An Overview of Uniswap v3 LP Automated Manager
Uniswap v3 generates more LP fees than Uniswap v2 by concentrating liquidity in a narrow range.This article is from Jinse Finance, and the original author is ビビドット (@vividot).
In this article, I will introduce a service called the Uniswap v3 LP Automated Manager. The so-called Automated Manager means it can execute rebalancing and compounding operations for Uniswap v3 LP on behalf of the user.
Before diving into the main content, I want to clarify that I am a supporter of the Automated Manager, and the purpose of writing this article is not to undermine it.
I view the Automated Manager as a type of Uniswap v3 investment fund, as its performance depends on its rebalancing strategy. Additionally, I believe these investment funds need to be transparent, which is why I started writing this article.
I hope the Automated Manager will improve over time.
Please note that this article is based on available information from June 27 to July 3, 2021.
Why do we need a manager?
Uniswap v3 earns more LP fees than Uniswap v2 by concentrating liquidity within a narrow range.
Therefore, those who want to earn fees will provide liquidity within a price range close to the current market price, as shown in the figure below, which causes Uniswap v3's liquidity to typically form a peak shape around the current market price.
On the other hand, if the market price exceeds the strategy range, liquidity providers cannot earn LP fees. Users outside the price range must wait for the market price to return to that range or immediately remove liquidity and provide liquidity again within another price range.
The latter behavior is referred to as "rebalancing," and the purpose of the Automated Manager is to help users automatically execute this operation. Additionally, in Uniswap v3, the manager will claim and reinvest LP fees that have not been automatically added to the Uniswap v3 pool.
The Automated Manager also has a common advantage, as the issued LP tokens can be made into ERC-20 tokens, making them easy to use for collateral and liquidity mining.
Considerations for Rebalancing Strategies
Before introducing the manager, I want to discuss rebalancing strategies, as the performance of the manager depends on rebalancing.
1. How to rebalance
According to Charm Finance, there are two ways to rebalance: (1) passive rebalancing and (2) active rebalancing.
"Active rebalancing" is a method of providing new range liquidity by swapping tokens managed during the rebalancing period.
The downside of this method is that each rebalancing incurs a small amount of swap transaction fees.
On the other hand, "passive rebalancing" does not swap tokens during operation but provides as much liquidity as possible for the tokens in the treasury, then offers the remaining tokens as unilateral liquidity. The term "unilateral liquidity" is not clear, but the point is that it is as close to the current market price as a limit order, so if the price moves in that direction, the second position will also become active and earn LP fees.
For example, the figure below illustrates how, when the manager controls 2500 USDC and 1 ETH, 1 ETH and 2000 USDC provide liquidity, and the remaining 500 USDC provides unilateral liquidity.
When the ETH price fluctuates between $1750 and $2000, the role of the unilateral liquidity portion is to gradually convert the surplus USDC into ETH.
Its advantage is that there are no fees during rebalancing. However, if the market fluctuates for an extended period (whether in a bull or bear market), the LP fees you earn will be less compared to active rebalancing (because unilateral liquidity does not work, and liquidity on both sides will also decrease).
2. How to choose a range
Compared to the method of rebalancing, how to choose the rebalancing range is more important.
When considering LP income, the best range is naturally as close as possible to the future price range. (Of course, this is unknown)
When only considering reducing impermanent loss, it is best to make the range as wide as possible, just like Uniswap v2 LP.
Therefore, each manager team will use various methods to choose the range. However, before we open the lid, we do not know which method will perform better.
3. Manager fees
Setting a performance fee is quite common; this fee is mainly used to pay for the gas fees of rebalancing and to repurchase governance tokens. It is important to note that the fees need to consider the strategy.
For example, suppose Manager1 uses a wide range strategy (small impermanent loss, low LP fees), while Manager2 uses a narrow range strategy (large impermanent loss, high LP fees). In this case, if Manager1 and Manager2 have the same performance fee (10%) and their performance (LP fees and impermanent loss) is also the same, then Manager2 will pay more rebalancing fees, which is undesirable for users.
4. When to rebalance
As competitors increase, the timing of rebalancing may become more important. Additionally, when the market moves in one direction, the unilateral liquidity of passive rebalancing becomes completely useless. Therefore, in trending markets, passive rebalancing needs to perform some active trading to concentrate liquidity near the market price to increase fee income. However, if done too much, it may face a downside, such as unilateral "selling low and buying high."
Introduction to Managers
1. Charm Finance (Alpha Vault)
Two days after Uniswap v3 launched on the Ethereum mainnet (May 7), Alpha Vault became the first automated manager to test the waters.
After being audited, they launched two new vaults on July 2, with a funding cap of $1 million (currently full).
Rebalancing method: Passive rebalancing
Range: Specified by the development team, centered around the time-weighted average price (TWAP) algorithm
Commission: 5%
The figure below shows the evolution of the strategy range for this manager's USDT-ETH. The deeper light blue portion represents the base liquidity, while the softer light blue portion represents unilateral liquidity.
From the figure above, it can be seen that this manager is trying to reduce impermanent loss (IL) by using a wide range to earn stable returns. If the market fluctuates quickly, the losses will not be too great.
By the way, its range width is determined through backtesting.
HP: https://alpha.charm.fi/
2. Visor Finance (Active Liquidity Management)
Visor Finance's Active Liquidity Management launched a beta version on May 18, and they also introduced Gamma Strategies to fund strategy research (the actual status is unknown).
Rebalancing method: Passive rebalancing (with swaps as needed)
Range: Based on Bollinger Bands indicators
Commission: 10%
Visor Finance displays the range on their website, but it is difficult to see clearly. You can see the gray Bollinger Bands, and when you hover over them, it appears that the range is set based on the Bollinger Bands indicators.
cite https://vault.visor.finance/analytics?view=usdt-eth
The author created the following chart using on-chain data. Thus, the deep green portion represents the base liquidity of the strategy, while the light green portion represents unilateral liquidity.
Even considering different time periods, these two charts look different. The Visor chart implies that the changes in liquidity range occur much more frequently than what the on-chain data shows.
The strategy I see on-chain is to specify a narrow range and earn high LP fees while accepting IL.
Additionally, once the range is set, the strategy delays realizing IL by not rebalancing before exceeding the range, making it a powerful strategy in range markets.
HP: http://visor.finance
3. Popsicle Finance (Sorbetto Fragola)
Popsicle Finance initially managed cross-chain liquidity and launched Sorbetto Fragola on June 26 to manage Uniswap v3 liquidity.
Rebalancing method: Active rebalancing
Range: Specified by the development team, centered around TWAP
Commission: 10% (repurchase $ICE)
The figure below shows the evolution of the USDT-ETH range. Sorbetto Fragola actively rebalances, so liquidity is always placed in a single position at each rebalancing.
I do not know how they determine the range, but in terms of width, it is between Charm and Visor. They have just launched this manager and may not have finalized their strategy yet, but due to the swap transaction fees at each rebalancing, they seem to be trying to reduce the frequency of rebalancing.
HP: https://popsicle.finance/dashboard
4. Harvest Finance
Many people may have heard of Harvest Finance, which actually offers a Uniswap v3 vault. On the other hand, this vault does not perform rebalancing but only collects and compounds fee income, so it is slightly different from the managers mentioned in this article.
If the range changes significantly, the developers will deploy vaults with different ranges, and users satisfied with the vault will increase liquidity.
HP: https://harvest.finance/
Note: The following section introduces automated managers that have not yet launched products.
5. Gelato Network (G-UNI)
G-UNI provided by Gelato Network is currently used for liquidity mining and liquidity supply management of the InstaDapp governance token INST and ETH.
Other pairs will also be offered in the future through Sorbet Finance provided by Gelato.
Rebalancing method: Active rebalancing
Range: Specified by the development team (briefly mentioned in a Medium article, but not detailed)
Commission: 10%
I previously wrote about Gelato Network, which aims to be a network of bots executing smart contracts, and here they will execute rebalancing strategies.
By the way, regarding the range for rebalancing, I asked the team, but they have not provided detailed documentation yet, only saying it will be announced soon.
HP: https://defi.instadapp.io/inst-pools
6. Aloe Capital
Aloe Capital claims to be a "crowdsourced liquidity allocation protocol."
The fundamental idea behind Aloe Capital is to use prediction markets to select rebalancing ranges.
Its users are divided into two categories: "liquidity providers" and "stakers."
"Liquidity providers" are those who want to use the manager to earn LP fees. The "stakers" stake Aloe Capital's governance token Aloe and predict the price of the target asset. The rebalancing range is then determined by aggregating the predictions of the stakers. If their predictions are correct, the stakers will receive more Aloe rewards; if their predictions are incorrect, their staked Aloe will decrease.
Interestingly, traditional price prediction is a zero-sum game (a negative-sum game when considering protocol fees), whereas Aloe Capital, by charging a portion of LP fees, becomes a positive-sum game.
Rebalancing method: Passive rebalancing
Range: Determined based on the predictions of the stakers
Commission: 17.5% (7.5% reserved for the pool, 5% for the development team, 5% for stakers)
One concern is that the protocol may not work due to a negative spiral. Without enough "liquidity providers," it cannot attract enough "stakers." And without enough "stakers," it cannot attract enough "liquidity providers."
HP: https://aloe.capital/
7. Lixir Finance
This may be the project that first proposed the concept of automated managers on Medium, but it has not launched yet.
The project's contracts have been implemented and audited, so I expect it to launch soon.
Additionally, the company raised $1.7 million through private placement and then raised $3.27 million through LBP. By the way, the final price of the project's LBP was $8.03, but due to the bear market and project delays, its token price is currently below the private placement price of $1.53.
After a quick glance at the source code, its rebalancing strategy seems to be as follows.
Rebalancing method: Passive rebalancing
Range: TWAP range + unilateral range at the current price
Commission: Variable unknown
HP: https://lixir.finance/
8. Mellow Protocol
The project's Twitter account has been frozen, and communication in the Telegram group is not possible, so our only source of information is the project's official website.
The white paper states that their goal is to "implement derivatives on Uniswap v3," and the first step is to implement leveraged liquidity provision. The meaning of leveraged liquidity provision at the beginning of the paper is slightly different from liquidity provision optimization, so they may have changed their strategy.
I have not checked the formulas in the white paper, so if you are interested, you can take a look yourself.
HP: https://mellow.finance/vault
9. Steer Protocol
Although there is no documentation or white paper, Steer Protocol was recently selected by Sushiswap as a Sushi Incubator product.
This product can be roughly understood as allowing users to create, participate in, and execute any DeFi strategy.
One of the strategies it introduces is automated liquidity management for Uniswap v3.
Honestly, I think this product's ambition is too great and may struggle to succeed, but since it was selected as a Sushi Incubator product, my expectations for it have increased.
HP: https://steer.finance/
Conclusion
After the announcement of Uniswap v3, many people mentioned that it could achieve XXX times the liquidity effect, but the reality is not so rosy. Considering impermanent loss, many strategies perform worse than Uniswap v2.
Currently, if you want to provide liquidity for Uniswap v3, you have two options: use a manager in the main pools or look for secondary pools and maintain them manually.
At present, well-performing LPs are those that have not exceeded their range and have set a wide range, as shown in the tweet below. However, if the price of ETH drops or rises significantly and exceeds the range, the performance of this strategy may be poor.
On the other hand, the reality is that the performance of the managers introduced in this article is not as good as imagined, which does not mean that you can earn a lot of money just by investing. In fact, I calculated the results using DuneAnalytics, and you will see that despite different strategies, all managers are struggling with impermanent loss (IL). Depending on the time period, Uniswap v2 LP may perform better.
Finally, regarding the question "If I want to use a vault, which one should I use?" If I compare performances in a short time, the result might be Charm > Fragola > Visor, but I cannot say for sure.
Additionally, some projects are dishonest because they display very high APR returns on their websites while completely ignoring the existence of impermanent loss (IL), so I do not recommend using their products.
I will closely monitor the performance of these strategies, as it requires research over a longer period, but interestingly, the results indicate that the wider the range of Charm (Alpha Vault), the better the performance of the strategy.
P.S. I am most interested in Aloe Capital; however, I know from the results of Charm, Visor, and Sorbetto Fragola that merely accurately predicting the range is not enough. So I wonder how to reward stakers or how to aggregate the predictions of stakers.
P.P.S. I believe the performance results I calculated include LP fees and impermanent loss, but I just started learning SQL and DuneAnalytics a week ago, so there may be some errors. I would greatly appreciate it if someone knowledgeable could check this SQL.
