Uniswap Becomes an L2

TL;DR

• Uniswap is launching its own Layer 2 chain, "Unichain," as part of Optimism's Superchain, enhancing Ethereum's rollup-centric architecture.
• Unichain will contribute 15% of net revenues to the OP Collective, offering token holders fee-sharing opportunities by delegating tokens to validators.
• Uniswap’s dedicated chain will cut fees by over 95% and reduce transaction times to ~250ms, improving the user experience.
• This migration will significantly reduce Uniswap's activity on Ethereum mainnet, impacting Ethereum’s fee generation and overall economic activity.
• As Uniswap and other gas-consuming applications move to L2s, Ethereum’s gas prices may drop, but its fee burns and valuation could suffer

Last week, the top decentralized exchange protocol Uniswap announced that it is deploying its own EVM layer 2 chain as part of Optimism’s Superchain, dubbed Unichain. As arguably the most successful DeFi application in the industry (Uniswap has facilitated ~$2.5T in swap volume to date), this is a major move that underscores Ethereum’s continued migration towards its rollup-centric architecture. This is also another example of vertical integration happening by the largest protocols (Uniswap already has its own front-end and wallet) and a continuation of the trend of any application that is sufficiently large ultimately becoming its own chain. DYDX was one of the first examples of this when they moved to a dedicated L2 and then again to a sovereign Cosmos chain. Aevo, Lyra (now Derive), Zora and Frax have all spun up their own L2s, and MakerDAO is going this direction with its SVM-based ‘NewChain’ in the future. 

As a member of Optimism Superchain, Unichain will contribute 15% of net revenues back to the OP Collective. Most importantly, this is also a way to find a more regulatorily-compliant ‘backdoor’ mechanism to return fees to token holders, as holders will eventually be able to delegate tokens to validators for a share of the fees earned. The added customization that comes with a dedicated chain will also enable Uniswap to reduce fees (relative to mainnet) by over 95% and bring transaction times down to ~250ms via sub-blocks. As part of the Superchain’s roadmap, native Interoperability across all member chains will enable same-block cross-chain messaging. However, this interoperability is siloed to member L2s – seamless interoperability for any chain outside of the Superchain ecosystem will be a larger issue in the near term as those technologies (e.g. shared sequencing) are very much still being ironed out. 

Undoubtedly, this migration is a major win for the OP Stack and the broader Superchain ecosystem. However, the impact to Ethereum’s mainnet economic activity is likely substantial. 

Over the past 30 days, Uniswap’s contracts (Universal router, V2 router, V3 Positions NFT) have collectively been the largest fee-generating contracts on Ethereum, responsible for ~7,400 ETH burned, out of a total of ~42,500 burned over that time period (nearly $20M). Put another way, Uniswap is responsible for 17% of all ETH base fee generation over the past month. Since the Merge slightly more than 2 years ago, Uniswap’s contracts have resulted in ~308,000 ETH burned (out of ~1,791,000 total), or $671M out of a total $3.9B burned. In short, Uniswap has historically been the economic engine of Ethereum.

Source: TheBlock

Compounding the impact, beyond just the Uniswap transaction activity leaving mainnet, so too will much of the activity generated from many of the other major gas consuming applications. If one looks at the remaining top 20 contracts, BananaGun, Maestro, 1Inch, MetaMask’s swap router, 0x, and many of the largest MEV bots all rely on or are tied to the substantial trading activity stemming from Uniswap. To get an idea for Uniswap’s importance to these top apps, ~80% of BananaGun’s and Maestro’s volumes are on Ethereum (i.e. Uniswap), and in 2Q 2024, 42% of 1Inch’s volumes went through Uniswap. If the bulk of Uniswap’s L1 activity largely leaves to L2s, the knock-on effect of all of these other applications migrating with it could be significant. MEV loss - and more importantly, the loss of priority fees paid to validators due to MEV - will also be substantial. Uniswap’s primary role MEV extraction is difficult to understate:

Source: Top Ethereum pools for 30D MEV extraction via EigenPhi

Further, transaction activity for stablecoins (USDT, USDC) is increasingly moving to L2s or alt-L1s given the low costs to transact. The majority of USDT transfers are now on Tron and USDC activity on Base now accounts for over 10% of USDC activity. Outside of trading, these two contracts are some of the biggest demand drivers for Ethereum gas, and as DeFi activity continues its move to L2s, stablecoin activity will increasingly follow. 

The ultimate impact of this move will likely lead to improved trading experiences for Uniswap users (cheaper, faster, less MEV extraction), more efficient liquidity provision and reduced fragmentation, while also providing stronger value accrual to the token. It will also likely lead to reduced gas prices on Ethereum mainnet as this signifies a major reduction in potential congestion vectors on the chain. While this helps improve the usability of the chain and helps support the EVM’s long term positioning, it is likely highly detrimental to Ethereum’s fee burns (e.g. revenues) and thus valuation (as justified by fundamentals) for the foreseeable future.