Gelato Network: Facing the automated Web3 world
Research institution: Mint Ventures
Researcher: Flying little toe (飞翔的小拇趾)
The last update time: 2021.12.11
1. Key points of the report
1.1. Core investment logic
Gelato provides automation services for smart contracts and intends to become the automatic execution layer of Web 3. In the long run, smart contract automation is in line with people’s expectations for the future, and Gelato has a certain first-mover advantage as the industry is in the ascendant. In the short term, the business logic of becoming a DevOps (software development (Dev) and IT operations (Ops)) outsourcing service provider in the crypto world is also feasible. Gelato has the first-mover advantage in the track, the support from the developer community, and the resource support brought by VC (venture capital). It recently won the first place in the BSC MVB3, laying a solid foundation for its expansion on BSC. But it also faces the threat of a strong competitor -Chainlink Keepers.
1.2. Main risks
We believe that the risks faced by Gelato are as follows: the risk of too strong competitors, the difficulty of capturing the token value, the risk of smart contracts and the risk of unsustainable competitive advantage. See the section of 3.4. Project Risk for details.
2. Project overview
2.1. Business scope
Gelato Network is a decentralized robot network for Web 3 developers. It can support the automatic execution of smart contracts on EVM-compatible blockchains. Currently it supports public chains include Ethereum, Polygon, Fantom, Arbitrum, BSC, and Avalanche. Its core service is simply to “automatically execute operations for smart contracts after certain events”. The main use cases of Gelato today include:
i. Provide limit order service to AMM (automatic market maker) DEX
ii. Provide “liquidation protection” services for lending protocols
iii. G-UNI, an automated management tool for Uniswap V3 LP
At present, Gelato’s main services are provided to other project parties, and their services are generally integrated into other project parties and used by users. In addition, they also have Sorbet Finance directly for DEX users (including limit order function and G-UNI function), and Cono Finance directly for lending users.
2.2. Historical development and roadmap
It can be seen that Gelato Network is a project that was launched very early, and it has very close cooperation with the mainstream EVM public chain and other DeFi projects. In its service of “automatically execute operations for smart contracts after certain events”, it has covered a wide scope of trigger events on the same chain.
In terms of future development planning, Gelato is focusing on expanding the scope of trigger events so as to include the events triggered by real-world events off the chain, events on other chains, or other WEB 2 applications into the scope of trigger events.
In 2022, Gelato plans to integrate the off-chain trigger conditions (for example, automatically purchase US$1,000 worth of ETH on the first day of each month), and realize the cross-chain automation between Ethereum’s L1 and L2 (that is, an event on L1 automatically triggers a certain behavior on L2).
In 2023 and beyond, Gelato plans to realize the automation between WEB 2 and WEB 3 applications (for example, automatically monitor and buy related tokens on the ETH chain when Elon Musk tweets about an animal), and projects to be compatible with non-EVM public chains (for example, an event on the SOL chain triggers a certain behavior on the ETH chain).
In addition, Gelato also plans to start offering token rewards this year to incentivize developers to make more use of Gelato Network’s functions, and to achieve the decentralization of the smart contract execution robot network as soon as possible (currently all smart contract robots are still run by the Gelato team, and other robots are not supported to join the Gelato network temporarily).
2.3. Business details
Next, we will respectively introduce Gelato’s main business today.
2.3.1. AMM limit order
For stocks, futures, foreign exchange or traditional CEX, the main trading mode is to use limit trading, so people have a better understanding of limit orders than AMM trading.
The limit order service provided by Gelato can meet this demand. The logic of an AMM limit order can be simplified as follows: “buy or sell token A if the price of token A (relative to token B) reaches a set threshold”.
The limit order function is the earliest automated function implemented by Gelato. They first provided limit orders to Kyber. Later, with the development of DeFi, they expanded this limit order service to more DEX giants. Currently, Gelato provides limit order services to these DEXs, including Uniswap on the ETH chain, Pancakeswap on the BSC chain, Quickswap on the Polygon chain, SpookySwap and SpiritSwap on the Fantom chain, and TraderJoe on the Avalanche chain. Among them, users access the limit order entries of Uniswap, Pancakeswap and Traderjoe through Gelato’s Sorbet Finance, while Quickswap, Spookyswap, and SpiritSwap directly introduce the limit order function on their official websites.
From a technical point of view, the limit order function is actually not difficult to implement. For example, 1inch and matcha, which mainly do aggregate transactions, and Sushiswap (on Polygon) provide limit order services. However, for DEX, the implementation of this function is cumbersome and requires certain development and maintenance costs. For Gelato, after its limit order function is improved, the marginal cost of providing services to other DEXs is not high, so the development of such cooperation is easy.
From the perspective of the result, this cooperation is a win-win for the three parties: users obtain the function of limit orders that can be used for free; DEX can have a decentralized and effortless way to meet user needs; and Gelato can gain brand exposure and a certain degree of user acquisition. According to their recent trends, Gelato actively provides limit order services to the leading AMM DEX on each EVM chain. On the one hand, it can establish a good relationship with each EVM chain, laying a foundation for their future cooperation in cross-chain automation and other deeper levels. On the other hand, it also hopes to quickly touch the traffic on each chain through DEX.
2.3.2. Lending liquidation protection service for debt position
The logic of the lending liquidation protection service for debt position can be simplified as: “If the price of the collateral falls to a certain threshold, it will sell a part of the collateral to repay the debt.” Generally speaking, a lending protocol has a liquidation threshold, that is, when the collateral-to-debt ratio falls below a certain value (such as 110%), liquidation will be triggered. The service provided by Gelato requires users to set a threshold (greater than the liquidation threshold of the lending protocol). When a user’s collateral-to-debt ratio is lower than the threshold set by the user, some collateral will be sold to repay the debt, so as to retain the collateral for the user as much as possible.
Liquidation is a necessary part of all lending or margin businesses, and liquidation usually occurs when the collateral drops below the collateral level required by the protocol. Generally speaking, both the lending users and the lending protocol do not want liquidation, so the liquidation protection service for debt position has its value:
For lending users, liquidation is obviously undesirable. This is mainly because liquidation will be accompanied by a certain percentage of liquidation penalty, resulting in a net loss of assets.
For the loan protocol, the timely liquidation itself is neutral, because it can maintain the overall security of the protocol. However, if the liquidation is not timely, it may also appear the similar black Thursday tragedy experienced by Maker at March 12th, resulting in a heavy pressure on the protocol debt and pessimism about the protocol. On the other hand, from the perspective of long-term development, lending users are the base of the lending protocol. The volatility of the crypto world itself is already very violent, so frequent liquidation on this basis will fundamentally reduce the willingness of all users to participate in lending behavior, further reducing the income of the lending protocol in the long term. Therefore, although the liquidation income is a non-negligible piece of income for all lending protocols, they hope that liquidation can occur as controllable as possible from the perspective of long-term development.
In addition, Gelato’s vault automation service in collaboration with Instadapp achieves a better liquidation protection effect for debt position through asset/debt transfer between protocols. The logic is as follows:
When a user’s ETH-A vault is close to the liquidation line, the vault automation function will move assets and debts to other vaults with a lower collateral rate through flash loan on the premise of ensuring safety (including Maker’s ETH-B vault, Aave and Compound), so as to ensure the safety of the user’s collateral to the greatest extent when the collateral and debt remain unchanged.
Theoretically, the combination of “automatic sale of collateral” and “vault automation” will form the most ideal scenario for automatic lending liquidation protection for debt position. That is, when the health factor of personal collateral is reduced to the threshold value:
· When there are other lending protocols with lower collateral rate requirements (usually with a higher interest rate), it will switch to the one with lower collateral rate requirements first;
· When it does not exist, it will sell the collateral to repay the debt according to the setting and ensure the safety of the remaining collateral.
Of course, the realization of this ideal situation requires users to authorize multiple protocols, and it may be difficult to implement the setting that does not rely on Instadapp DSL (DeFi Smart Layer) (for Instadapp, please refer to our previous article https://mp.weixin.qq.com/s/E8xvG5lU933xeAiJT61Ltg).
2.3.3. G-UNI: a position management tool for Uniswap V3
The V3 released by Uniswap gives LP a fine operation space, however, for users, they often encounter the problem that their liquidity positions may easily get out of range. Users here include not only ordinary LP farmers who want to earn transaction fees, but also project teams who want to better incentivize the liquidity of their tokens.
The logic of G-UNI provided by Gelato can be simplified as “if the token price is far away from the set threshold, rebalance the LP to near the current transaction price and do the auto-fee-compounding regularly.”
The main benefits of using G-UNI are as follows:
· Always rebalance LP positions around current market prices, and always maintain capital efficiency. This is particularly important for projects that intend to increase the liquidity of their tokens.
· Automatically reinvest the fees earned by Uniswap V3 back to Uniswap for compound interest.
· Under the G-UNI mechanism, the same amount of G-UNI LP at any time always corresponds to the same amount of two tokens. That is to say, the G-UNI LP held by the user is a homogenous ERC-20 tokens (G-UNI’s LP on Uniswap v3 is still non-homogeneous, however the LP issued by G-UNI is homogenous), thus making Uniswap v3’s non-fungible liquidity positions fungible and facilitating integration by other protocols.
Current users of G-UNI include: Instadapp, Zerion, Rari Capital, Fei Protocol, Float Protocol, etc. G-UNI’s current TVL is around US$130 million, or 3.1% of Uniswap V3’s total TVL, which is higher than other Uniswap V3 position management tools.
In particular, it is worth mentioning that Maker supported G-UNI’s DAI-USDC LP as its collateral on September 21 of this year. This is the first Uniswap V3 LP supported by Maker, demonstrating Maker’s recognition of G-UNI.
In addition to the above three main application scenarios, Gelato is actually used in many other scenarios, such as:
· Qidao, the stablecoin protocol on Polygon, uses Gelato for auto-compounding
· The aggregation protocol Bprotocol uses Gelato for arbitrage trading
· NFT fragmentation protocol Unicly uses Gelato to distribute staking rewards
· Funds management protocol Parcel uses Gelato to pay the salaries of DAO members
And so on.
2.3.4. Gelato Ops
The above functions are all customized scenarios by the Gelato team for cooperating project parties. The Gelato Ops released by Gelato on November 18 allows developers to create automatic execution rules for any contract for free (until the end of 2021). This allows individual developers to create use cases for automated contract execution with the help of Gelato’s robot network.
At present, Gelato Ops supports three trigger conditions for contracts without parameter requirements, namely: timing execution, execution once available, and execution triggered by behaviors of other smart contracts. For contracts with parameter requirements, it supports two trigger conditions: preset parameters and trigger by reading other contract parameters, which can cover the trigger conditions of smart contracts on the chain widely.
Currently, Gelato Ops supports 6 chains of ETH/Polygon/Fantom/Arbitrum/BSC/Avaxlanche. At present, users have created more than 200 tasks, of which the most active ones are from the Polygon chain. Polygon currently performs more than 4000 tasks per day.
2.4. Team background
The two co-founders of Gelato, Hilmar Orth and Luis Schliesske, are both developers, who wrote the core functions of Gelato products from the beginning. The two of them have been close friends since they were in college and have been working together ever since. Before Gelato, they also co-founded a startup that focused on helping large European companies use smart contracts to explore new business models. Later, they participated in a series of hackathons such as ETHParis, ETHBerlin, ETHCapetetown, and Kyber Defi Hackathon, and achieved good results and influence. That’s why they were able to get Grants from Gnosis and MetaCartel and founded Gelato Network.
According to official website information, the core team of Gelato has a total of 12 people. In addition to the two founders, there are five developers, two product managers, two growth managers, and one community manager. In addition, according to the recruitment website, Gelato is still recruiting BD manager, BD director, marketing director, as well as front-end and back-end developers.
On the whole, Gelato is a medium-sized crypto team with reasonable staff allocation, and from the perspective of recruitment, the team has a strong willingness to continue to expand in BD and the market.
2.5. Investment background
Gelato has a total of 3 rounds of financing, including two rounds of private placement and one round of public placement. The details are as follows:
The seed round took place in September 2020 and raised US$1.2 million. Investors include: IOSG, Galaxy Digital, D1 VC, The LAO, Ming Ng, MetaCartel, Christopher Jentzsch. The token cost for this round is US$0.019.
In September 2021, Gelato announced a US$11 million fundraising. Investors include: Dragonfly, Parafi, IDEO, Nascent, Stani Kulechov (founder of Aave). The token cost for this round is US$0.2971.
The public placement, also in September 2021, raised US$5 million at a cost of US$0.2971.
On the whole, the investment background of the team is good, and the team has sufficient funds at the current stage.
3. Business analysis
3.1. Industry space and development prospect
From the classification of the crypto world, the smart contract automatic execution business done by Gelato falls into the broad category of infrastructure/middleware.
Specifically, Gelato provides services for the automatic execution of smart contracts, which basically means “automatically execute operations for smart contracts after certain events occur.” We look at the short-term and long-term separately:
In the short term, Gelato has an obvious business logic, that is, to become an outsourcing service provider for Web 3 DevOps (software development (Dev) and IT operations (Ops)).
In the world of Web 3, there are a large number of functions that need to be triggered on a regular basis or under specific conditions, such as regular auto-fee-compounding, regular salary payment, liquidity rebalancing, and so on. Before there was a service provider like Gelato, if a project team wants to automate these behaviors, it needs to build a complete set of robot programs that triggered the operation of smart contracts. It will take a lot of time to establish, operate, and maintain these robots. In particular, a large amount of work is actually repetitive work from a higher perspective. For example, all limit order functions are the same, whether it is for Uniswap or Quickswap. Fei Protocol and Float Protocol also require official liquidity for their governance tokens to ensure the transaction depth of the current price as much as possible.
From this perspective, developers can outsource their web3 DevOps demands so that they can focus on building core products instead of reinventing the wheel. In addition, outsourcing to a professional service provider such as Gelato can also reduce the risk that the project team may cause from a single point of failure.
For Gelato, under the premise that the core logic of the product has been relatively complete, as long as it can stimulate a complete smart contract robot system, the marginal cost of serving one more project team is significantly lower than that of the project team itself. It also makes this business logic established.
In fact, Gelato has two main competitors, Keeper.network and Chainlink Keepers, who primarily target this market too.
In terms of the long-term vision, Gelato actually hopes to become the automatic execution layer of Web 3. They can connect more trigger events, including events from the real world off the chain, cross chain, and from the Web 2 world, so as to realize the vision of making smart contracts smarter.
In the current WEB 2 world, the services provided by IFTTT (ifttt.com) established in 2010 are exactly what Gelato hopes to achieve in WEB 3. IFTTT is the abbreviation of “If this, then that”, and the core function it implements is also “automatically execute operation B after some event A occurs.” For example, the popular “If someone posts a new tweet, sync to my TG/Discord”, “Notify me if the temperature drops”. Or after connecting to a smart device, you can realize automatic functions, such as “go home and turn on the light, or turn off the light when leaving home”, “close the garage door after the car leaves”.
According to Crunchbase data, IFTTT has received four rounds of financing totaling US$62.5 million, including US$7 million in Series A financing led by a16z in 2012, and US$30 million and US$24 million in financing received in 2014 and 2018 respectively. This also shows the recognition of this model by Web 2 capital.
Therefore, in the long run, when Gelato establishes a perfect smart contract automatic execution robot network to ensure the robustness of the smart contract execution system, Gelato can start to work on the “trigger”:
· Incorporate more real-world behaviors into smart contract triggers and keep this process decentralized
· Incorporate actions on other chains into smart contract triggers
· Incorporate the behaviors of Web 2 applications into smart contract triggers by connecting to the notifications or interfaces of Web 2 applications
Moreover, with the help of the existing Gelato Ops and effective incentives, more developers’ trigger ideas can be incorporated into the entire automated smart contract system. At that time, Gelato will be able to make smart contracts truly smart. Smart contracts in the world of Web 3 will be able to react in real time to the real world and to what happens in the world of Web 2, and Gelato will likely become a bridge between the real world as well as Web 2 API and Web 3 smart contracts.
In summary, Gelato’s industry is in the early stage of development, with broad prospects and many possibilities, in terms of Gelato’s practice on Web3 and the situation of its Web 2 “rival” IFTTT.
3.2. Competition analysis
In fact, many services provided by many protocols overlap with those provided by Gelato, such as 1inch and matcha that provide limit order services with aggregation transactions, and Sushiswap. Similarly, Visor, Charm, and Yin also provide Uniswap V3 LP management services, however, we believe that these do not constitute a competitive relationship with Gelato. This is because the “smart contract automation services” provided by these protocols are only a means to serve their main business (trading or asset management), whereas the purpose of Gelato is to provide “smart contract automation services”.
The ones that really compete with Gelato from the perspective of business logic are keeper.newtork (KP3R) and Chainlink.
The business of keeper.newtork is carried out in a form similar to crowdsourcing: each user can publish a job of contract execution type. Developers can pick up a task after binding a certain KP3R, and they can obtain the reward set by the job publisher after completing the task. At present, there are 46 jobs on keeper.network, a considerable part of which are yEARN, Hegic, and KP3R’s own jobs, with an obvious Andre Cronje mark.
Chainlink officially released Chainlink Keepers on its mainnet in August this year. This is a product potentially competitive with Gelato. The goal of Chainlink Keepers is also to automate smart contracts, and the current use cases that Chainlink Keepers can implement are exactly the same as Gelato.
We believe that the Web 3 devops outsourcing track is not technology-oriented, but more product-oriented and resource-oriented.
From the perspective of product, Gelato has been working on this track for a longer time, with a broader product range. In addition, it is more focused on this business, and more likely to customize some functions for project teams. In this regard, kp3r, constrained by its founders and team, has not actively expanded this business, with slow progress. Although Chainlink’s products have been launched for a short time, it has covered its use cases in the DeFi field comprehensively.
From the perspective of resources, Gelato has a good developer community influence, investment background, and rapid public chain expansion speed, while Kp3r has the support of Andre Cronje’s projects. But since this is a business for developers, Chainlink has obvious advantages in this area because its oracles and VRF (Verifiable Random Function) services can naturally reach more developers. This is evident from the fact that Chainlink keepers has established cooperative relationships with DeFi protocols such as Aave, Synthetix, Bancor, and Alchemix soon after it was launched.
In addition, Gelato has an implicit relative advantage over Chainlink and KP3R in that its token rewards will be more focused on the smart contract automation business. KP3R’s tokens are currently in full circulation, and there is no plan to mint new tokens. This will affect its use case expansion in smart contract automation. However, Chainlink has other more core businesses (such as oracles), and its token reward cannot be focused on the smart contract automation business.
What’s more, the KP3R team doesn’t seem to be focused on expanding smart contract automation at the moment, and the founder’s commitment has been in doubt, which is also an X factor of its development.
Of course, the entire track is still in the early stages of development and does not yet require involution (fierce competition). But in the long run, Chainlink keepers will be in full competition with Gelato in terms of business form. This is not friendly to Gelato, which is currently in the early stages of development. Whether it can better motivate developers and users to form a certain first mover advantage and network effect through tokens may be the game changer for Gelato.
3.3. Token Model
3.3.1. Token distribution
The total amount of Gelato Network governance token $GEL is 420,690,000, and its distribution is as follows:
· 50% is allocated to community development
· 4% is allocated to the public offering in September this year, of which 1.68% will be locked up for six months.
· 21% is allocated to private investors, and half of the investors’ tokens will be released in September 2022 and September 2023.
· 25% is allocated to the team, of which 15% is allocated to the current team and 10% will be allocated to the future team. After the team’s tokens are locked for one year, 25% will be released, and the rest will be released linearly in three years.
The current circulating amount of tokens is about 9.77 million (almost all come from public offerings).
3.3.2. Token use cases
Gelato’s native token $GEL has two main use cases, governance and staking (not yet live).
Staking means that if developers want to become robots in the network (thus obtaining token rewards from the system’s additional issuance and fees paid by users), they need to stake a certain amount of $GEL tokens. When they perform well, they can get rewards, and when they perform poorly, they will be punished. However, currently Gelato’s web robots are not open to ordinary users, and no more details about this have been announced.
As can be seen from the token use case, Gelato has no intention of pegging any revenue to the $GEL token.
3.4. Project risks
We believe that the risks of the project in descending order of importance are as follows:
3.4.1. Risk of too strong competitors
Chainlink Keepers is Gelato’s main competitor. Compared with Gelato, Chainlink has a stronger brand, more rigid demand developer access channel, richer on-chain and off-chain synchronization experience, and richer team fund reserves. It also can form a bundling effect with oracles and VRF services. Thus, it is not easy for Gelato to win the competition with Chainlink Keepers.
3.4.2. Risk that its tokens are difficult to capture value
IFTTT, Gelato’s equivalent product in the Web 2 world, has undergone a painful commercialization process. IFTTT’s final revenue model is mainly derived from corporate users, and it still exists in a form more similar to corporate service outsourcing. Outsourcing implies difficulties in customization and scale effect. According to Growjo data, the expected revenue of IFTTT in 2021 is only US$8.6 million, which is not good for an app with nearly 20 million active users and one billion monthly calls.
Gelato is likely to encounter this problem in subsequent product promotion due to the similarity of product forms. In addition, Gelato has no intention of linking its protocol revenue to its tokens in its planning of token use cases.
3.4.3. Risk of smart contracts
The services provided by Gelato require direct access to users’ funds. Although Gelato’s contracts have also been audited by Certik, the risk of smart contracts is inevitable. This is also proved by the G-UNI risk problem on December 12.
3.4.4. Risk of unsustainable competitive advantage
As mentioned earlier, the Web 3 DevOps industry does not have a very high barrier to entry, so it is not difficult for outsiders to enter this industry. Both the project team and the end users are only sensitive to functions, without loyalty to the service provider that provides them. Therefore, Gelato’s current competitive advantage may not be sustainable.
4. Valuation
4.1. Five core questions
What business cycle is the project in? Maturity stage or the early and middle stage of development?
This project is a project with a long history of development. Its product path is clear, and several core products have been relatively stable. However, as far as its planning is concerned, the current product is still only a small part of its planning and in the early stages of development.
Does the project have a solid competitive advantage? Where does it come from?
The competitive advantage of the project comes from the first mover advantage in the track, the support of the Ethereum developer community, and the resource support brought by VC (including the potential support from Binance after winning the BSC MVB3 winner recently).
Is the medium and long-term investment logic of the project clear? Is it in line with the industry trend?
Smart contract automation tools have use cases on all public chains that use smart contracts. Among the current top ten crypto projects, ETH, BNB (BSC), SOL, ADA, DOT, LUNA are all smart contract public chains, and Gelato has cooperated with the current top 5 EVM public chains in TVL. In the longer term, the significance of smart contract automation is also consistent with people’s vision of the future. In short, the project is in line with the general development trend of the industry.
What are the main variables in the operation of the project? Are these variables easy to quantify and measure?
In the short term, Gelato is a ToB-based project. The main variable of its operation comes from whether it can effectively introduce developers into the Gelato system. The measurement method is to observe the cooperation trend between Gelato and other project parties and monitor the activity of the Gelato robot network.
In the long run, its operational variables come from its implementation scheme of its cross-chain triggers and off-chain triggers, whether it can effectively attract users and find the path of realization.
What is the management and governance of the project? What is the level of DAO?
The project adopts the DAO method for governance, and the voting for its first governance proposal began on December 1. The time is too short to judge the level of its DAO.
4.2. Valuation
Gelato will not charge any fees for its products until the end of 2021, which is related to the fact that the project is still in the early stage. Moreover, since the project is now multi-chain development in parallel, and there is no statistics on the number of calls of the entire network, we are unable to simulate the revenue and profit. Therefore, we do not conduct the valuation calculation of P/E and P/S of the project itself.
In terms of horizontal comparison, among Gelato’s competitors, KP3R’s current core use case is to obtain the protocol revenue of Fixedforex, which has nothing to do with Gelato’s business. In addition, Chainlink Keeper has just launched, and many data are unavailable, so we are unable to accurately compare the horizontal valuation of Gelato.
Here we only list the fully circulating market cap and the circulating market cap of the above three protocols for comparison.
4.3. Summary
Gelato provides automation services for smart contracts and intends to become the automatic execution layer of Web 3. In the long run, smart contract automation is in line with people’s expectations for the future, and the industry is developing in the ascendant, so Gelato has a certain first-mover advantage. In the short term, the business logic for Gelato to become a DevOps (software development (Dev) and IT operations (Ops)) outsourcing service provider in the crypto world is also feasible. Gelato has the first-mover advantage in the track, the support from the developer community, and the resource support brought by VC. It recently won the first place in BSC MVB3, laying a good foundation for its expansion on BSC. But it also faces the threat of a strong competitor — Chainlink Keepers.
5. Reference
https://thedefiant.io/the-defi-future-is-automated/
https://medium.com/gelato-network/gel-token-sale-date-and-whitelist-opening-fc697c814662
Token market cap data: https://www.coingecko.com/
Protocol TVL data: https://defillama.com/home
