A cluster of 6 wallets, funded via Bitget deposits, coordinated a leveraged long position exceeding $10 million on XPL likely a perpetual futures contract on the Hyperliquid decentralized exchange.
The wallets deposited around $1.5 million in collateral. They aggressively built long exposure, pushing or riding the price of XPL upward. As unrealized profits grew, they withdrew approximately $3 million; realizing gains while the position was still open or partially managed. When the price reversed sharply, the cluster’s positions were liquidated.
This liquidation cascade triggered over $10 million in additional liquidations and an ADL (Auto-Deleveraging) backstop on Hyperliquid for XPL, amplifying the downside move. One analysis on X suggested it might stem from an API key leak affecting a large trader who had a separate $50 million long position likely lower leverage and still open elsewhere, possibly on Binance or another venue.
This appears to be a classic case of coordinated leveraged trading rather than outright on-chain token price manipulation via spot buys and sells. The profits came from futures PnL on Hyperliquid. The $3M withdrawal happened on the way up, but the cluster ultimately got liquidated on the reversal, so the profit was extracted before the full blow-up.
Register for Tekedia Mini-MBA edition 20 (June 8 – Sept 5, 2026).
Register for Tekedia AI in Business Masterclass.
Join Tekedia Capital Syndicate and co-invest in great global startups.
Register for Tekedia AI Lab.
Such events highlight risks in high-leverage perp markets: clustered positions can create feedback loops of liquidations, especially on lower-liquidity tokens like XPL. Hyperliquid’s mechanism handled the bad debt via its insurance and backstop, but it still caused volatility.
Hyperliquid’s Auto-Deleveraging (ADL) is a last-resort solvency mechanism for its perpetual futures markets. It kicks in only after all other risk controls fail, ensuring the platform never has bad debt (negative equity that can’t be covered). It does this by forcibly closing a portion of the most profitable positions on the opposite side of the bankrupt trade—at the prevailing mark price—so the underwater position can be offset without draining external funds.
Hyperliquid’s liquidation process is layered and designed to maximize trader retention of capital while protecting the platform: Normal (Book) Liquidation Trigger: Account equity falls below maintenance margin typically 1.25%–16.7% of notional, depending on the asset’s max leverage tier. The system sends market orders to the public order book to close the full position (or 20% initially for large positions >$100k USDC, then full after a 30-second cooldown).
The liquidated trader keeps any remaining equity. No liquidation fees. Fully competitive—anyone can take the flow. Equity drops below 2/3 of maintenance margin and book liquidation fails. The entire position + associated margin is transferred to the Hyperliquid Liquidity Provider (HLP) vault.
All cross positions and margin go to HLP ? trader equity goes to zero. Only the isolated position/margin is taken. HLP absorbs the position. On average, these are profitable for the community (PNL flows back to HLP depositors). Maintenance margin is not returned to the trader, this buffer ensures HLP profitability.
Hyperliquid’s ADL logic is deliberately simple and mirrors mainstream centralized exchanges, but executed fully on-chain. ADL is rare by design. The first cross-margin ADL occurred during a major volatility event on Oct 10, 2025, where ~$2.1 billion notional was deleveraged in ~12 minutes across many markets.
In the recent XPL incident you referenced, the cluster’s liquidation cascade exhausted local liquidity + HLP buffers, triggering ADL and amplifying the move via forced closures.
Hyperliquid’s production queue sometimes over-utilized ADL relative to an optimal policy—closing ~28× more notional than the theoretical minimum needed to cover the shortfall, resulting in an estimated $45M–$52M in excess PnL haircuts to winners.
The paper argues better algorithms could reduce unnecessary deleveraging while still guaranteeing solvency. Hyperliquid’s co-founder has pushed back, noting that ADL has net delivered hundreds of millions in realized profits to users by closing winners at favorable prices rather than letting HLP take more risk.
Hyperliquid’s ADL is a robust, transparent nuclear option that prioritizes platform solvency above all while trying to minimize socialization to only the most profitable opposing traders. It has proven effective at preventing insolvency in real stress events, though debates continue on whether the current ranking algorithm is optimally efficient. The mechanism continues to evolve, with the core philosophy remaining: keep it simple, on-chain, and strictly solvent.