Moderator: Alex, Research Partner at Mint Ventures
Guest: Lawrence, Researcher at Mint Ventures
Recording time: 2024.12.27
Hello everyone, welcome to WEB3 Mint To Be initiated by Mint Ventures. Here, we continue to ask questions and think deeply, clarify facts, explore reality, and find consensus in the WEB3 world. We clarify the logic behind hot topics, provide insights that penetrate the events themselves, and introduce multiple perspectives.
This episode is the third episode of the "Current Status and Future of Web3 Track" series of podcasts, and it is also a special episode. Let's talk about the recent very popular on-chain trading platform project, which is also a Layer1 project: Hyperliquid. In the last two episodes, we talked about the two topics of Crypto AI and Defi. In the subsequent series of programs, we will invite corresponding guests to talk about Meme, public chain, Depin, games & social, Payfi, and related topics of web3 policy.
Disclaimer: The content discussed in this podcast does not represent the views of the institutions where the guests work, and the projects mentioned do not constitute any investment advice.
A one-sentence introduction to Hyperliquid
Alex: Hyperliquid can be said to be one of the most popular projects in the market recently. Our researcher Lawrence has also been researching this project recently, so I invited him to talk to us about all aspects of this project. We will release a review research report on Hyperliquid later. If you want to know more about this project, you can read the report. Before we officially start, a warm reminder that this podcast involves more professional concepts about the project. Although we will try to express them in more popular language, if you are a listener who has no knowledge of the project, some parts may still require you to check the text version and do some searches to understand.
Now let's get to the point of today. Since Lawrence is an old friend, I won't introduce him to you any more. Let's talk about the details of the Hyperliquid project. The first thing we want to talk about today is that many listeners may not know much about the Hyperliquid project. If Lawrence wants to use a simple and popular expression to explain what Hyperliquid is to a Web3 novice user, how would you explain it?
Lawrence: If I have to describe it in one sentence, I think the "Binance on the chain" mentioned by Messari is relatively simple and direct. Of course, it may not be very accurate, because Hyperliquid's main business now is a derivatives exchange and a spot exchange, and it also has an EVM chain like Binance. Of course, the BNB chain has been running for more than three years, and Hyperliquid's EVM chain has not yet been officially launched. But overall, the main business scope is still relatively close to Binance.
Hyperliquid’s Value Proposition
Alex: OK, "Binance on the chain". The point of comparison between it and Binance is that it is a trading platform itself, and it has its own exclusive chain, a Layer 1. So now when we talk about it, we will say that it is not only a trading platform, we emphasize its public chain attributes. So compared with existing exchange products, such as Binance you just mentioned, or existing on-chain trading platforms, for example, we know that there are actually many various projects like on-chain derivatives trading and even spot trading, such as GMX, spot has Uniswap and so on. Compared with this type of exchange product or DEX product, what is the value proposition of Hyperliquid? Or what differentiated value does it provide?
Lawrence: Compared with centralized exchanges, Hyperliquid is actually an order book product in terms of its core trading mechanism, including all its rates, funding rates, and transaction fees. The various mechanisms are basically consistent with centralized exchanges, or simplified on these basis. In terms of experience, it is basically close to centralized exchanges such as Binance, but the UI/UX experience may not be as good as Binance. The main difference between it and centralized exchanges is that it is KYC-free. Compared with these decentralized derivatives exchanges, Hyperliquid is relatively similar to DYDX, both of which are based on the order book as the core architecture. Like GMX, including Jupiter, they use a set of innovative product mechanisms to solve the performance problems of the on-chain order book, and Hyperliquid is inconsistent with other exchanges in this regard.
Compared with spot DEX, Hyperliquid is not a spot exchange that everyone can access. Although it is permissionless, it is not free. The subject matter traded on it still has to go through a set of mechanisms. Although this mechanism is decentralized, it is different from the familiar Solana Pump.fun. It does not allow everyone to issue coins. This is a difference between it and spot DEX. Overall, I think the value of Hyperliquid as a product lies in that, on the one hand, its business scope is close to Binance, but it is a decentralized project; on the other hand, everyone will think that its decentralization still has some flaws. I think it is a good balance between decentralization and product experience in the current market environment. Compared with all centralized exchanges, it is obviously more decentralized; and compared with all decentralized exchanges, its experience may be better.
Alex: I see. Actually, we talked about this issue yesterday. According to our definition, Hyperliquid is an on-chain trading platform. Its feature is that it is KYC-free, which means that everyone can trade. Although most people can trade normally on Binance, there are always some people who hope that they do not need to do KYC, and there is such a demand. However, we just talked about that its characteristics can be defined as decentralization, but in my opinion, its characteristics may be more in line with two attributes. The first is that it is permissionless, which means that its accessibility is very high and there are no permission restrictions. The second is that the core information of the entire exchange operation is transparent and public. I think these two points are indeed the two points that most decentralized projects have. But can these two attributes fully represent decentralization?
Lawrence: If we look at the Layer1 currently running on spot and derivative exchanges, it can actually be understood as a stand-alone chain. Because all nodes are in the hands of the team, including all funds, the funds on the entire chain are a three-quarter multi-signature, a two-thirds multi-signature plus another cold wallet signature. From this perspective, even if everyone thinks that the BNB chain has few nodes, the BNB chain also has more than 40 nodes, and the same is true for Solana. And its number of nodes can currently be understood as only four, not to mention compared with Ethereum and Bitcoin, and compared with Solana and BNBChain, it is also very small.
This is why everyone would say it is a stand-alone chain. In fact, judging from the current situation of Hyperliquid, if they want to change this, I don't think it is particularly difficult. Because it is a new thing with a token that has been online for less than a month. There are already many experienced validators on other chains. As for Hyperliquid, the currently running Layer1 is not open to these validators, but it is hard to say whether it will be open in the future. From the perspective of validators, there are still many validators who want to join HyperEVM. If Hyperliquid wants to change its decentralization problem, I think it is relatively not difficult. Of course, the scope of decentralization is even wider, for example, its nodes may also need to achieve geographical decentralization, etc. Indeed, its current decentralization can be said to be very incomplete.
Alex: Can we understand it this way? If we compare it with the current large-scale trading platforms such as Binance, the service value provided by the two is actually quite similar, which is the trading of derivatives and spot goods. I think the main difference between it and Binance may not be whether it is decentralized. Because as you just said, it is actually a three-quarter multi-signature at present. We also know that Binance's asset cold and hot wallets are also controlled by multi-signatures, and they are definitely not controlled by a single point. So in this regard, there is not much difference between the two parties. But the difference is that Binance's exchange operation data, capital flow, including its business data, and program operation logic are a black box. But after all, Hyperliquid's entire contract, including core operation data, is on the chain, and it is transparent and verifiable. In addition, it does not require KYC, which means that everyone can access it. So perhaps the advantages brought by this permissionlessness and data transparency are a difference or value compared to the current large-scale trading platforms. Can we understand it this way?
Lawrence: Right.
Hyperliquid's overall product structure
Alex: Let's introduce the overall product structure of Hyperliquid. We just said that it is not only a trading platform, but also has its own Layer 1 operation plan. How is the design structure of its entire product planned, and what are its parts?
Lawrence: It currently has two products online. One is the derivatives exchange that they have been building from the beginning, which is also its main product. The other is the spot exchange that they have just launched recently. These two parts are actually built on a public chain called Hyperliquid L1. They actually already have a public chain, but this public chain is not EVM, and not all developers can develop on it. According to their current plan, this RustVM chain should mainly support these two core applications, which are derivatives trading and spot trading, which are the products of these two order books. This is the product they have launched. The product that it will launch in the future is HyperEVM.
HyperEVM is a Layer 1 of EVM. Of course, it is not clear whether it is Layer 1 or Layer 2. Because in their architecture, RustVM and HyperEVM are actually two execution layers, and the consensus layer is the same, that is, they use a set of HyperBFT, an algorithm based on the previous Meta consensus algorithm. This is a combination of two execution layers and a consensus layer. There is no clear precedent for how to implement it. In addition, they also want to make the spot trading on RustVM and the transfer on HyperEVM interoperable, that is, interoperable. Therefore, how to link HyperEVM with the parts that have been launched, and whether their relationship is more like two chains or more like a Layer 1 and a Layer 2, is not particularly clear at present. Of course, the team said that the two chains are interoperable, both are Layer 1, and both are main chains. However, from the current perception, the relationship between the two is not particularly certain.
In general, from the perspective of their architecture, there is a consensus layer and two execution layers, and the two execution layers are interoperable. One of the execution layers is currently running the core product, and the applications should be these two, which will not be open to all developers. The other execution layer is EVM, which is a code language that Web3 developers are more familiar with, allowing developers to develop various applications on it, such as lending, games, etc.
Alex: I just mentioned that although there may be two chains, there is a certain degree of interoperability between the two chains. A previous research report mentioned that one scenario of their interoperability is that the transaction chain can provide oracle price feed services to the EVM chain. Is it similar to this kind of interoperability?
Lawrence: Yes, oracle price feed is one, and it can also achieve interoperability of asset transfer. Of course, this is definitely not the full amount, because it is permissionless, and there are definitely more assets on EVM than on RustVM. Some special processing may be required, that is, only the core assets shared by both chains, such as HYPE tokens, which are used as gas on EVM, and also include native assets that already exist on RustVM. Based on the current information, this is not described very clearly.
Alex: The relationship between HyperEVM and RustVM may be similar to that between Binance and BSC. In fact, the asset transfer between them is relatively convenient, but in terms of specific execution and program operation, they are not in the same system. We can understand it as the relationship between an exchange and a chain issued by the exchange.
Lawrence: Yes, my current understanding is probably closer, or easier to understand, than the relationship between Binance and BNB chain, and between Coinbase and Basechain.
Alex: I just mentioned that the two modules belong to the same execution layer. They have a separate consensus layer called HypeBFT at the bottom, which is the so-called Hype Badge On Team fault-tolerant consensus mechanism. This consensus mechanism comes from the Meta team. As we all know, Meta previously worked on a global permissionless project called Libra. Later, the Libra project was suspended due to regulatory issues. Their consensus mechanism is based on the modified Libra BFT. Since the RustVM chain is already running, is this consensus algorithm currently online?
Lawrence: It is already online. They should have officially switched to HypeBFT in May this year. From May to now, all spot and derivative transactions have been on this RustVM execution layer.
Hyperliquid’s business model
Alex: OK, so we have a general idea of Hyperliquid's entire product structure, which is two execution layers, which can be understood as two chains, and the bottom layer is the consensus layer shared by the two chains. If you were asked to briefly describe Hyperliquid's current business model, what would you say? How is its business revenue distributed within the system?
Lawrence: Its current business model consists of two parts, a derivatives exchange and a spot exchange. The business model of its derivatives part is the same as that of a centralized exchange, which is that it charges fees, including handling fees and funding fees. Of course, the funding fees are not all charged by the team, but are a balancing mechanism for the entire market, and there are also liquidation fees on the other hand. The spot part has spot transaction fees, and it also has a listing fee that can be understood as a spot token, which is the auction fee for its HIP-1.
In general, the fees are divided into several parts. In terms of income distribution, at present, except for a part of Hyperliquid's income that is used to subsidize HLP, which is equivalent to adding liquidity to its official market-making fund, all its surpluses are currently flowing into HYPE tokens. HYPE's spot trading pair against USDC will directly destroy Hype, of course, the amount is relatively small, and the remaining part of all its fee income will now be used through their assistance fund, that is, the assistance fund, to repurchase HYPE tokens.
Actually, we just talked about Hyperliquid's products. Of course, its derivatives exchange is a very successful product. From the current stage, HYPE tokens can also be regarded as a very successful product. It uses this cold start form in a very short time, and coupled with its relatively radical mechanism for income distribution, basically all the income flows to users. This is the first point, because the team at least does not directly obtain income from the surface. In addition to the money it wants to distribute to HLP, the rest is all allocated to HYPE tokens.
Alex: What is the approximate ratio of profit distribution between its HLP and Hype tokens?
Lawrence: The team didn't tell you the ratio directly, but told you that all the fees went to these two directions. HLP is divided into three strategies, and the three strategies have their own PNL and HLP has its own PNL. The team probably subsidized PNL with fees, and by the end of November, it should be more than 40 million. Then the initial repurchase funds of AF are transferred on the chain, which can be checked. It is about 53 million, so the total is about 100 million. About half, half was given to HLP and half to HYPE tokens.
Alex: We can understand that the total revenue of Hyperliquid since its launch until recently was 100 million, and then maybe more than 40 million was given to HLP, and then nearly 50 million entered this fund. Most of the fund was used for the repurchase of HYPE, right?
Lawrence: Yes, the time point is the end of November, because after the end of November, the transactions of its AF fund address have increased a lot, and now there is no very good tool to visualize its AF address balance. But in fact, Hyperliquid's data in the last month is relatively good, so this data is from the time it was launched in 2023, and the official start of the points plan in November, about a year and a half of data.
Alex: Although the actual distribution may be about 50-50, the project's documents or public rules do not actually specify how the revenue from this product will be distributed. This has not been determined yet, right?
Lawrence: It only determines the destination, which is HLP and the aid fund, but it does not mark the specific ratio. I feel that this is also quite reasonable, because how much subsidy to HLP should be given is dynamic, because in fact, from July 23 to now, the market has basically been in a rising state for a long time. Generally speaking, from the perspective of derivative exchanges on the chain, retail investors as a whole are always long. Then the three strategies of HLP should still be a small profit, because its position can be seen on the statistical website of its official website. It basically holds a net short position in the bull market, and its own strategy is still a small profit. In this case, if it wants to maintain a profit of HLP, it must make up for it through handling fees, funding fees and liquidation fees. So there is no way to stipulate how much of this thing should be allocated to HLP.
Alex: But let’s go back to what we talked about before, that is, the spot trading platform that is now online, the handling fees and the auction fees for listing coins are all under the control of funds, right?
Lawrence: From the current situation, all the fees of Hyperliquid are uniformly allocated, so a part of all the income of the platform is allocated to HLP, including spot fees and listing auction fees. There are two types of spot fees. For example, for the HYPE-USDC trading pair, the fees that can be received are HYPE and USDC. The USDC part is also allocated to HLP and AF, and the HYPE part is directly destroyed at the contract level to reduce the maximum circulation.
(The original audio answer in this paragraph has some deviations, and the text version has been updated according to the actual situation)
Alex: What if it is other currencies, such as PURR, PURR and USDC transactions, is it handled in the same way? Are they uniformly transferred to the foundation's address?
Lawrence: It is unified. All USDC fees are transferred to HLP and AF. AF’s part is used to repurchase HYPE regularly, and the PURR part is also destroyed.
(The original audio answer in this paragraph has some deviations, and the text version has been updated according to the actual situation)
Hyperliquid’s spot auction mechanism
Alex: I see that the market is also paying close attention to Hyperliquid's spot auction. Could you explain this mechanism in more detail?
Lawrence: First of all, it is an open auction mechanism. All developers who want to participate can participate. It is a Dutch auction, which is a descending price auction mechanism. The first bidder can get the qualification. Then it has a scarcity treatment, that is, a round of auctions will be opened every 31 hours. Within these 31 hours, the price will drop linearly from twice the last transaction price to 10,000 U. The value of 10,000 U can be adjusted. Before it, it was not raised. This value was 100 U, of course, because there was not such a high demand at that time. From twice the last transaction price to 10,000 U, this is the current mechanism. In this process, anyone who wants to bid can bid, because the first successful bid will be auctioned at the current price to get such a qualification.
This qualification corresponds to a ticker of Hyperliquid's spot exchange, which is a trading pair, and its payment is to pay USDC. In the past ten days, the price of this auction is still relatively high. The highest was God on December 16. This ticker was auctioned for nearly 1 million US dollars, and it has been around 300,000 US dollars recently. This is still a relatively special mechanism. There were many voices in the market questioning centralized exchanges before, especially Binance, whose listing strategy was unclear and there may be some problems of interest transfer in the middle. Hyperliquid's strategy is open and transparent, with a transparent charging mechanism, and all the money it receives is currently fed back to the HYPE token. So from this perspective, it is equivalent to a response to the trend of this narrative.
Alex: As you said, in the past ten days or so, the listing fee is basically hundreds of thousands, the highest may be 1 million, and the average may be more than 300,000. This listing fee is actually comparable to most of the top second-tier exchanges, such as Bitget or GATE, whose listing fees may also be hundreds of thousands or at least hundreds of thousands depending on the situation of the project. So the influence of its listing, the pricing level that the project side is willing to pay for, has reached the level of these exchanges, right?
Lawrence: Yes. Of course, because the sample size is relatively small and the duration is relatively short, but this is real money that needs to be paid. For the project side, whether to list on GATE or Hyperliquid is a choice in terms of the budget for listing. This is still very objective. Of course, it may not be very accurate because of the short time, because after all, there are very few qualifications, and there may be relatively large fluctuations.
Alex: But I think a good point is that the entire auction process is very transparent, whether it is just a user who is watching, or a holder of Hyperliquid's tokens, or a shareholder of this project, or a participant in the game of listing coins. Once the information is transparent, everyone gets the same information, so it will become very efficient when making decisions, and there will not be a lot of unnecessary wear and tear, or unnecessary costs caused by misunderstandings. Who are the main sources of these projects that are currently auctioning Tickers? Because it seems that the earliest ones may not necessarily be me who auctions this thing, but more of a Meme operator. It seems that some project parties have recently started to auction this Ticker. What is the composition of its customers?
Lawrence: Yes, before Solv auctioned this Ticker in early December, most Tickers were still similar to domain name logic. Because it is unique, and the previous auction price is relatively low, the lowest is a few hundred dollars, so there are many common meme-type Tickers. Solv is a staking protocol in the Bitcoin ecosystem, and it is invested by Binance Labs. From it, we can basically see that on the one hand, it directly raised the price to more than 100,000 US dollars, and then basically exceeded 100,000 US dollars.
On the other hand, the participants are mainly project parties, and more of them are pan-entertainment projects. There are many games, for example, the most expensive ticker of GOD is a game invested by Pantera, and the ticker of MON is also a corresponding game, and Sovrn, these are all projects that cost 300,000 US dollars, they are all making games, and Anime is rumored to be a ticker of Azuki's Token. Then, including Cream of Machi, which is an old loan project and is often visited by hackers, it also bid for a ticker, and ETHC, another project of Machi, also bid for a ticker. This is also normal, because it is not that you can bid for something by spending a few hundred U or a few thousand U before. With the increase in the price of HYPE itself and the increase in the influence of Hyperliquid, this thing has become a spot listing fee. Many of these game projects are not only listed on Hyperliquid, but also released on HyperEVM. Some games are also prepared to run on HyperEVM. So it is equivalent to a deeper investment in ecological binding.
Hyperliquid's business data performance
Alex: OK, we just talked about the spot listing part. In terms of overall business performance, how is the core performance of Hyperliquid's entire business data since its launch? If we look at its core data, which indicators do you think are more critical?
Lawrence: For trading, the core is still to look at the trading volume and the transaction fees. First of all, its derivatives have no rivals in the decentralized derivatives exchange. Its competitors have a trading volume that is orders of magnitude lower than its counterparts in the past month, and few can reach 1/10 of its volume. Then in the past few days, its trading volume, including the amount of open interest, is about 1/10 of the corresponding data of Binance. And then including the mainstream trading pairs BTC, ETH, including some short-term hot new hot currencies, such as Penguin (PENGU) a few days ago, are probably of this magnitude.
According to Coingecko's derivatives exchange ranking, Hyperliquid ranks around 20th. Its data is about 1/10 of Binance, 1/5 of Bybit and Bitget, and about the same level as Kucoin and HTX. This is its derivatives trading part. As for the spot part, its spot trading volume fluctuates quite a lot. When its trading volume was the highest, it had a daily trading volume of more than 500 million US dollars. Among all the DEXs on the chain, this trading volume of 500 million US dollars can probably be ranked in the top five. There are only a few like Uniswap, PancakeSwap, Raydium and Aerodrome, which are the largest exchanges on Solana and Base. After Uniswap and PancakeSwap, Hyperliquid can be ranked fifth.
However, since Hyperliquid's spot trading can only trade the assets acquired through the previous auction, which is equivalent to only its derivative assets, the vast majority of the trading volume here is HYPE, so the fluctuation is relatively large, that is, the price fluctuation of HYPE will also have a relatively large impact on the fluctuation of its trading volume. I think the key indicators are also relatively clear. One is the total transaction amount of the contract, the second is the total transaction amount of the spot, and the third is the listing fee just mentioned, which is the fee that the project party is willing to pay for its ecosystem. I think these are the three most critical indicators.
Alex: We see these indicators, including transaction volume, position size, and the TVL of this chain. It took off significantly at the end of November, and then took off even higher in mid-December. If we only look at the TVL, before the end of November, I think the TVL of the entire chain was about 200 million. Of course, it may be higher than most application chains. But from 200 million to mid-December, when it was the highest, its TVL was still close to 3 billion, that is, 2.6 or 2.7 billion. What happened in the middle that caused its transaction volume and the amount of funds on the chain to surge by 10 times?
Lawrence: The main thing is the official issuance of the HYPE token and its continued rising price performance. I think from this period of time, Hyperliquid has done a very good job on the HYPE token product. First of all, they have an AF fund. If you pay attention to it earlier, it is equivalent to using all my previously accumulated income to buy back my own tokens. Because its token distribution actually has no investor share, 31% of the early circulation was all airdropped to its early users. This is equivalent to a turnover of chips, that is, many airdropped chips will be equivalent to being bought back by its team. In this way, its price will continue to rise. After its price rises to a certain order of magnitude, it will trigger widespread discussion on social media.
For example, some people are looking at when its market value will surpass Uniswap, and of course it didn't take too long to surpass it. Including its subsequent price increase, its ecological tokens have also had more wealth-creating effects, such as Farm, its native first test token Meme token PURR, PURR has risen to a market value of 500 million, and Farm has also quickly gained more than ten times or even twenty times the profit. This will cause a lot of funds to quickly enter Hyperliquid to get a piece of the pie. Of course, in recent times, in the past week or so, everyone has found that in fact, under the current mechanism of Hyperliquid, something like Pump.fun cannot appear. Because its Ticker has to pay 300,000 US dollars. Including a storm about its security issues some time ago, so TVL has continued to decline in recent times. This is probably its situation.
Alex: Yes, I think it has fallen back by more than 30% from the peak recently. After Hyperliquid issued the coin, it gained brand exposure, and the industry's attention soared a lot. Everyone is analyzing how its strong growth was achieved. It seems that from the performance just mentioned, although its growth data in the derivatives trading platform on the chain was good before the coin issuance, it is still far from the current level. So its turning point was after its coin issuance. Now we look back and see that it is the wealth effect of the coin issuance itself, which is very powerful, that has attracted a large number of traders to come to this chain to seek gold, and there are many arbitrage opportunities brought by the transaction, which has attracted many people who use large funds for arbitrage, forming a mutual promotion and rotation of its trading volume, on-chain funds, and traders. But now its price has reached a very high level, which may be tens of billions of US dollars. It seems that it is very difficult for the price of Hype to rise from 3 US dollars to more than 20 US dollars as before.
Including, as you just said, the growth of the number of its wealth-creating targets is not as large as that of SOL, it is limited. From a positive perspective, its supply is limited. From a negative perspective, the number of innovative projects it can carry out will also be limited, or its threshold is relatively high. In this case, the two factors that led to its previous increase of more than ten times were the rapid rise of Hype and the strong wealth effect of some early projects. Both of these seem to be stagnant or in a state of resistance. So where can its subsequent growth momentum come from? If it still needs to go up at this stage?
Lawrence: The subsequent upward momentum is that under its current mechanism, there will still be the auction gas fees, listing fees and handling fees mentioned just now. These data are continuing to rise. The auction amount is currently more than 300,000 US dollars a day, and if we collect transaction fees at a comprehensive fee rate of 20,000, this value will also be several million US dollars. So the daily net income should still be in the order of one or two million US dollars. At present, a relatively large proportion of this part will be allocated to repurchases, that is, its AF fund in the past may also accumulate for a period of time, and then use up all the bullets in its hand. Repeating past behavior is still there, but because the market value has increased a lot on the one hand, and the increase in its recent data is also slowing down on the other hand, this force must be weakened. So it is very difficult for it to increase 10 times from 3U to the highest 30U as in the past.
On the other hand, its current market value is already 30% of Binance in terms of FDV. It is now close to 30 billion US dollars in FDV. If it increases 10 times again, it will be close to the position of Ethereum. Even if it can be achieved, it will take a very long time. Because it has increased 10 times in a month, the magnitude is already very drastic. So I think under the current circumstances, there should not be such a big impetus for price increases in the short term. We may still have to wait for the emergence of some subsequent catalysts, such as the launch of its HyperEVM. Because HyperEVM will have its staking when it goes online, including HYPE can also do gas. After EVM goes online, some people's valuation system for HYPE will change. Because the valuation of public chains is much higher than that of exchanges. So we may have to wait for the emergence of some new catalysts.
Alex: Let's summarize the information just now. First, Hype's business data and technical performance are good, but its main performance surge, including the turning point of performance, is after it issued the coin. The surge in the price of the Hype token after it issued the coin brought about the wealth effect and the out-of-circle effect, which brought a large number of gold diggers and caused a surge in performance. This is also related to what we said earlier that in the more than one year since it went online, the AF Fund has nearly 50 million US dollars in funds. After Hype went online, it used this 50 million US dollars to repurchase all of Hype, which formed a lot of buying pressure in a short period of time, and quickly rose from a few US dollars to more than 10 US dollars, which shocked everyone.
This shocking advertising effect attracted a large number of people to come in anxiously. The current situation is that although it still has an income of more than 1 million per day, I just checked the data from Defi Lama. Its business income in the last 24 hours is basically about 1 million, which is an income from handling fees. Of course, part of this 1 million income should still be given to HLP, and the remaining 300,000 is a net income that can be used to support Hype. So, although the net income of one or two million US dollars a day is not small, compared with its initial strong buying of 50 million US dollars, it will still be more difficult to maintain such a high price. So let's talk about its next story, which is the problem of its EVM chain. Although it has not yet been launched, can you tell us what key information it has at present, such as its ecological construction plan, some of its basic or special mechanisms, including possible value recovery, etc.
Lawrence: Although it is not online yet, the testnet has been launched for a while. We can also see that there are some relatively well-known validators on its testnet, such as Nansen, Chorus One, and Figment, which are relatively well-known in the field of staking. These validators have participated in the verification of its testnet. On the other hand, there are many projects that want to be done on HyperEVM. In its DC, we can still see many developers asking questions every day, including some of its early community members. We currently have about 20 to 30 native projects, and there are at least a dozen projects that already have businesses on other chains and are preparing to launch projects on HyperEVM.
Alex: What are the categories of these projects?
Lawrence: There are all kinds of categories. First of all, there are some trading tools based on it, such as PVP.Trade, which is also a social trading product that people may use more in TG. Then there are the solv protocol for Bitcoin pledge mentioned earlier, and some other defi protocols. There are more NFTs, games, and memes, because at present, speculative funds still like relatively short-term and fast projects. At present, there is actually no way to do these projects on Hyperliquid RustVM, so there should be more NFT and Meme projects, such as the games corresponding to GOD and SOVRN mentioned earlier, which are quite rich in variety. Of course, there is no high cost for them to test it. For Hyperliquid, if it does not develop well later, this project will not be released there, and there is no high cost, because it has risen so much some time ago. There must be many projects willing to come, or at least willing to express their willingness to release it.
Alex: What are the current expectations for the potential value capture or value delivery of Hyperliquid tokens through Layer 1?
Lawrence: The most obvious one is HYPE Staking, because the team has not publicly stated how to empower, we can only guess based on similar projects in the past. The first is to use Gas. This is mentioned in his document. Hyper will become the Gas on EVM, but its HyperEVM must be a very cheap Gas, which will definitely not be high, so if you really want to say how much destruction can be brought, this may not be particularly high. This is the first point. The second point is Staking. Staking can share part of the income, because after going online, like dYdX, its income is shared with Staking, which is less risky from a legal point of view. Of course, the regulatory environment is not quite the same as before, and it is unclear whether there will be any problems with it directly sharing, but after all, if Staking is opened and then shared, it will definitely be no problem from the past. In fact, our previous analysis shows that Hyperliquid has already achieved a relatively extreme economic model. Except for the part that is to be given to HLP, all of it is given to HYPE tokens, but the form is not to say that I will share HYPE or U with you, but to say that I will buy it back directly from the market. In my current understanding, from the perspective of the economic model, that is, from the perspective of actual distribution, 60% of all revenue will be given to the Hype token. I think the increase in this ratio has little to do with whether HyperEVM is online. Whether this ratio can be increased depends more on how much money it needs to subsidize HLP. If it needs less money to subsidize HLP, then it can allocate more money to the Hype token. I think this is probably the case.
Alex: Actually, from this perspective, the current market value of Hyper is so large. Judging from its spot and derivative transactions, it has actually reached a level that is not cheap. The valuation may include a lot of expectations for its subsequent development of the HyperEVM chain. If this chain can do very well, for example, it can become the next Solana Killer worth looking forward to, I think it may have a market value of hundreds of billions, plus it has good business data on the derivative side to support it, so it is possible to rise again. If it is said that the EVM is average, there may not be much improvement, because if it is done, its valuation may have to be cut down further. What do you think?
Lawrence: It is possible. The core is that it has risen too much in the past month. In terms of its magnitude, there has not been a very decent correction. At most, it has fallen to 23 or 24. Its related catalysts are not easy to see in the short term. So the next stage may be that the price performance will not be so good.
Hyperliquid Team
Alex: I have seen many comments praising Hyperliquid for having a very small team. It seems that they had only about 10 people before, but I don't know what the situation is now. The reason why they say their team is small is mainly compared to those large centralized exchanges. We know that Binance has a large number of employees, basically several thousand people. Although Hyperliquid's business data has not surpassed Binance or reached the level of large exchanges, it is already very impressive. In your opinion, why can such a small team maintain such a high level of performance?
Lawrence: I think their entire marketing strategy is quite straightforward. Since Hyperliquid went online, there has not been much marketing budget. If we look at it from this point in time, it is equivalent to that all of its marketing budget comes from the funds accumulated before, that is, the funds for pulling the market equal the marketing budget. From the current point of view, the more than 50 million accumulated by the end of October is still quite worthwhile. Because most centralized exchanges have no obligation to disclose data, we cannot see it, but we can see that because Coinbase is a listed company, the core revenue structure of Coinbase is not completely consistent with that of exchanges such as Binance and Bitget, and it also has a custody business.
But we can see in the third quarter report that Coinbase has more than a billion in revenue, of which 800 to 900 million are costs. Its operating costs and R&D costs account for a large part of its revenue. In the case of Hyperliquid, it does not need to do a lot of functions like centralized exchanges, but focuses on its core functions, which means it does not need to recruit too many people, and then it has lower R&D costs. All its marketing budget comes from the fact that it does not do any SEO or very complicated channels, it just uses price performance and this kind of community spontaneous dissemination. I think from the results so far, this is bold on the one hand, and the results are also quite good on the other hand.
Alex: I think it is very interesting that Hyperliquid has such a streamlined staff structure while its competitors have such huge staff and management expenses. There are two possibilities. The first is that Hyperliquid has grasped the essence of this industry and this product or business model. It has done its best in its most critical functions and used the most effective methods in marketing, which has led to it saving a lot of unnecessary budgets and forming such a situation. Another possibility is that Hyperliquid's current performance level is temporary. It spent 50 million yuan in the early stage to pull the market, which is also the marketing cost, to form the current performance level. However, if it does not invest the same amount of funds to pull the market in the future, it may be challenging to maintain this performance level in the future.
If it is the first case, I think all trading platforms should reflect on why they can do this with such a small staff cost. If it is the second case, Hyperliquid may face many challenges in the future, and it may also need to increase a lot of additional expenses. For example, it may also need to expand its product line vertically, and it may also need more business teams and a larger marketing budget. The cost of this part will gradually catch up. Then in the long run, its small team structure may also need to expand gradually. Which of these two possibilities do you think is more likely at present?
Lawrence: From my current observation, Hyperliquid should maintain a relatively small team size. From what we can see from what they have done, such as its HIP-1 spot trading listing mechanism, Binance or other centralized exchanges will have a listing team. From Binance's perspective, they need to pay these people's salaries, and there are also many possible interest issues in the listing process.
Alex: There is a regulatory team.
Lawrence: Yes, and then there needs to be an internal control team. For Hyperliquid, we can see that it solves such a problem in a very simple and fair way. Of course, this is not without problems. It definitely has problems, but from the current perspective, at least it can be seen as a relatively sustainable solution. From the perspective of long-term operation, it saves a lot of costs, including capital costs and team manpower costs. The overall cost reduction is still very large.
I watched some interviews with the founder Jeff, and I feel that he is a person who can solve problems from the perspective of first-line principles. From his initial intention, he did not accept investment. He had a clearer understanding of what he wanted to do, and he did not pay much attention to how the current market participants solved the problem. He was still willing to think about a way to solve these problems from the bottom up. I think this is a core quality that some of the better entrepreneurs in our industry need to have. Someone said before that good entrepreneurs need to have autistic tendencies and childhood trauma. I think that in terms of the first point, it is very reasonable to have this kind of autistic tendency.
Hyperliquid Token Economic Model and Valuation Benchmarks
Alex: Yes, the number of people in the Hyperliquid team and the number of people in the large exchange team remind me of the first principle you mentioned just now, which Musk often talks about. Musk also mentioned another concept before, the so-called idiot index. What is he talking about? It means that when I use the same raw materials, I may have to spend dozens of times the cost to build this thing, such as building a rocket. The ratio of the total cost divided by the raw materials is the idiot index. He believes that only products with a low idiot index are good products. For example, the raw materials may cost ten yuan, and in the end, you may spend 15 yuan to build something. For example, Tesla used batteries from CATL in the early days, because CATL's idiot index is very low, maybe only a few percent. It is only a few tens of percent higher than the cost of raw materials.
From this concept, if the Hyperliquid model is sustainable in the long term, it may mean that the idiot index of most trading platforms in the exchange operating costs may be significantly high. If its logic works, I think many exchanges may have to reflect on themselves, slim down, lower this index, put more energy and funds on truly effective products, and think about how to build my business from the first principles. I think they have made a lot of contributions to the promotion of this industry. Finally, let's talk about the economic model and valuation of Hyperliquid tokens. Because this may involve whether HYPE can be bought at the current price of more than 20 yuan. If it can be bought, where is the reasonable range? Otherwise, please give us a brief introduction to the economic model of HYPE.
Lawrence: Its token distribution is relatively simple. The total amount is 1 billion, and 30% is given to the team and the foundation. The team's part is locked for one year, which is equivalent to being released from December 2025. The biggest pain point of the release is from 27 to 28. 31% of the remaining 70% is the initial creation allocation, which is the airdrop. The remaining 39% is used as an incentive for the community in the future. The AF just mentioned here is its assistance fund. It has repurchased more than 1.1% of the total circulation. At present, more than 31% of the circulation is in circulation, which means that more than 3% of the circulation has been repurchased by its AF. It is equivalent to its actual circulation of about 30%. This is the situation of its chips. Then the economic model has actually been roughly explained just now. At present, all its expenses have not been mentioned by the team. They are all given to HLP and AF, and AF is to repurchase. The amount was just mentioned. By the end of November, the total revenue was close to 100 million. Then the revenue of several blocks in December should add up to more than 10 million. Because I see that the gas of its HIP-1 has increased again in the past two days. It was still relatively low in the first few days of December. Overall, this part should be 4 to 5 million. Add to that its transaction fee income. There are many ways to value it. I think it is more reasonable to compare it with BNB, because its benchmark may be BNB. If we look at it separately, we can only look at it in terms of volume, because there is no way to uniformly consider the specific proportion of the transaction fee and its cost. Simply looking at the volume, its derivatives trading volume is about 10% of Binance, and the spot trading volume is about 1% to 2% of Binance.
Another point to value it is to value its HyperEVM. Based on what we mentioned earlier, we regard the relationship between HyperEVM and its existing exchanges as more of a Binance migration relationship, that is, I don’t think all of Hyperliquid’s current TVL will be migrated to this EVM chain. If we compare the TVL of Binance’s chain and the exchange, and the TVL of Coinbase’s chain and the exchange, the ratio of the TVL of Binance’s chain to the exchange is about 3.7%, and Coinbase is about 6.6%. Of course, Coinbase’s data is because it does not have such open and transparent reserve information on the chain, it is announced in the quarterly report. So Coinbase’s data is based on the data of the third quarterly report, and then it is processed according to Binance’s current increase. Then Coinbase is 6.6%, that is, the TVL of the Base chain is about 6.6% of its trading part of the custodial assets. Of course, Hyperliquid can be considered hot. Even if we estimate it at 10%, its data is roughly equivalent to 3% of the BNB chain. Its current market value is already 10% of Binance, and from these indicators, Hyperliquid is at least not ridiculously expensive. Because we have to consider the economic model of Hyperliquid. You can think that the part given to HLP is its cost, and 100% of the net profit is used to repurchase HYPE. BNB clearly stated earlier that it would use 20% of its net profit to repurchase BNB, but it later changed this rule due to regulatory reasons. It was changed in 2021. Judging from the changes in data before and after the change, at least in 2021, it was still close to a ratio of 20% net profit repurchase.
From the perspective of economic model, first of all, Hype is definitely better than BNB, but the extent of its improvement is not particularly easy to quantify. The overall conclusion is that the current valuation is not high if BNB is used as a reference indicator. But there is a big premise here. Our valuation is definitely calculated based on data. Hyperliquid has the current data, more because Hype has a higher price. The causal relationship here, I think, is that Hype has risen so the data is good, so the valuation indicators look good, rather than saying that because the valuation indicators look good, Hype is very attractive, and then it rises. The causal relationship here is relatively clear. Another point is that Hyperliquid's data this month actually fluctuated greatly. When we valued, the ratio of Hyperliquid to BNB data was 3% or 10%, but its own data was a data that first doubled 10 times and then fell 30% to 40%. Judging from the magnitude of the change in such a data, the validity of the specific numerical indicator is definitely questionable. But it is more likely to use such a framework for estimation, which I personally think is more reasonable. Because we also see many other ways of valuation, which include doing its three business parts separately. For example, referring to GMX and DYDX to value its leading products, and then referring to some DEX to value its spot exchanges.
I personally think the previous method is better. Of course, there is another valuation method, which is to calculate its PE or PS. Because it has a token repurchase and destruction mechanism, and the revenue is equivalent to directly affecting the Hype token, it can also be valued using this indicator. This valuation is based on the indicators we just mentioned to calculate that it has a PS of about 38 times. This valuation is relatively close to AAVE now. Overall, even though its price has fallen a bit now, we use these relatively reflexive indicators to see that its valuation is not very expensive, but it is definitely not cheap.
Alex: I understand. For more valuation algorithms, including valuation angles and business data, you can read Lawrence's upcoming research report on Hyperliquid, which will make the deduction clearer. Last week, there was a hacking incident, which led to the withdrawal of funds from many Hyperliquiquid users, especially some huge investors. How about Lawrence help us introduce this incident again, and roughly what the cause and effect are.
Lawrence: Okay, the reason is that there is a famous white hat hacker named Taylor. The founders of Ronin and Dragonfly have come out to prove that this person is very capable and is a very experienced white hat hacker in events related to North Korean hackers. He came out and said that he followed the addresses of several North Korean hackers to trade on Hyperliquid. At that time, many media should have published his articles. In fact, these hackers were liquidated. Then Taylor published this matter mainly to tell the market that North Korean hackers have set their sights on Hyperliquid, and he actually notified Hyperliquid two weeks ago and made a quotation. But Hyperliquid did not accept his service. The fermentation process of this matter is mainly because Hyperliquid community members do not approve of Taylor's behavior. They do not understand Taylor's background. They think Taylor has bad intentions and think that North Korean hackers just came to speculate on coins and lost money. What is there to care about this matter. Taylor said this was a typical behavior pattern of North Korean hackers before they attacked a protocol. This matter became more and more intense, and then many people we just mentioned came out to prove to Taylor that this was a good person with a good background and ability, and he would not mess around.
After these things happened, a Hyperliquid team member gave a reply on Discord, and the conclusion of the reply was that there was no attack at present. Hyperliquid's security issues do exist. We have talked about this before. One is that all of its current funds come from the arbitrum bridge. Another is that all of its funds are actually in the three-quarters signature just mentioned, which is actually in the hands of this team member. From the perspective of attack and the vector of attack, the main core is actually these two points. Especially the signature of this team. If you want to do something like social engineering attack on these team members, it is not particularly easy to prevent. After all, he only needs to deal with three signatures.
However, many analysts believe that even if the attack occurs, since the TVL of Hyperliquid is basically USDC, many people believe that first, USDC, that is, Circle, will not sit idly by. Second, many people believe that Arbitrum will not sit idly by, because Hyperliquid has contributed 50% of Arbitrum's TVL and more than 60% of USDC's circulation. So if Hyperliquid has a problem, the Arbitrum chain will not sit idly by from the perspective of interests. I think these analyses are also reasonable, that is, its current security is that as a project with a market value of 30 billion, its security issues are still quite vulnerable. However, this problem may not be big when its market value is relatively small and development is the top priority, but with its current market value, it will give everyone some concerns. If this problem cannot be handled properly, or if it is really hacked, it may turn out that Circle or Arbitrum will indeed come to the rescue, but it may be a blow to the team's reputation, including the possible external handling methods that may appear in the whole process. So the concern about funds is still quite reasonable. Just because a hacker didn't attack doesn't mean he will never attack you.
Alex: OK, today we spent a lot of time talking about a single project. I can make a brief summary. I don’t know if I understand it correctly. Lawrence can give us your opinion later. From what I see, the current business model of Hyperliquid is consistent with the business model of most of the leading centralized exchanges. It has an exchange, but it is different from our centralized exchange.
First, its accessibility is permissionless. Second, its data is transparent and verifiable. This is what makes it different from most centralized exchanges. But like most centralized exchanges, it is also a place controlled by multiple signatures. Its biggest room for valuation growth in the future comes from its Layer 1, which has not yet been launched. There are already many potential projects in reserve. The interoperability between its two chains is also worth looking forward to. One challenge it may encounter in the future is that its current market value has reached a relatively high level. The promotion of this market value and the explosion of performance have a lot to do with the fact that the price of Hype was pushed up from a few dollars in the early days through the team's very decisive violent purchase of more than 50 million US dollars.
The current budget in the hands of the team is not as abundant as it was at the beginning, and the market value has reached the current level. The subsequent growth may still face some resistance challenges. There does not seem to be much fermentation or driving incentives in the short term, so the short-term price performance may be under some pressure. Lawrence, do you have anything else to add?
Lawrence: I would like to add that I think it is more reasonable to treat it as a project similar to the centralized exchange business in the long run. In the medium and long term, the core is that its cost is extremely low, which leads to a lot of funds that can be operated in the medium and long term. This fund can subsidize its users and its holders. The team's share of total expenses is surprisingly small at present, and there should be a trend of continuation in the medium and long term. I think this may be a question that Hyperliquid brings to the exchange, the longest-lasting and most solid industry in crypto, in the medium and long term.
Alex: We have seen this advantage repeatedly in many DeFi projects. For example, AAVE may only have a few dozen people running a protocol, which is too few compared to a bank. Hyperliquid has done a good example this time, and compared with large centralized exchanges, such a small number of people can also do a big business. The last question is if the price of this hype falls back in the future, will Lawrence consider buying it?
Lawrence: I will buy. I had some airdrops before, but I sold them at a relatively low price of around 10 yuan. I regret it because I didn’t do a thorough research at the time. I am not optimistic about the short-term, but I am optimistic about the medium and long term. So if there is a pullback, I will consider buying.
Alex: Okay, that’s it for today’s podcast, and next you can look forward to Lawrence’s review report on Hyperliquid.
Welcome to BlockBeats the BlockBeats official community:
Telegram subscription group: https://t.me/theblockbeats
Telegram group: https://t.me/BlockBeats_App
Official Twitter account: https://twitter.com/BlockBeatsAsia
