for professional market-making teams, who can carefully select liquidity ranges based on their risk control and use financial instruments like call options to hedge against price impacts and maximize returns. The official also reminds users to set a portion of their liquidity within a range and use part of their funds to hedge against downside risks or invest in other projects for returns.

The introduction of franchise pools in SushiSwap's liquidity expansion plan embodies such a strategy. Some exchanges (like Binance) integrate the Sushi protocol, funneling liquidity into SushiSwap to provide returns for their users as liquidity providers.

Having exchanges act as professional market-making institutions benefits them by offering users more potential returns, which helps retain users. For Sushi, it can gain more liquidity and provide a better user experience. It remains to be seen whether Uniswap will attract more professional market-making teams through such designs, while ordinary users may entrust their funds to such institutions, similar to how retail investors often choose to join a mining pool for more stable returns.

2. Flexible Fees

Compared to Uniswap V1 and V2, the V3 version of Uniswap grants the owner key to liquidity providers, allowing them to change the fees for each pool, offering three different fee levels: 0.05%, 0.30%, and 1.00%.

According to official data, LPs within the 0.1% liquidity range will see a capital utilization rate 4,000 times higher than before, while the V3 version can technically support liquidity ranges as low as 0.02%, theoretically boosting capital utilization by up to 20,000 times. Of course, this also means more complex contract designs, requiring more gas fees, making it more suitable for implementation on Layer 2.

For low-risk trading pairs like stablecoins, a relatively small liquidity range can be set (the official notes that the exchange ratio for the DAI/USDC trading pair is concentrated between 0.99-1.01). Even with reduced fees, higher returns can still be achieved compared to before. This is undoubtedly a significant blow to Curve, which focuses on stablecoin trading, leading many KOLs on Twitter to label Uniswap V3 as the Curve killer.

In special market conditions, such as a protocol hack, a 1% trading fee can be set to reduce users' desire to participate in trading, minimizing LP losses.

Additionally, Uniswap V3 can allocate 10% to 25% of fee revenues as protocol fees, transferring them to an address held by Uniswap, which is not charged by default. In fact, the "protocol fee" mechanism was introduced in Uniswap V2, allowing community governance to activate a unified 5 basis points (16.66% of LP fees) protocol fee, with Uniswap V3's protocol fee being relatively more flexible.

However, this mechanism was previously criticized for harming liquidity providers' interests. Last year, increasing liquidity was seen as the best way to strengthen one's moat, as ample liquidity means better trading depth and can attract more traders.

Uniswap developers believe that the design of protocol fees is a necessary function to support Uniswap's growth and sustainability, which can be voted on by the community to decide whether to activate this mechanism. Just as YFI ultimately chose to issue more tokens and Aave opted to merge its existing lend while reserving some funds for ecosystem operations, all to ensure the protocol's long-term development.

If Uniswap also has a continuous stream of protocol fee income, this revenue can be used to upgrade the protocol and the entire ecosystem, increasing acceptance and ultimately allowing liquidity providers to gain greater benefits. Theoretically, this could compensate LPs for any losses incurred under this mechanism. It also marks a significant step towards commercialization for Uniswap, as previously all fee income was given to LPs.

Moreover, one of the most criticized aspects of Uniswap is that the governance token UNI lacks a clear mechanism for capturing value. In Sushiswap, only 2.5% of the 3% fee charged to traders is allocated to liquidity providers, with the remaining 0.5% used for repurchasing Sushi, integrating the fee value into the governance token.

Although the V3 version update did not adequately address this issue, I believe that whether a token can capture value is not the most critical point at present. As the DEX with the most users, it would be straightforward for Uniswap to propose a fee buyback and burn mechanism. However, for those DEXs without users, even if all fees are used for buybacks, they may not attract many real users.

3. Range Orders and NFT LP Tokens

Even before the release of Uniswap V3, many predicted that it would adopt an order book model and support NFT trading. Although we now know that it differs from expectations, careful observation reveals innovative aspects of Uniswap's product design.

Range orders simply mean setting a price range in advance and depositing a certain token. If the market price enters this range, an asset will be sold along a smooth curve in exchange for another asset, while also earning trading fees in the process.

In Uniswap V1 and V2, the "value" between LP Tokens was the same, allowing them to be represented in ERC-20 token form. However, in V3, each LP corresponds to different liquidity ranges, so they can only be represented in a non-fungible form as NFTs. This primarily affects the composability between various DeFi protocols: for instance, several LP Tokens could previously be used as collateral on MakerDAO to generate DAI, as well as the newly launched AMM Market on Aave.

4. More Advanced Oracle Pricing Mechanism

Although the bZx flash loan attack raised doubts about Uniswap's role as a price oracle, since the implementation of the TWAP Oracle (Time Weighted Average Price) in the Uniswap V2 update, the cost of attacks has significantly increased. It has now integrated with over 50 protocols, including Compound, Augur v2, and Empty Set Dollar, second only to Chainlink.

Uniswap's price information transmission mechanism allows developers to calculate the average price of a token based on its price volatility over several blocks (which can also be represented as a time interval between two timestamps). This time interval can be set according to user needs.

Specifically, a segment of blocks is selected on-chain as the time interval, and the cumulative price of a token within this segment (the token's price in each block) is divided by the timestamp interval (the ending block's timestamp minus the starting block's timestamp) to derive the average price of the token over this segment.

While Uniswap provides token prices, it does not store any historical prices on-chain. DApp developers need to select a cumulative price over a period to calculate the average price of a token during that time.

Uniswap V3 has made significant improvements to the TWAP Oracle, allowing TWAP data to be read on-chain at any time, with records available for up to 9 days. V2 effectively weighted the current price of a token (or the exchange ratio of a trading pair) over time, while V3 stores this data in an array.

Based on the V3 Oracle's array storage of historical data, Uniswap can build more advanced oracles. For example, to prevent derivatives from "piercing the needle," exponential moving averages (EMA) can be applied to the basis over a short period, as well as simple moving averages (SMA), and mechanisms like Coinbase Oracle that filter outliers that deviate significantly from expected volatility.

Currently, the prices of many long-tail assets primarily depend on prices on Uniswap, so this revision provides significant security for the design of derivatives for some long-tail assets and further solidifies its pricing power.

Additionally, despite significant design improvements, the gas fees required to use the V3 version of the oracle have decreased by 50% compared to V2, and the cost of calling and calculating TWAP in external smart contracts is also much lower.

5. Software Business License

The core code of Uniswap V3 will be protected under the Business Source License (BSL), restricting improper commercial use for two years. This makes Uniswap the first major DeFi protocol to adopt a code licensing mechanism, possibly related to last year's complete code replication by Sushiswap, which temporarily exceeded the locked amount of the original project.

Uniswap states that its copyright protection restrictions can be changed through protocol governance, allowing for a transition from Business Source License protection to GPL (General Public License) ahead of schedule.

6. Contract Audits and Bug Bounties

Currently, the contracts of Uniswap V3 have been audited and certified by three authoritative security audit firms: Trail of Bits, ABDK, and samczsun, with the audit reports publicly available in the project's GitHub repository. Uniswap has also launched a bug bounty program, offering up to $500,000 for anyone who discovers significant vulnerabilities within the next 30 days.

Conclusion

Like last year, the V3 version chose to release its white paper in March and officially launch in May. Before the official launch, it will first be deployed on the Ropsten, Rinkeby, Kovan, and Görli testnets, followed by the deployment of the Layer 2 version on Optimism. Compared to Sushiswap's choice of a zkRollup solution, Uniswap places greater emphasis on Optimism's compatibility with Ethereum EVM.

In addressing the issue of capital utilization, V3 employs a more flexible liquidity provision method and flexible fees, providing traders with a better trading experience. Meanwhile, SushiSwap has chosen to collaborate deeply with the YFI ecosystem and utilize bentobox lending to absorb its liquidity overflow.

In the short term, there is no clear winner between the two approaches. As Mable Jiang, a partner at Multcoin Capital, stated: Uniswap and SushiSwap share the same code, but they attract different users and have evolved into entirely different solutions. User behavior and patterns will continue to influence the direction and roadmap of each project, and this divergence is likely to continue developing in the future.