Hyperliquid liquidation incident: Sober reflections after the leverage storm

Key Points

• Research shows that on March 13, 2025, a whale on the Hyperliquid platform liquidated ETH long positions through high-leverage trading, resulting in a loss of approximately $4 million in the HLP Vault.

• The evidence tends to believe that this was a legitimate operation, and the whale triggered liquidation by extracting unrealized profits, circumventing the slippage of market orders.

• This incident has sparked a discussion about the difference in capabilities between DEX and CEX in high-leverage trading.

• Hyperliquid has lowered its leverage limit (BTC to 40x, ETH to 25x) and plans to improve its margin system to prevent similar manipulation.

Hyperliquid Liquidation Event

The whole story and background

On March 13, 2025, a whale liquidation event on the Hyperliquid platform became a hot topic in the DeFi community. Hyperliquid is a decentralized perpetual contract trading platform known for its efficient on-chain order book and high leverage trading (up to 50x). Its HLP Vault is a community-driven liquidity pool used to support liquidation and market making, but this mechanism has exposed its vulnerability in the face of extreme market behavior.

According to Coindesk reports and NFT Evening reports, the incident originated on March 12, when a whale (the on-chain address is often mentioned as "0xf3f4") executed a high-risk trading strategy. The whale deposited about 15.23 million USDC and opened a long position of 160,000 ETH, with a total value of about 307 million US dollars, and leveraged between 13.5 times and 19.2 times (the specific leverage varies slightly due to different sources of data). This position initially showed an unrealized profit of 8 million US dollars, and the whale subsequently withdrew 17.09 million USDC (including principal and profit), significantly reducing the margin level.

As ETH prices fell back (possibly to around $1,915), positions triggered liquidation, Hyperliquid's automatic liquidation system took over, and HLP Vault was forced to take over at a high price and gradually close positions. Due to insufficient market liquidity, HLP lost about $4 million in the liquidation process. This loss accounts for 1% of Hyperliquid's total locked value (TVL) and about 6.6% of the historical cumulative profit (US$60 million).

Giant Whale Strategy and Impact

The strategy of the giant whale is described as "liquidation arbitrage", that is, reducing margin by extracting floating profits, inducing liquidation, and transferring risks to HLP Vault. This operation circumvents the huge slippage of market orders and allows the giant whale to exit a $300 million position at a lower cost. In the end, the whale made a profit of about $1.86 million, while HLP Vault took a loss of $4 million.

The incident triggered a net outflow of $166 million, and the price of HYPE tokens fell from $14 to $12.84, reflecting the market's shaken trust in the platform. ETH prices were also affected by liquidation selling pressure, falling to around $1,915, ending the previous small rebound.

Interestingly, the same whale continued to be active after the incident, opening a new 25x leveraged ETH long and 40x leveraged BTC short on March 13, showing that it has not completely exited the market, which may further test the platform's risk management capabilities.

Comparison of leverage management between DEX and CEX

This incident has sparked a discussion about the difference in the capabilities of DEX and CEX in high-leverage trading. Some people believe that this is a "trading turmoil and a crisis of trust" and question the sustainability of DEX's high leverage. Others believe that this is a "liquidity game" within the rules and there is no need to over-interpret it. CEX can adjust the leverage limit based on the size of the position through centralized risk management (such as dynamic risk limit mechanism). For example, in CEX, a position of this size may be limited to 1.5 times leverage. However, DEXs such as Hyperliquid lack similar controls due to their decentralized nature, and users can diversify risks through multiple KYC-free accounts to circumvent restrictions.

In theory, the advantages of CEX include market monitoring tools and open interest (OI) limits to detect and prevent market manipulation. The challenge facing DEX is to maintain the principle of decentralization while preventing abuse. For example, Hyperliquid cannot freeze accounts or track user identities like CEX, which increases the risk of manipulation.

Hyperliquid's response

Hyperliquid officially responded afterwards, denying that the incident was a vulnerability or hacker attack, emphasizing that it was the result of high-leverage trading. Based on the information we collected, the platform has taken the following measures:

• Reduce leverage limit: Reduce the maximum leverage of BTC to 40 times and ETH to 25 times to reduce the risk of large position liquidation.

• Improve margin system: Plan to introduce a new margin design to ensure that liquidated positions are loss-making for traders (relative to the entry price or the last margin transfer). For example, at 20x leverage, liquidating a position requires a loss of at least 18.3%, making manipulation uneconomical.

In addition, Hyperliquid emphasizes that with the addition of market makers, platform liquidity will increase and the cost of price manipulation will increase significantly. According to Hyperliquid's Medium article, this strategy will attract more professional participants and enhance the robustness of the system.

Event Data Overview

The following is a summary of the key data of the event:

Community and Block unicorn's views

The community's views on the incident are divided. Some people concluded that there needs to be a centralized force to detect and limit malicious behavior. This completely violates the purpose of decentralized finance (DeFi) and everything Hyperliquid stands for. This forces users back to the Web2 world where the platform has the final say. True decentralized finance is worth it, even if it is 10x harder to build. Just a few years ago, no one believed that DEX/CEX volumes would reach today's ratios. Hyperliquid is leading the way and has no intention of stopping.

Others believe that approaches copied from CEXs will work in DeFi as well. The most common suggestion I have seen is that as position size increases, the margin requirement ratio per address should also be adjusted accordingly, because CEXs only offer higher leverage for smaller positions. However, this approach does not prevent manipulation attempts on DEXs, as a sophisticated attacker can easily open positions on multiple accounts. Nonetheless, this will reduce the impact of "organic whale" positions to some extent and is one of the features to be implemented.

Another suggestion is to implement some features that severely limit the usability of the platform in exchange for security. For example, if unrealized profit and loss (PNL) is not extractable, many attacks cannot be carried out. In fact, Hyperliquid has pioneered segregated-only perpetual contracts for illiquid assets that have this safety mechanism. However, this change will have a serious impact on capital arbitrage strategies, as unrealized profit and loss (PNL) will need to be extracted from Hyperliquid to offset losses in other venues. In system design, real user needs are the primary consideration.

There are also proposals to innovate the design through margin settings based on global parameters. However, the liquidation price needs to be a deterministic function of price and position size. If global parameters such as open interest are used as inputs to margin requirements, users will lose confidence in using leverage.

So what is the answer? We all want DeFi, but permissionless systems must be resistant to manipulation at all scales.

The answer lies in understanding the real problem with large positions: they are difficult to mark. The first-order approximation of mark price multiplied by size breaks down when the market impact approaches the maintenance margin. Accurately simulating market impact is impossible because order book liquidity is a path-dependent function of time and the behavior of other participants. Without simulating market impact, liquidation can become a low-slippage exit method, but the price is unfavorable to the liquidator.

Hyperliquid's margin system update therefore has the following ideal properties: any liquidated position is either a loss relative to the entry price, or at least a loss of (20% - 2 * maintenance margin ratio / 3) = 18.3% relative to the last margin transfer (using 20x leverage as an example). An organic 20x leverage user who earns a 100% equity return after a 5% move is still able to extract most of the profit and loss without closing the position. However, by introducing a separate margin requirement between transfers and opening new positions, a profitable manipulation attempt would require moving the mark price by nearly 20%. From a capital perspective, such an attack is not feasible.

I would like to point out that as market makers continue to scale on Hyperliquid, the mark price issue will resolve itself. So it is likely that traders yesterday lost money overall. The $1.8 million P&L on Hyperliquid could have been made up when pushing prices higher on other venues or using other accounts on Hyperliquid. HLP took a bad position and lost $4 million. The only market participants who are certain to make money overall are market makers, who are able to make millions of dollars in P&L in a matter of minutes. As liquidity improves, it will become increasingly expensive to change prices. So while improvements to the margin system will be of great help, the appeal of easy P&L to market makers will provide an independent source of robustness over time.

Finally, Hyperliquid's response reflects its adherence to the principles of decentralization while reducing the risk of manipulation through technical innovations such as improved margin systems. However, given the openness of DEX, it may be necessary to explore more CEX-level risk management tools in the future, although this may trigger controversy in the community over privacy and decentralization. This incident provides a valuable lesson for the DeFi ecosystem, emphasizing the potential risks of high-leverage transactions and the importance of platform governance.