Back to Markets
Crypto▼ Bearish
L2BEAT: Perp DEX Traders Rely on Operator Trust, Not Crypto Proof
L2BEAT's analysis of Hyperliquid and Lighter reveals most perpetual DEX traders depend on operator trust, not crypto proof. The March 2025 JELLY incident illustrates the risk.
Continue with
Blockchain research firm L2BEAT published a report finding that most traders on decentralized perpetual exchanges such as Hyperliquid and Lighter are effectively trusting platform operators, not the cryptographic proofs the exchanges are built on. The evaluation covered two main criteria: property rights and order fairness. A third criterion, position fairness, was also assessed. On all three, the report concluded that the cryptographic layer does not fully guarantee outcomes without operator intervention.
The report highlighted the March 2025 JELLY incident on Hyperliquid as a concrete example. Validators forced the liquidation and delisting of the JELLY token at a price of $0.0095, preventing estimated losses of $13 million. The episode showed that the platform's validators can override market mechanics, a power that aligns with trust-based systems, L2BEAT researchers said. The JELLY token was delisted shortly after the forced liquidation, and the incident drew criticism from some traders who argued it violated the principles of decentralized finance.
L2BEAT's methodology involved analyzing the smart contract code and governance structures of both platforms. The firm also examined the validator sets. The firm found that while the core trading logic is on-chain, the ability to update contracts or force liquidations rests with a small set of validators or a multisig. On Hyperliquid, the validators are elected by token holders. The report noted that the concentration of voting power gives effective control to a few entities.
Property rights, in L2BEAT's framework, means traders have exclusive control over their positions. Order fairness ensures no participant can see or front-run pending orders. Position fairness prevents liquidations based on selective or delayed data. On Hyperliquid and Lighter, the report found that operators retain the ability to intervene in each of these areas. The JELLY incident was an extreme case. The criteria themselves identify structural vulnerabilities, the researchers said.