During the October 2025 crypto crash, Hyperliquid's backstop mechanism absorbed $576 million in forced liquidations off its public order book, helping to prevent a deeper systemic selloff and highlighting the impact of venue-level risk controls
When crypto markets plunged in October 2025, Hyperliquid, an on-chain perpetual futures platform, faced a wave of forced liquidations. According to a new research preprint, the venue's internal backstop mechanism absorbed the majority of these forced sales, preventing them from hitting the public order book and potentially triggering a deeper cascade of liquidations.
Backstop Absorbs the Shock
At 21:19 UTC on October 10, 2025, Hyperliquid processed approximately $641 million in forced sales within a single minute. Of that, about $576 million was routed to the platform's backstop, while only $64 million reached the public order book. This split is significant: when order books thin out during high volatility, prices can spiral lower, forcing more leveraged positions to liquidate. By absorbing most forced sales internally, Hyperliquid's backstop helped interrupt this feedback loop, reducing the risk of a self-reinforcing crash within the venue.
How the Liquidation Mechanism Works
Hyperliquid's liquidation process first attempts to close positions using market orders. If certain conditions are met-such as insufficient liquidity or extreme volatility-a liquidator vault, part of the Hyperliquidity Provider (HLP) protocol, can take over the position instead. The research found that after the initial onset of forced selling, 62.6% of the value was absorbed off-book by the backstop. The event was highly compressed in time: 87.8% of forced sales after onset occurred within 30 minutes, and 96.5% within an hour.
Systemic Risk and Market Feedback
The study modeled the liquidation cascade using a branching ratio, which measures how many additional liquidations are triggered by each forced sale. A ratio near 1 would indicate a self-sustaining chain reaction. Hyperliquid's branching ratio remained below 0.2 in all measured regimes, peaking at 0.195 during the initial phase and dropping to 0.140 at the event's peak. This suggests the backstop mechanism dampened internal feedback, though the effect was limited to Hyperliquid itself-shared prices across exchanges could still amplify liquidations market-wide.
Research Context and Limitations
The research places Hyperliquid's October 2025 event in the context of seven major Bitcoin perpetual futures liquidation cascades from 2022 to 2025. Its earlier companion study found no universal early-warning indicator for such events. The October 2025 crash is the only in-flight case study available for Hyperliquid, as its fill-log archive began in May 2025. The authors suggest that venues without a similar backstop may experience higher realized branching ratios, a hypothesis for future cross-venue analysis.
During the 15.7-hour window following the onset, the paper tracked $733 million in book-directed forced sales, with $644 million occurring during the initial nucleation phase. The 62.6% backstop share is reported as a separate off-book series, so the figures describe different aspects of the liquidation process rather than a single combined total.
Hyperliquid's approach to managing forced liquidations stands in contrast to other market responses to liquidity stress. For example, some exchanges have responded to shrinking reserves and compliance pressures by selling assets or proposing structural changes, as seen when Fold sold Bitcoin and considered a reverse split to address listing requirements-an episode covered in detail in EgonCoin's reporting on Fold's treasury management.
According to the research preprint, the branching ratio for Hyperliquid's liquidation cascade peaked at 0.195 during the nucleation phase and dropped to 0.140 at the event's peak, with a separate amplification calculation implying a ratio of 0.122. The backstop absorbed 62.6% of forced-sale value off-book after onset, and 87.8% of forced selling after onset occurred within 30 minutes, with 96.5% within one hour. The study tracked $733 million in book-directed forced-sale value over a 15.7-hour window, including $644 million during the initial phase.
Backstop mechanisms like Hyperliquid's are designed to absorb forced liquidations internally, reducing the risk of a feedback loop that can drive prices sharply lower during periods of extreme volatility. While these systems can help stabilize trading venues during market stress, their effectiveness depends on the specific design and liquidity of each platform. The broader market impact remains subject to cross-exchange dynamics, as price movements on one venue can still influence liquidations elsewhere. As more exchanges and protocols experiment with internal risk controls, understanding the trade-offs and limitations of these mechanisms will be critical for both users and market operators.
