KETJU Research

← The Register

other

Hyperliquid HLP

Rejected
Max sleeve
Reviewed
2026-08-19 · v1
Next review
2026-09-19
Research basis
Individual research
Chains
Hyperliquid / HyperEVM · freezable

Rejected venues wait the longest for re-review; a rejection has to earn another look before the scheduled date.

REJECTED ON BOTH ELIGIBILITY AND A TWICE-REALIZED MARKET-MECHANIC LOSS. Hyperliquid’s Terms define persons residing, located, incorporated, or registered in the United States as Restricted Persons barred from its interface. HLP is also not passive protocol revenue: it runs market-making strategies, supplies USDC in Earn, and acts as the backstop liquidator when ordinary order-book liquidation fails. That last role produced roughly $4M of actual loss in March 2025 after a trader forced liquidation of a $271M ETH long; later that month the JELLY manipulation transferred a toxic short to HLP and ended only after validators voted to delist the market and impose a settlement price. A current SEC-filed third-party prospectus reports approximately $13.5M of HLP losses from the JELLY activity. The response added margin tiers and isolated the liquidator strategy, but it also proved validators can alter a market outcome under pressure. A four-day deposit lock further means the client cannot exit as the risk first appears.

The research file

Mechanism and return source

HLP is a protocol-owned strategy vault whose depositors share the PnL of multiple market-making strategies, backstop liquidations, USDC supplied through Earn, and a portion of exchange fees. When a trader falls below maintenance margin, Hyperliquid first sends liquidation orders to the public book. If the account falls below two-thirds of maintenance margin and the book cannot close it successfully, the position moves to HLP’s liquidator component. The vault then owns the position and its mark-to-market risk. This design directs liquidation profits to community LPs in normal conditions, but it also assigns them the tail loss when position size, collateral, oracle marks, or market liquidity defeat the normal closeout path.

Categorical US exclusion

The current Hyperliquid interface Terms, made available by Hyperliquid Corp., define Restricted Persons to include anyone residing in, located in, incorporated in, or with a registered office in the United States, as well as Ontario and sanctioned jurisdictions. The chain may be reachable through other software, but recommending an alternate interface or direct call does not turn a prohibited client into an eligible one. Current SEC filings for proposed HYPE investment products independently reproduce the same terms-based US restriction and explicitly identify VPN circumvention as a potential regulatory and enforcement risk. For this US-advisor mandate, that is a standing zero-allocation decision before the vault’s economics are considered.

Realized losses and validator intervention

On 2025-03-12, a trader opened a roughly $271M ETH long with about $6M of collateral, withdrew margin, and forced the remaining position into liquidation. HLP lost roughly $4M even though Hyperliquid said the system operated as designed; max leverage was then reduced and an added margin requirement applied to transfers. On 2025-03-26, coordinated JELLY positions moved a very large short into HLP before the external spot price was pushed upward. Validators voted to delist JELLY and settle positions at a validator-selected price rather than the prevailing market. A 2026 SEC-filed prospectus reports about $13.5M of HLP losses from that activity. These were market-design losses, not stolen keys, which makes them more—not less—relevant to underwriting HLP.

Control and remediation

The JELLY response contained real remediation: the liquidator vault was capped to a small share of total HLP collateral, automatic deleveraging was introduced when losses breach a threshold, and open-interest limits and margin tiers were made more sensitive to asset capitalization and position size. Yet the event exposed a second control surface. The validator set can coordinate rapidly to delist a market and select a settlement outcome, a power that protected platform solvency while overriding normal price formation. This registry’s separate Hyperliquid Bridge memo also found a 27-validator custom-consensus system with closed-source node software and material foundation-affiliated stake. HLP depositors depend on that same control plane for every trade, liquidation, and emergency response.

Exit and comparison

HLP deposits are locked for four days after the most recent contribution. During that period the depositor cannot withdraw even if a large account is approaching liquidation or validators are debating an intervention. After the lock, redemption remains a claim on vault equity after open-position PnL rather than a fixed one-USDC claim. Compared with Jupiter JLP, HLP has a simpler USDC-denominated interface to a broader strategy set but a materially worse realized tail-loss record; both categorically exclude US users. Compared with holding USDC or Treasury exposure directly, HLP adds leveraged trader counterparty risk, backstop liquidation, validator discretion, oracle marks, and a four-day inability to react. No client-portfolio need requires accepting those added risks through an expressly prohibited venue.

Sources

The claims above trace to these. Where a number could not be independently verified, the thesis says so.

Inherited controls

The verdict above grades the protocol layer. Every position also inherits the asset it holds and the chain it settles on. The least safe layer sets the position’s grade, and the position table names which one that is.

ChainVerdictGradeControl constraint
Hyperliquid / HyperEVMRejected freezable a 21-validator permissioned set operates both the chain and its bridge — one compromise reaches both.
The memo is public. The watching is the product: the terminal reads your clients’ wallets against this Register and flags the events above when they fire. $49 per advisor per month, first 14 days free. Start the trial.