ChainCatcher Space Review: Behind Hyperliquid’s game with whales, where is the future path for on-chain trading?

Looking back at the amazing record of this 50x leverage whale on Hyperliquid: five games and five wins, making $15 million in 10 days. He escaped by "actively compressing liquidation prices", but caused the HLP vault to lose $4 million in 24 hours. This game between giant whale and the platform has sparked heated discussion: Why is his strategy so accurate? Will the HLP mechanism drag down the ecosystem? Will the HLP problem repeat itself in CEX?

This week, ChainCatcher invited six front-line VCs and researchers to discuss investment challenges and opportunities in turbulent markets with the theme of "Behind the game between Hyperliquid and whales, where is the future path of on-chain trading?"

For details, see X:

https://x.com/i/spaces/1zqKVjgRLvdKB

The following content is a summary of Sapce Chinese.

1. Host Ray: Hyperliquid 's 50-fold leverage giant whale made the HLP vault lose $4 million in 24 hours. What problems have been exposed when the on-chain exchanges were "swallowed"?

Sean: Platforms like HyperLiquid do expose some structural problems when facing high leverage funds entering the market. Especially in the bear market, the amount of funds has declined and the platform is more susceptible to manipulation. The key to the problem reflected in this incident lies in how to improve its mechanism.

The response given by the platform is to lower the maximum leverage multiple of BTC and ETH, with BTC reduced from 50 times to 40 times, and ETH reduced from 50 times to 25 times. But this does not fundamentally prevent similar events from happening again. Because the attacker can still operate in batches through multiple addresses to launch an attack on the HLP's fund pool. At this stage, the platform mechanism is still unable to effectively defend against such premeditated operations.

Lucio: Judging from this incident, the whale established a large number of long positions through 50 times leverage, and used a smaller principal to control huge positions to guide the market trend. When the position is floating, he withdraws the profit part of the margin, and even if the position is liquidated, it will not cause actual losses. Finally, the platform took over through forced weight, resulting in the vault bearing liquidation losses.

The main problems are three points: First, the allowable leverage multiple is too high, second, the position limit is too low, and third, the liquidation price setting is too loose. If the user who sets the platform to open 50 times leverage triggers a strong flattening at a 0.5% drop, the risk will be greatly reduced. Overall, this is a loophole in the platform's product design and risk control mechanism.

Jarseed: I followed the whale's liquidation process throughout this incident. During the observation, it was found that even during the period of partial liquidation, the account was still profitable.

Looking back at similar cases in centralized exchanges, such as SBF that helped CZ clear large orders, there is also a background of such high leverage operations. I think the key problem reflected by the current event is still the design of leverage and margin mechanisms. For example, the platform's openness to leverage, the speed of response of the liquidation process, etc., still have room for optimization. There is not much problem with the overall mechanism, but the details of risk control need to be strengthened.

Yuyue: I agree with the previous point of view. Due to the lack of a KYC mechanism on-chain transaction, the platform cannot block the address, which is essentially different from centralized transactions and limits the risk control ability. HyperLiquid's mechanism is currently not suitable for carrying large amounts of funds. Especially under the premise that data is completely transparent, it is easy to be reversely manipulated by larger funds, amplifying slippage and liquidation risks.

Similar to Texas Hold'em, if the opponent's capital volume far exceeds you and you can still see your trump card, the game will be extremely unbalanced. There is a lack of protection mechanism on the chain, and strategic large players can easily suppress small and medium-sized players and even platforms.

In contrast, centralized exchanges can risk control by hiding key transaction information to prevent malicious attacks. This is also the reason why I lowered my expectations for HyperLiquid, which is structurally difficult to support higher volume trading activities.

In the future, some "hybrid" protection mechanisms may be explored, such as by hiding some addresses or behaviors, but this will also bring about transparency and compliance challenges.

Danny: This incident revealed controllable and uncontrollable problems in the platform mechanism.

In terms of controllability, the platform has not set enough risk exposure upper limit and the hedging mechanism is not perfect. Platforms should avoid systemic risks by limiting the scale of leverage and dynamic margin.

In terms of uncontrollability, the openness of on-chain transactions and the difficulty of identifying abnormal behaviors brought about by the lack of KYC are challenges faced by all decentralized platforms. If this issue is not handled properly in the future, it may affect compliance and large-scale user acceptance.

Overall, this is the testing phase that HyperLiquid must go through as a decentralized platform during the bull and bear transition period.

Jt Song: I think this incident reflects from a more macro perspective that the current on-chain products are still in their early stages. The biggest contradiction lies in the balance between decentralized concepts and performance, safety, and risk control.

Centralized exchanges can suspend trading or modify parameters in times of crisis, while on-chain products are limited by contract mechanisms, and their response speed and flexibility are not as good as centralized platforms. Once a platform is attacked or arbitrage, it is difficult to stop losses quickly.

In addition, excessive transparency has become a double-edged sword. When trading volume is still mainly concentrated on centralized exchanges, large funds on the chain are easy to manipulate prices. If the chain becomes mainstream in the future and an ecosystem of "transparency vs. transparency" is formed, this type of problem may be alleviated. At this stage, whales' advantages are still obvious.

2. Host Ray: Is the current game relationship between giant whales and platforms zero-sum games or can they promote development with each other? Can the community restrain whale behavior through collective action?

Sean: We can review the stage when centralized exchanges began to place contracts from 2018 to 2019. At that time, what many platforms were doing was actually what HyperLiquid wanted to achieve now.

In theory, if CEX also has its own bot or HLP mechanism like it, its APY should be higher than HyperLiquid, because a large amount of smart money and whale funds are gathered on the HyperLiquid platform.

The relationship between a whale and a platform is not necessarily an opposition. For whales, HyperLiquid is one of the few places that allow maximum leverage and maximum capital exposure; for the platform, whales place large orders on the chain will attract a lot of attention and form a traffic effect.

But for ordinary HLP users, the existence of whales may compress their earnings expectations and even constitute damage. Whether it is a win-win situation depends on whether the platform mechanism can be optimized. I suggest that mechanisms can be designed from the perspective of market impact, such as setting the upper limit of the impact of orders on market prices, and looking for decentralized and permissionless solutions. This may be a healthier win-win path.

Danny: I think the whale's operation behavior is essentially the same as some people who call orders on Twitter or trade on the chain. They attract market attention through trading, drive price fluctuations, and at the same time hedge their own risks through various means to ensure that they do not lose money.

This type of behavior must be profitable for whales; for the platform, whales' active behavior can attract users and topics and form positive traffic. So at this stage, this is a win-win situation. Platforms, whales and users all gain different levels of value from it.

Jarseed: I agree with the previous ones. The whale's ordering behavior itself triggers tracking from other users, building a market around its trading behavior.

This incident is more about whales discovering loopholes in the platform mechanism and achieving low-risk arbitrage. He has extremely strong hedging ability, and may even be linked to multiple platforms.

From an ecological perspective, whales' activity is conducive to improving the activity and exposure of the on-chain platform. But if whales cause persistent HLP losses and affect the platform's existence, the game itself will end. Therefore, the platform mechanism design still needs to be improved.

3. Host Ray: Some communities now propose "whale hunting operations", hoping to counter whales through collective hedging. Is this possible an effective constraint mechanism?

Jt Song: I think this whale hunting action may become part of the whale strategy. For example, suppose I have a $100 million position on the exchange, and I open a $5 million short position on HyperLiquid, inducing the community to snipe me and increase the price by 5%.

On the surface, the community has won, but in fact, I have made a profit on larger positions on CEX far exceeding the losses on the chain. This combination of "open conspiracy + conspiracy" can easily maximize benefits.

Therefore, the whale hunting operation may backfire, and instead helps the whale complete off-chain shipment, it is a double-edged sword.

Lucio: I think the relationship between whales and platforms is ultimately a game between users and systems. When there are loopholes on the chain, users will naturally exploit them, which is not illegal or illegal.

Similar situations have also happened in GMX. For example, at the end of 2023, YFI coins doubled in a short period of time. Some whales opened long positions on GMX, and coordinated the market with off-market funds, which eventually led to large losses on the platform.

So the question is not whether the user does evil, but whether the platform design is stable enough. If the platform mechanism can withstand such operations, even whales will not harm the system. As for whale hunting, I don't think it's realistic. Community members have unequal cognition and cannot be unified in their strategies, and they are more likely to be reversely exploited by whales.

4. Host Ray: Mainstream DEX Revenue Decline: What is the root cause and how to break through in the future? Does CEX have mature experience for reference?

Jt Song: I think there are two directions worth paying attention to. The first is to improve the performance and response speed of the chain, so that DEX's product functions are stronger and more stable, thereby enhancing sustainability.

The second is to combine smart contracts and AI models to achieve smarter risk control. For example, when an abnormal market or attack occurs, the system can automatically adjust the leverage multiple, or identify malicious transactions that use rules to arbitrage to intervene.

In the future, we can consider using AI to customize user behavior and provide personalized risk parameter settings. Such a platform system is closer to the risk control efficiency of centralized exchanges, while retaining the decentralized characteristics.

Sean: I'll share with you the reasons for the decline in revenues of mainstream DEX. Uniswap has tried to increase revenue by expanding to more chains in the past, but the actual results are not good. Multi-chain strategy leads to increased maintenance costs and dispersed transaction volumes, which does not bring higher returns.

In addition, a number of products with "destructive innovation" have emerged, such as GMGN, Phantom, Pepeboost, etc., which have directly changed the user interaction model and transaction logic, and have also diverted the traffic and revenue of mainstream DEXs such as Uniswap.

However, from the perspective of somatosensory, the overall transaction revenue on the chain in this cycle should be much higher than the previous cycle, especially at the levels of on-chain lending, bots, contract DEXs, etc. Therefore, not all DEXs are declining, but the revenue differentiation of different types of products is obvious.

The key to breaking through the future is to innovate product design, rather than just imitating CEX. Especially in the current geopolitical and regulatory context, CEX itself is also facing increasing restrictions, and DEX has the opportunity to overtake in some areas.

Danny: The core of DEX’s reference from CEX is its anti-cycle ability. CEX has spot business in bull markets and contract business in bear markets, which can smooth revenue. Most DEXs have no trading volume in bear markets and have poor anti-cyclical ability.

The solution is through business integration. For example, package gambling products or volatility products and aggregators on the contract chain to form a more complete trading ecosystem. Currently, on-chain projects like Manta and Merlin are trying to introduce game-based transactions into DEX.

The second point is capitalization. If DEX can successfully raise funds at the market high point, integrate projects within the ecosystem, and expand revenue sources, it will have a better chance of resisting the pressure of the bear market and even surpassing some centralized second-tier exchanges.

For example, Jupiter continued to raise funds at the peak of its business and looked for ecological projects for acquisition and integration, gradually expanding its influence.

Finally, in response to the data mentioned by Teacher Sean, according to my observation, Uniswap's trading volume is indeed lower than the previous cycle, but the trading volume of platforms such as GMX, Jupiter, dYdX has been significantly higher than the previous cycle. The overall growth trend of DEX is clear.

Jarseed: I'll add to the evolution path of current DEX from the perspective of product classification. It can be roughly divided into three categories:

The first category is the lowest basic DEX, such as Uniswap and Raydium, which provide original liquidity;

The second category is aggregators, such as 1inch and Jupiter, which integrate multiple sources of DEX liquidity;

The third category is trading products for C-end users, such as Trading Bot, Pepeboost, and GMGN, which serve end users more directly, emphasizing ease of use and participatory ability.

One of the core driving forces of this round of asset outbreak is the emergence of a new issuance mechanism. For example, on Solana, platforms like Pump.fun have greatly lowered the threshold for currency issuance, using light liquidity and rapid rise mechanisms to drive on-chain activity.

In its early stages, CEX also acquired traffic by actively launching high-volatility assets. For example, Binance creates wealth effects and gains user trust by launching waist-level currencies. Nowadays, the asset price discovery process is gradually moving forward on the chain.

Whoever can gain an advantage on the asset side in the future will be able to dominate the competitive landscape of DEX.

5. Host Ray: You can speak freely and predict the next breakthrough point of on-chain transactions from the innovative models, potential trends, and narrative explosion paths you have observed. This breakthrough may be reflected in product form, asset form, chain structure, or new logic of a new chain.

Jarseed: I'll tell you one direction. Currently, when rushing small coins on the chain, many scenarios can only do unilateral trading (i.e., buying) and cannot short it. If a certain mechanism can appear in the future, allowing users to participate in short selling in the early stages of new assets, such as building short positions through borrowing coins or contracts, this will greatly enrich the game structure.

Of course, such a mechanism requires extremely high risk control capabilities, because the prices of new assets fluctuate violently and have extremely high risks. But if anyone can do a solid job in risk control, this type of product is both market attractive and has huge profit potential. I think this is one of the important evolutionary directions for on-chain trading.

Sean: Several independent narratives that I have recently paid attention to are focused on the "specialized chain". We have invested in Berachain and are optimistic about projects such as Celestia. What they have in common is that they focus on solving structural problems in specific scenarios.

For example, HyperLiquid focuses on trading; Berachain focuses on stability construction of liquidity proofs; Celestia provides modular block space support. The logic of these projects is: Ethereum and Solana cannot fully meet all scenarios, so in the future, there will be chains with independent business structures to serve specific needs.

Whether an independent narrative can explode must meet several conditions: a scenario with strong demand pain points, an application product that can actually be run, and technically, it is not required to license, self-host and trustworthy in the contract. If all these conditions are met, it is not just an L1, but a decentralized platform with independent business structure.

Jt Song: Our 0G chain is also working hard in this direction at present. As a decentralized AI chain, we believe that future asset issuance must be combined with AI and upgraded based on the original logic.

One of our focus is to launch the "intelligent NFT" standard, which directly binds the training data or small models of AI models to NFT. Users not only hold assets, but also customize and continuously train this AI model, and even connect it to the Twitter API to automate social behavior. Under this model, AI not only participates in asset construction, but also becomes directly part of the asset.

In addition, our main network is benchmarked against AWS at the storage layer, is competitive in terms of expenses, and achieves a balance between on-chain transparency and high-performance computing. We hope to promote the on-chainization of AI data and execution processes, improve the traceability and security of the entire industry, and solve problems such as "AI illusion".

Lucio: My perspective is more executable. I'm not very good at predicting what the next narrative will be, but I'm better at identifying trends from the data. Once a chain or a certain product begins to have a clear preference in terms of user growth, popularity, data performance, etc., I will pay attention and support it immediately.

The directions mentioned just now are very inspiring. Whether it is specialized chains, AI assets, or more professional on-chain contract platforms, as long as the data is good enough and the user behavior is real, it is possible to explode.

For our organization, we are also willing to support such teams to promote their development from the perspectives of liquidity, resources, markets, etc.