Pricing, Matching & Stress Behaviour

Pricing and Execution Fundamentals

All liquidation activity and internal loan-related swaps are executed directly on the Chainflip DEX using the JIT AMM. This provides several consistent pricing characteristics:

• LPs compete on price for every order
• Execution generally occurs at or near market prices
• Both soft and hard liquidations use oracle-indexed minimum prices via Live Price Protection (LPP)
• No dependency on external liquidators or third-party settlement networks

Because the DEX already processes large volumes of cross-chain swaps, liquidation flow is handled naturally as part of normal market-making activity.

Matching Engine Behaviour

The matching engine treats liquidation orders similarly to internal DCA-style swaps:

• Orders are submitted in fixed-size chunks (per parameters)
• Each chunk has a fill-or-kill behaviour per block
• If not filled, a new order is submitted at updated oracle prices and LPP bounds
• LPs may fill partial or full amounts based on their quoting strategy

Liquidation orders remain fully permissionless, any LP can fill them.

The system prioritises:

• Safety (LPP limits)
• Fairness (LP competition)
• Determinism (consistent scheduling and update cadence)

Liquidations do not override or block non-liquidation swaps; both types of flow coexist seamlessly.

Behaviour During Soft Liquidation

Soft liquidations behave like a controlled unwind:

• Frequent, small liquidation chunks
• Tight LPP bounds (e.g., ~0.5% below oracle)
• LP competition ensures execution close to the live market
• Liquidation halts automatically when LTV improves

This approach reduces borrower losses and prevents unnecessary hard-liquidation escalation.

Behaviour During Hard Liquidation

When LTV breaches the hard threshold:

• Liquidation chunks increase in size
• LPP minimums widen (e.g., up to ~5% below oracle)
• Submission cadence remains block-by-block
• LPs still compete on price, keeping execution near fair value

LP competition continues to anchor fills to market rates.

Handling Many Liquidations Simultaneously

Chainflip is designed to handle large-scale liquidation events without degrading normal swap UX.

During high-stress periods:

• Quotes remain available
• Refund logic in the DEX behaves normally
• LP bots continue operating
• Liquidation and non-liquidation flow are matched simultaneously
• Competing LPs naturally rebalance inventories via external venues

The DEX’s architecture avoids the “clear the books first” behaviour seen in more primitive AMMs.

LP Behaviour and Strategy Under Stress

LPs may adjust spreads or participation during heavy liquidation flow. Important properties:

• Liquidation opportunities appear as normal LP fill opportunities
• LPs quote based on available liquidity on-chain and externally
• Inventory-sensitive bots (hedged/asymmetric) can adapt dynamically
• No LP has priority, the system remains fully permissionless

Pricing remains competitive because:

• many LPs operate concurrently
• liquidation orders update every block
• oracle pricing stabilises minimum bounds

Chainflip’s DEX has proven the ability to rebalance across severe global volatility.

Swapper UX During Liquidations

Normal user swaps continue unaffected:

• Quotes remain real-time
• LPP continues to protect users
• Refunds follow standard logic
• Execution latency remains stable

Liquidations do not block or override normal swaps; they run on the same schedule and fill conditions.

This is a major advantage versus systems where liquidations compete for gas or block space (e.g., Ethereum).

Summary of Stress Behaviour Guarantees

Chainflip’s pricing and matching engine ensures:

• Deterministic execution
• Oracle-anchored pricing
• Block-by-block liquidation progression
• Fair LP competition
• No privileged liquidators
• Swapper UX remains intact
• DEX behaviour remains stable under load
• Gradual unwinding instead of catastrophic liquidation

These properties significantly reduce the risk of cascading failures common in external-liquidator models.