Cover photo

Liminal’s breakdown: October 10 market event.

TL;DR

On October 10, the crypto markets experienced an unprecedented liquidation cascade that wiped out over $19 billion within hours.

Hyperliquid remained fully operational with zero bad debt, while several major exchanges experienced downtime.

Hyperliquid’s Auto-Deleveraging (ADL) mechanism was triggered for the first time at this scale in over two years to maintain system-wide solvency.

As Liminal is running delta-neutral strategies (spot + short perp), some users with highly profitable shorts were closed via ADL to offset market-wide losses.

Following these closures, Liminal converted all spot assets (of positions affected by ADL) to USDC to prevent further directional exposure, protect user balances, and shield them from deeply negative funding rates observed after the crash.

Withdrawals have remained fully open throughout the event, deposits have reopened, and executions are progressively resuming.

Users running customized strategies must confirm their strategy in the app before reactivation, while Regular users within Liminal’s strategy do not need to take any action.

Overall performance: +$355,354 net realized PnL at the protocol level, with 75% of affected users in profit and an average return of +1.8% per account.

Among users running the Liminal strategy and affected by ADL, 1,800 users (99.94%) ended in profit. The average return per affected user was +3.43%, with a median of +3.40%.

Some users recorded losses as the scale and speed of the event led Hyperliquid’s ADL to forcibly close highly profitable short positions from Liminal users at uneven times and prices in order to offset market-wide losses and preserve Hyperliquid’s solvency.

Liminal deliberately chose to keep withdrawals open rather than mutualize user outcomes, ensuring fairness and preserving each account’s autonomy across both custody modes.

We remain mindful of users who were negatively impacted during the event and are exploring ways to reflect their experience in future community-focused initiatives.


Market context

On October 10, the crypto market faced its most severe stress event to date, a full-scale liquidation cascade triggered by a sudden macro shock that wiped out more than $19 billion in positions within hours, according to Coinglass.

What began as ordinary volatility quickly evolved into a global wave of forced liquidations as traders, funds, and automated systems were compelled to unwind positions across major exchanges. Liquidity evaporated, prices diverged sharply across venues, and emergency mechanisms such as Auto-Deleveraging and backstop programs were activated throughout the industry to contain the damage.

Quantifying the collapse

The scale of the October 10 unwind has no historical precedent in crypto. Within minutes, the largest assets on Hyperliquid suffered double-digit losses:

  • BTC: ≈−15%

  • ETH: ≈−20%

  • HYPE: ≈−26%

  • SOL: ≈−33%

  • XPL: ≈−63%

  • PUMP: ≈−74%

  • FART: ≈−85%

These moves, occurring within an hour, erased a significant share of global open interest and triggered liquidations across every major exchange. More than 1.6 million traders were liquidated in a single day, the largest mass liquidation event in industry history. The total $19 billion in wiped-out positions far exceeded the scale of any prior crash.

Even stablecoins temporarily deviated from normal behavior. USDT and USDC traded at a slight premium on certain exchanges as traders rushed to de-risk, reflecting an extreme demand for stability where even “cash” became scarce.

Beyond price action, market infrastructure itself came under pressure. Major platforms, such as Binance, Lighter, or Kraken experienced outages or severe latency, leaving thousands of users unable to manage positions. Hyperliquid, by contrast, remained fully operational, processing record volumes without interruption, a rare display of resilience during widespread system stress.


Hyperliquid’s performance during the event

During the October 10 liquidation cascade, Hyperliquid processed billions of dollars in liquidations within minutes while maintaining 100% uptime and zero bad debt, a rare outcome under such extreme volatility. The event served as a full-scale stress test for the Hyperliquid’s execution, risk, and solvency systems, confirming their ability to perform under exceptional market pressure.

As markets unraveled, Hyperliquid’s engine executed liquidations at record speed and scale. All positions were settled transparently, and the platform’s consensus layer, HyperBFT, sustained peak throughput without latency or disruption.

For the first time in more than two years, Auto-Deleveraging (ADL) was triggered at scale on most perp markets available on Hyperliquid simultaneously to preserve system solvency. ADL activates only when both market and backstop liquidations are exhausted, closing a portion of profitable positions on the opposite side to offset undercollateralized losses. Its function is protective for the solvency of Hyperliquid, ensuring that no bad debt can accumulate within the system.

Despite more than a thousand traders losing over $100,000 and more than 200 exceeding $1 million in losses, the platform itself remained solvent, transparent, and operational. Hyperliquid’s handling of the event demonstrated the robustness of its margining, liquidation, and execution layers, reaffirming its position as one of the few trading systems capable of maintaining integrity under systemic stress.


Auto-Deleveraging (ADL) on Hyperliquid

Auto-Deleveraging, or ADL, is the final safeguard that ensures the solvency of every derivatives exchange. It activates only after all other liquidation mechanisms, including market and backstop liquidations, have been exhausted and the system still contains accounts with negative value. Its function is to maintain solvency under all conditions and prevent bad debt accumulation.

When a user’s account or isolated position turns negative and the market lacks sufficient liquidity to close it through standard liquidations, the system identifies traders on the opposite side who are in profit. These traders are ranked using a transparent index:

(mark price / entry price) × (notional position / account value)

This formula captures both profitability and leverage, ensuring that the most profitable and highly leveraged traders are prioritized for closure. This process offsets losses from underwater positions while maintaining overall market balance.

If liquidity becomes insufficient, the system intervenes directly, closing profitable traders’ positions based on the same index. This controlled reduction restores equilibrium and ensures full solvency. Although this mechanism may appear unfair to affected traders, ADL is a necessary element of perpetual futures markets, where every gain mirrors an equivalent loss. Its role is to preserve systemic balance and prevent solvent users from absorbing others’ losses.

The October 10 event was an extraordinary and rare occurrence in Hyperliquid’s history. For the first time in over two years, Auto-Deleveraging (ADL) was triggered at scale across most perp markets simultaneously to preserve system-wide solvency. ADL activates only after all market and backstop liquidations have been exhausted, closing a portion of profitable positions on the opposite side to offset undercollateralized losses. Its function is protective for the solvency of Hyperliquid, ensuring that no bad debt can accumulate within the system.

The magnitude and speed of the market dislocation produced highly asymmetric outcomes across users. Some benefited significantly from favorable position timing, while others experienced sharp drawdowns as liquidity thinned and prices diverged rapidly across assets.

On Hyperliquid, ADL is triggered only after the Hyperliquid Liquidity Provider (HLP) vault, acting as a non-toxic backstop liquidator, has fully absorbed all possible liquidations. If HLP itself becomes undercollateralized, it is treated identically to any user in the ADL queue. This guarantees procedural fairness and complete transparency.

Each ADL event involves two sides: the triggered side, consisting of undercollateralized accounts, and the providing side, made up of profitable traders ranked by the ADL index. While ADL-providing trades are generally profitable overall, some can be unfavorable when only parts of a diversified portfolio are closed.

In essence, Hyperliquid’s ADL mechanism is protective for the solvency of the platform and its users’ positions and assets. It acts as a controlled rebalancing process that preserves solvency, transparency, and systemic integrity. The events of October 10 serve as a rare but valuable reminder of its importance, proving that even under extreme conditions, the system can remain fully solvent, operational, and fair to all participants.


How Liminal operates on HyperCore

Liminal is built to provide sustainable, market-neutral yield through fully automated delta-neutral strategies. Its purpose is to make basis strategies accessible to all users.

At its foundation, Liminal runs a delta-neutral engine designed to neutralize directional exposure. When a user deposits USDC, the system opens two offsetting positions on Hyperliquid: a long position in the spot market and a short position of equivalent size in the perpetual futures market. The result is a total delta of zero, meaning that price movements in the underlying asset have minimal effect on overall portfolio value. This structure allows the strategy to focus on capturing yield from structural inefficiencies such as funding rates.

Liminal operates directly on Hyperliquid’s execution layer, HyperCore, which provides access to deep liquidity, real-time data, and low-latency execution. These characteristics are essential to running a delta-neutral system efficiently. During major market events, however, when liquidity deteriorates or spreads widen, this same infrastructure can also transmit market stress into strategy performance.

A defining aspect of Liminal’s current product design is the segregation of user accounts. Each user’s strategy operates independently, preventing any imbalance or loss from spreading between participants.

This isolation enhances transparency and safety, although it also means that results can differ for each user depending on timing, asset allocation, leverage, and other parameters. This design favors clarity and capital protection, even if it can produce uneven outcomes under extreme conditions.

Liminal offers two operational modes combining flexibility, security, and control.

In Regular Mode, Liminal manages an externally owned account for the user, executing all transactions within an offline enclave secured by hardware-level cryptography.

In Institutional Mode (Self-Custodied), funds remain entirely under the user’s control within their own Hyperliquid sub-account. Liminal interacts with these accounts through limited agent permissions that cannot withdraw or transfer funds, allowing users to revoke access at any time and continue managing their portfolios directly if desired. Even during market disruptions, this configuration ensures uninterrupted access to assets.

Every position, trade, and performance metric is verifiable on-chain through Hyperliquid’s infrastructure. While the system cannot entirely eliminate market risk, it minimizes it through automation and segregation.

The recent market events have reinforced that even the strongest architectures can face extraordinary stress. Liminal’s framework, grounded in transparency and user control, proved resilient but not immune to extreme volatility.

Risk management and market exposure

Liminal’s risk management framework relies on automation, segregated architecture, and precisely defined exposure limits. These mechanisms substantially reduce risk, though no system can completely eliminate it, particularly under extreme or disorderly market conditions.

Market neutrality minimizes exposure to price direction, yet periods of high volatility or sudden funding reversals can still affect performance. Rapid movements in spreads between spot and perpetual markets may require rebalancing at less favorable prices, leading to temporary drawdowns.

Liquidity risk can also emerge during market stress when order book depth decreases and spreads widen sharply. In such environments, execution becomes more challenging, especially for smaller or less liquid assets. Execution risk can arise when volatility accelerates between the trade signal and order completion, but continuous monitoring and adaptive rebalancing help to mitigate this effect.

During the October 10 market event, Hyperliquid operated as intended, maintaining full uptime and zero bad debt despite extreme volatility. This marked the platform’s first cross-margin Auto-Deleveraging (ADL) event in over two years. While some ADL-related trades resulted in unfavorable outcomes, the system maintained solvency and ensured uninterrupted user access throughout.

HyperCore’s infrastructure remains one of Liminal’s greatest strengths, providing low latency and deep liquidity. However, it also means that when liquidity shocks occur across markets, Liminal’s strategies are naturally exposed to the same stress. In such moments, performance may be affected by spread widening or reduced order book depth. These outcomes reflect broader market conditions rather than structural weaknesses.

To prevent concentration risk and limit systemic exposure, Liminal applies strict position caps at both the protocol and per-user levels. They are calibrated to balance opportunity and risk, maintaining meaningful market participation while avoiding concentrations that could amplify stress during dislocations.

Caps and leverage parameters were as follows during extreme market volatility:

post image

These thresholds are determined through quantitative metrics, such as bid–ask spreads, spot–perpetual price differentials, and order book depth. The purpose is to include only the most liquid and attractive assets while avoiding those that could introduce execution or liquidity risk under stress. Market conditions can evolve rapidly, and even the most rigorously designed parameters may be tested during periods of extreme volatility.

Liminal applies intentionally conservative leverage parameters, particularly for assets with higher volatility. This cautious approach prioritizes capital preservation and stability over aggressive yield optimization. By maintaining low leverage limits across riskier markets, the system reduces the likelihood of liquidation or excessive slippage during volatile periods while still capturing structural yield opportunities when conditions are normal.

The design philosophy of Liminal’s current product emphasizes compartmentalization rather than mutualization of risk. Each user’s positions and balances are isolated, ensuring that any adverse outcomes remain contained within individual accounts and cannot propagate through the system. This protects both users and the protocol, but also means that results may vary between accounts depending on leverage, timing, and asset composition.

Overall, Liminal’s current product structure aims to maintain controlled, transparent, and manageable exposure. While market shocks can influence short-term results, the system remains focused on long-term transparency, operational safety, and preserving user access to capital under all circumstances.


ADL impact on Liminal and operational response

The market dislocation of October 10 had a direct and exceptional impact on Liminal’s positions. All strategies on Liminal are delta-neutral, composed of short perpetual positions balanced by spot holdings. As prices fell, the short side of user strategies accumulated substantial unrealized gains, while the spot leg lost value, keeping overall exposure neutral.

As volatility surged and Hyperliquid’s Auto-Deleveraging (ADL) mechanism was triggered, many Liminal users whose short perpetual positions were showing substantial unrealized profits were forcibly closed.

These closures, executed by Hyperliquid’s ADL, were part of a system-wide process to offset losses from undercollateralized traders and preserve overall platform solvency. The most negatively affected positions were concentrated in XPL and FART, where liquidity thinned quickly.

By contrast, positions in major assets such as BTC and ETH remained largely unaffected, as their markets retained sufficient depth. Out of thousands of active positions on these assets, only one Liminal user ended with a negative PnL.

Because of these forced closures, some users’ portfolios temporarily shifted into a net long configuration due to their remaining spot holdings. This imbalance, caused by the ADL process, required system action to prevent further directional exposure.

Liminal’s system response

Following the ADL event, the protocol’s priority was to protect users from unintended risk. With many short positions forcibly closed, portfolios became net long, a configuration outside Liminal’s delta-neutral framework. To eliminate this exposure amid severe volatility, the system closed the remaining spot positions and converted all assets to USDC.

At that stage, market conditions were still unstable, and this precautionary step ensured that users’ capital was fully protected and insulated from further downside risk. While Liminal captured positive spreads across many executions, the same conditions also led to asymmetric outcomes.

Users whose short positions were closed earliest by ADL, during peak illiquidity, experienced less favorable results, as order books were extremely thin and spreads widened sharply. Others benefited from better performance and wider positive spreads as liquidity recovered.

Every position affected by ADL was converted to stablecoins, a necessary measure to preserve system integrity during one of the most severe market dislocations in crypto history.

Throughout the event, withdrawals on Liminal remained fully functional, allowing users to access their funds at any time. As a precaution, new deposits and strategy updates were temporarily paused until liquidity and market depth are normalized.

These actions were taken to prioritize user safety and maintain system integrity, ensuring that all balances remained fully redeemable while avoiding any unintended risk-taking in an unstable environment.

System-level overview

At the protocol level, Liminal concluded the event with a net result of approximately +$355,354, representing about +0.56% of total TVL and an average PnL per ADL event of +$31.41.

These figures reflect that, despite exceptional volatility and disparities in individual outcomes, the system remained resilient, operationally effective, and able to preserve overall capital stability across the platform.

User outcomes

Among all users directly affected by the ADL sequence, 5,147 users (75.2%) ended in profit, and 1,698 users (24.8%) recorded a negative outcome. This distribution reflects the extraordinary volatility and timing effects of the ADL process rather than differences in user strategy or behavior.

Across the entire group, the average return per affected user was +1.8% on their Liminal balance, with a median of +2.18%. Most accounts, therefore, ended net positive despite exceptional market turbulence.

Among users running the Liminal strategy and affected by ADL, 1,800 users (99.94%) ended in profit, and only 1 user recorded a negative outcome. The average return per affected user in Liminal strategy was +3.43% on their Liminal balance, with a median of +3.40%. The only user who recorded a loss had a return of -0.71% (-$6.81).

For users who recorded losses, the average loss was −1.55%, with a median of −1.19%. These outcomes primarily reflected the timing of Hyperliquid’s Auto-Deleveraging (ADL) events, which occurred during the most illiquid moments of the cascade.

Affected users were initially running delta-neutral strategies, with short perpetual positions showing substantial unrealized gains that were forcibly closed by ADL to offset losses elsewhere in the market. The combination of abrupt ADL timing, extreme market stress, temporary fragmentation in liquidity and execution conditions ultimately led to these results.

Among users who ended the event with positive results, average gains reached +2.91%, with a median of +2.95%, reflecting how many accounts were able to benefit from transient market inefficiencies during the dislocation.

Asset-level breakdown

post image

The distribution of outcomes varied significantly by asset, primarily reflecting differences in liquidity, volatility, order book depth, and the timing of Hyperliquid’s ADL process.

Overall, the largest positive contributors were HYPE, SOL, and BTC, which benefited from deeper liquidity and more stable spreads.

Losses were concentrated in smaller and mid-cap assets such as XPL, FART, and PUMP, which quickly became less liquid during the event. These assets also had the lowest per-user exposure caps and a maximum leverage of 1x, meaning that individual users could not allocate more than $500,000 per asset. Execution dispersion was naturally higher on these markets due to thinner order books and wider spreads.

Understanding the dispersion in outcomes

Differences in user results were primarily driven by the timing of Hyperliquid’s ADL triggers, variations in market depth and liquidity, and spread dynamics. During peak volatility, order book depth changed rapidly, spreads widened across markets, and the relationship between spot and perpetual prices fluctuated sharply.

This meant that two users running similar strategies could experience different short-term outcomes depending on execution timing, asset mix, leverage and ADL timing. These discrepancies were most noticeable in lower-liquidity assets and diminished as conditions normalized.

We understand that these differences, especially for users who experienced losses, were difficult. The extreme volatility and timing of Hyperliquid’s ADL had a disproportionate impact on certain positions. These affected users are being carefully considered as we design future user-aligned initiatives aimed at recognizing their experience.


Protocol decision framework

During the market dislocation, the Liminal protocol faced a defining operational decision with meaningful implications for fairness and user trust.

In theory, the protocol could have suspended withdrawals and temporarily pooled user outcomes to balance gains and losses across accounts. While this might have appeared to reduce short-term differences, it would have conflicted with the principles at the foundation of Liminal Customized (current product): self-custody and the segregation of user assets.

Beyond the philosophical inconsistency, such an action would not have been technically possible. Liminal cannot move or reallocate funds from self-custodied accounts without explicit user authorization. Furthermore, users operating in self-custody mode would still have been able to withdraw directly from their Hyperliquid sub-accounts even if withdrawals through the Liminal interface had been paused.

This would have created an unequal situation where Regular-mode users were restricted while Institutional users retained full and immediate access to their funds.

The only approach consistent with both fairness and Liminal’s architecture was therefore to keep withdrawals fully open. This ensured that all users maintained continuous access to their capital and that each account’s performance accurately reflected its own market exposure.

No participant was advantaged or disadvantaged based on custody mode, and user autonomy remained intact throughout. This decision was guided not by convenience but by fairness.

Upholding transparency and user control, especially during market stress, is central to how Liminal operates. These principles define the protocol’s integrity, and it is precisely in moments of turbulence that they matter most.

While this framework ensured fairness and transparency for all users, we understand that the experience was not positive for everyone. Some users faced negative outcomes due to the timing of Hyperliquid’s ADL and the exceptional market stress. Their situation is not being overlooked. It has been taken into account, and we are committed to ensuring that those affected are recognized through future initiatives.


What’s next

With markets and funding rates now stabilized, deposits have reopened and executions are progressively resuming. Users running fully customized strategies must review and confirm their configurations in the app before reactivation, as these strategies will not be re-executed until validated. Regular users running Liminal’s strategy do not need to take any action, while Institutional users should complete the required step displayed in the interface.

Withdrawals remained fully available and will continue to be so under all circumstances. This principle lies at the heart of our commitment to user autonomy and unrestricted access to capital, regardless of market stress or system conditions. Ensuring that users always retain complete control over their funds is a core element of how Liminal operates.

We will also expand the Stats section of the platform to provide more granular performance analytics and real-time market metrics for each asset. By offering deeper visibility into performance drivers and execution conditions, we aim to give users greater clarity and confidence in understanding how the system behaves during different market phases, ultimately helping them make more informed decisions.

While the protocol operated to protect users from more significant downside risk, the overall experience was not positive for everyone. Some users encountered outcomes that were understandably frustrating given the exceptional speed and intensity of the market event. These moments of stress are opportunities for reflection and improvement, and we are determined to learn from them.

We understand that recent conditions led to outcomes that may have felt confusing or disappointing for some users. Such moments can be unsettling, especially when markets move in ways that challenge expectations. Our priority now is to respond with transparency, care, and to ensure that every user feels supported and confident as we move forward.

The entire Liminal team remains deeply committed to refining our framework and strengthening our safeguards. The lessons from this event are already shaping how Liminal will adjust its operations going forward to reduce such risks as much as possible.

Beyond enhancing transparency and strengthening our framework, we are also exploring ways to support users who were affected during the recent volatility. Our goal is to ensure that participants in the Liminal ecosystem continue to feel valued as the protocol evolves through future community-focused initiatives.

We extend our sincere gratitude to all our users for their trust, patience, and constructive feedback. Your confidence and engagement continue to drive us forward as we work to deliver a platform defined by clarity, security, and stability, no matter how turbulent the market becomes.

- The Liminal Team