Oracle error on Hyperliquid perp market wiped $1.5M in May
DeFi & Yields ·
A stock-split misread by a price oracle triggered a 45% crash in a pre-IPO perp market in under 30 minutes, liquidating 400 traders, underscoring data-layer risks as pre-IPO perp trading expands.
Pre-IPO perpetual futures markets on Hyperliquid have been expanding, but a May incident involving $SPCX highlighted how fragile these markets can be when price feeds are misconfigured. One Hyperliquid perp market tracking the pre-IPO asset fell 45% in under 30 minutes after an oracle misinterpreted a routine stock split as an actual price collapse, according to a post from RedStone. The misread triggered cascading liquidations across the market.
The scale of the fallout was concentrated but significant: 400 traders were liquidated and $1.5M in notional value was erased in the span of half an hour. The event illustrates a structural risk in Hyperliquid's HIP-3 framework, which allows anyone to launch a perp market on nearly any asset while leaving the choice of oracle to the market's deployer. That flexibility means a single wrong oracle configuration can destabilize an entire market almost instantly.
The core issue was mechanical rather than market-driven — the oracle treated a routine corporate action, a stock split, as if it reflected a genuine price crash. For assets that can be split, repriced by funding rounds, or otherwise restructured outside normal trading, this kind of misclassification is a recurring hazard for any real-world-asset market built on perpetual futures.
RedStone has positioned its oracle stack as a response to this specific failure mode, powering HIP-3 markets with a primary and fallback price system. The design separates live price feeds from mechanical resets like stock splits, aiming to prevent an oracle from conflating a structural adjustment with an actual collapse in asset value.
The episode has been documented in a research spotlight examining how HIP-3 pre-IPO markets handle pricing, with the incident cited as a case study in the risks of deployer-chosen oracle configurations. What remains unresolved is how widely such misconfigurations may recur across other HIP-3 markets tracking pre-IPO or split-prone assets, and whether broader adoption of primary/fallback oracle designs will become standard practice for deployers launching similar markets going forward.