Uniswap v3 — Current Limitations and Supporting Protocols and Tools

Introduction

Uniswap v3 was announced in March 2021 and went live on the Ethereum mainnet in May 2021. The crucial update being ‘concentrated liquidity’, which allows liquidity providers (LPs) to select the range of the price curve along which to allocate their liquidity for any given asset pair. It’s a dramatic improvement in capital efficiency — the platform can now facilitate more transactions and generate higher fees for each dollar of liquidity provided (4000x the capital efficiency of Uni v2).

This step function in improvement is evident by the market share Uni v3 now commands of the DEX space. Within a few months, per Messari, it already accounts for 51% of DEX trading volume, outperforming both Uni v2 (17%) and its closest competitor Sushi (13%) as of August 2021.

While Uni v3 introduced significant improvements in capital efficiency, it also brought with it some drawbacks. Below we discuss some of those along with protocols and tools that are being built or should be built to address the shortcomings.

Increased Complexity / High Transaction Fees

• Increased complexity. To benefit from Uni v3’s capital efficiencies, LPs now must play a much more active role in managing their positions via range orders. LPs that instead prefer to be passive are forced to either (1) spread their liquidity across a wide price range, thus reducing swap fees earned (and essentially negating the key benefit of Uni v3) or (2) if they select a narrow price range around the market price, they take on a higher risk of impermanent loss and may potentially stop earning swap fees altogether if the market price moves outside of their selected range.
• High transaction fees favor larger LPs. With high gas fees on the Eth mainnet, it becomes uneconomical for smaller LPs to constantly rebalance their pools, as the gas fees and swap fees begin to eat into the earned swap income. Even if LPs who are comfortable with the complexity factor chose to actively manage their liquidity pools, the transaction costs may not justify doing so for all but the largest LPs.

Many protocols have now emerged that attempt to address these drawbacks. One category of such solutions is the automated liquidity provisioning tools. Protocols such as Visor Finance and Charm.fi have stood up Uni v3 tailored “smart vaults” for LPs to deposit their tokens into. These vaults in turn provide concentrated liquidity into Uni v3, and employ various automated strategies for periodic rebalancing around the market price and maintaining appropriate token ratios. Below we detail 3 of the more popular of such solutions available today:

1. Charm.Fi’s Alpha vaults (~$5m TVL)

• LPs add liquidity to one of Charm’s Alpha vaults (three are offered currently: USDC/WBTC, USDC/WETH, USDT/WETH)
• The vault in-turn deposits liquidity into the respective Uni v3 pool
• Every 12 hours, the vault rebalances its Uni pool via on-contract limit orders to manage the selected price range and token ratio
• Takes 5% of the Uniswap fee income which goes to cover gas fees required in rebalancing

2. Visor Finance (~$13m TVL in Uni v3 pools)

• Similar to Charm’s Alpha vaults, the key difference being that Visor uses Bollinger bands to determine its selected price range. Bollinger bands widen with increased volatility in the underlying asset — thus, during times of higher volatility, liquidity is spread over a wider price band leading to reduced fee revenues

3. Gelato’s G-Uni pools (~$4m TVL)

• Uses an active rebalancing strategy vs Charm and Visor’s passive rebalancing. As such, price ranges are selected and optimized off-chain. As well, tokens are outright swapped to maintain an equal ratio (thus paying higher swap fees and gas fees) vs. Charm and Visor’s passive rebalancing mechanisms through limit orders.

Other less popular automated liquidity strategies include Popsicle Finance, Lixir Finance and Aloe Capital. In addition, Mellow Protocol and Steer Finance are currently under development.

While these solutions were quick to launch following Uni v3, most are still in their early stages. A very limited number of token pairs are currently offered; there is limited data available on their performance / return profiles given recency; many vaults (e.g., Charm.fi) have reached capacity with no timeline provided on when capacity is expected to increase; many of these protocols’ smart contracts have also yet to be audited.

We believe these automated liquidity strategy protocols are an important piece of the ecosystem surrounding Uni v3. Addressing some of the above issues will be crucial to support the Uni v3 network. As this segment matures, we would fully expect more complex strategies — strategies currently utilized by professional market makers — to be offered in the market, catering to the passive retail LP base and enabling them to take advantage of Uni v3’s concentrated liquidity to earn outsized yields.

Non-compounding of Earned Fee Revenue

The nature of customized liquidity ranges in Uni v3 requires the protocol to utilize NFTs to represent LPs’ pool interests. A byproduct of this is that swap fees earned from the pool are not deposited back into the pool. They are accrued separately and thus do not compound until harvested and re-deposited, which incurs additional gas fees.

A second category of solutions is solving Uni v3’s non-compounding fees problem. These include xToken and Harvest Finance’s Uni v3 pools. Similar to the automated liquidity strategies, LPs deposit their tokens into each protocol’s respective vaults which in turn provide liquidity into Uni v3 pools. However, these vaults do not provide any automated rebalancing. Instead, they simply harvest and re-invest the Uni v3 swap fees earned on a regular basis on behalf of the stakers in the vault. LPs are thus able to compound their positions while saving on gas fees and potentially benefitting from a more favorable tax treatment.

In our opinion, these solutions by xToken and Harvest aren’t very innovative; they seem to be a “quick and dirty” attempt to tailor the protocol’s existing capabilities to Uni v3. Given Charm, Visor and the other automated liquidity strategies discussed above also provide regular fee harvesting, they are much more comprehensive and superior solutions. We believe the market and LPs will converge towards those rather than the more simplistic ones provided by xToken / Harvest Finance.

Composability / Non-Fungibility

Uniswap’s prior version v2 (as well as most other DEXs in the market currently) provide ERC20 tokens to represent LP’s pool interests. These in turn could easily be used as collateral across the DeFi landscape. Due to the non-fungible nature of Uni v3 pool interests, they present unique challenges from a composability perspective and their ability to be used elsewhere in DeFi.

Currently, none of the leading lending protocols (Aave, Maker, etc.) accept Uni v3 NFTs as collateral for loans facilitated on their platforms. Since liquidity is the oil that the DeFi engine runs on, we believe it’ll be important for these lending protocols to create tools to accept Uni v3 NFTs as collateral. For example, a tool that allows LPs to deposit their non-fungible Uni v3 LP tokens into a vault, which then provides a corresponding fungible ERC20 wrapper to be used elsewhere in DeFi as collateral. This will significantly increase the appeal of providing liquidity on Uni v3 and further integrate Uni v3 into the ecosystem.

Liquidity Mining

Uniswap and Compound popularized liquidity mining in 2020 which played a tremendous role in kicking off DeFi summer. DeFi would likely not be nearly as prominent as it is today if it were not for liquidity mining. Due to the technical nature of Uni v3’s non-fungible LP tokens, however, liquidity mining was not immediately available upon launch in May. It was as recently as July that Uniswap deployed staker contracts on Eth mainnet which now allow for liquidity mining on Uni v3 pools.

Uni v3 liquidity mining is only now beginning to be utilized. Becoming the very first protocol to take advantage of Uni v3’s staker contracts, Index Coop on August 20th announced their liquidity mining program for the DPI-ETH Uni v3 pool. If prior liquidity mining programs on Uni v2 are any indications, it’s safe to assume this program should be successful in attracting additional funds to the pool. As the staker contracts become more “battle-tested”, we would like to see additional protocols put in place Uni v3 liquidity mining programs to incentivize liquidity.

In addition to 3rd party protocols implementing liquidity mining programs, we believe Uniswap itself should also provide mining incentives via its Uni tokens much like it did with Uni v2 in 2020. While transaction volume on v3 has overtaken v2, the number of wallet addresses providing liquidity on v3 still lags that of v2. This goes to show that v3 has successfully attracted the professional market makers, but has a long way to go in migrating retail LPs. A liquidity mining program that incentivizes v2 LPs bringing their liquidity to v3 could be another successful tool in increasing Uni v3’s network dominance.

Siloed Liquidity and a Fragmented Market

The introduction of Uni v3 adds yet another DEX in an already crowded marketplace that has seen liquidity fragmented among various DEXs (Uni v2, Uni v3, Sushi, Balancer, Bancor, Quickswap on Matic, Pancakeswap on Binance, etc.)

To solve this, many aggregators have now come online (1inch, Matcha, Paraswap) that allow users to quickly search across all DEXs and provide traders with optimal pricing. Upon Uni v3’s launch, aggregators quickly integrated with v3. They have played an important role in routing transaction volume to Uni v3, as orders originating from the aggregator will get fulfilled on Uni v3 if it provides superior pricing.

Where the aggregators fall short currently is their ability to perform cross-chain transaction. An aggregator will only execute trades from a DEX on the same chain as the originating order. E.g., if a user on Polygon chain is seeking to swap USDC for ETH — and the best pricing is found on Uni v3 mainnet, the aggregator will not execute on Uni v3 mainnet and instead settle for inferior pricing provided by one of the DEXs on Polygon. There is a need for an “aggregator of aggregators” tool that can compare and execute across all L1 and L2 chains. While creating such a solution certainly would have its technical challenges, we believe such a tool would serve a much-needed vacuum in the market. It would promote healthy competition among DEXs, and would also lead to greater market share aggregating the provider that offers the best pricing. More times than not, Uni v3 does tend to have the best pricing due to its concentrated liquidity — a cross-chain “aggregator of aggregators” tool could thus further cement Uni v3’s market leadership and also serve as an effective Uni v3 distribution channel to users on other chains.

Conclusion

Uni v3 is a remarkable improvement in capital efficiency over both v2 and other DEX’s in the market today. However, it does come with multiple drawbacks. Luckily there are already many solutions that have been built to address these— ranging from automated LP strategies, fee harvesting vaults, liquidity mining, and DEX aggregators. These have all played a key role thus far in propelling Uni v3 to market dominance.

Additionally, there is still room for further tools to support the Uni network— expanded automated LP strategies; tools that enable Uni v3 composability; liquidity mining programs; and a cross-chain ‘aggregator of aggregators’ are just a few possible solutions that would serve a current market need that the community may wish to consider.