Liquidation Auction

A Dutch auction (or descending-price auction) starts at a high price that decreases over time until a buyer accepts. This model is designed for speed and is used by protocols like MakerDAO for surplus auctions and Compound v2 for liquidations.

• Mechanism: Price starts above market and falls on a predetermined curve.
• Goal: Incentivize rapid bidding to minimize protocol bad debt.
• Example: A vault's collateral is auctioned starting at 110% of the debt; the price drops until a keeper buys it.

An English auction (or ascending-price auction) starts at a low price, with bidders incrementally raising their offers. The highest bidder wins. This model aims for maximum recovery for the protocol (or the liquidated user).

• Mechanism: Bidding opens at a reserve price and increases through competitive bids.
• Goal: Maximize the sale price of the collateral.
• Use Case: Some protocols use this for collateral auctions where the goal is to recover the full debt plus a penalty.

In a fixed discount auction, liquidators can purchase collateral at a fixed, predefined discount to the oracle price. This is a simplified, first-come-first-served model.

• Mechanism: The protocol sets a discount (e.g., 5%). The first keeper to transact buys the collateral at oracle_price * (1 - discount).
• Advantage: Extremely fast, minimizing price risk during liquidation.
• Protocol Example: Aave and Compound v3 use variants of this model, where liquidators receive a bonus for taking on the liquidation.

A batch auction (or sealed-bid auction) collects all bids within a time window, then clears them at a single, uniform clearing price. This reduces MEV (Miner Extractable Value) from front-running.

• Mechanism: Bidders submit price-quantity pairs secretly. After the period ends, the protocol calculates a market-clearing price.
• Goal: Fair price discovery and MEV mitigation.
• Protocol Example: Gnosis Auction and CowSwap use this model; it's proposed for decentralized liquidation systems.

Direct liquidation is not a true auction but a fixed-price, immediate settlement. A designated liquidator (often the protocol itself or a permissioned actor) closes the position at a predefined penalty rate.

• Mechanism: The undercollateralized position is automatically closed, with collateral seized and sold at oracle price minus a liquidation penalty.
• Characteristic: No bidding process; fastest resolution but less price optimization.
• Example: Centralized exchanges and some lending protocols use this for speed and simplicity.

Some protocols use a dealer network or Over-the-Counter (OTC) model for liquidations. Instead of an open auction, a whitelisted set of market makers or dealers is offered the liquidation at a fixed discount.

• Mechanism: The protocol or a liquidation engine pings registered dealers with an offer.
• Advantage: Reduces market impact for large positions and can offer better pricing.
• Use Case: Employed by institutional-focused protocols or for handling whale positions that could destabilize open markets.

Protocols select auction mechanics based on asset liquidity and desired market fairness.

• Dutch Auction: Price starts high and decreases (e.g., MakerDAO). Efficient for discovering market price.
• Fixed Discount / Incentive: Pre-set bonus for liquidators (e.g., Aave, Compound). Prioritizes speed and simplicity.
• Batch Auctions: Multiple positions liquidated simultaneously at a clearing price. Reduces gas costs and MEV extraction.

The choice impacts liquidation efficiency, keeper profitability, and protocol risk.

The liquidator (or keeper) is the active participant who triggers the liquidation by repaying the borrower's outstanding debt to the protocol. In return, they receive the collateral at a discount, creating an arbitrage opportunity. Their role is critical for system solvency.

• Incentive: Profit from the liquidation penalty or discount on the collateral.
• Function: Monitors on-chain state, identifies undercollateralized positions, and submits the liquidation transaction.
• Risk: Bears gas costs and execution risk if the auction is competitive.

The borrower is the user whose position has fallen below the required collateralization ratio. Their role shifts from active manager to a passive, penalized party in the auction.

• Consequence: Loses their collateral, which is sold to repay their loan.
• Remainder: In many systems, if the auction generates excess proceeds after repaying the debt and fees, the surplus may be returned to the borrower.
• State: Their position is closed, and any remaining debt is cleared.

The protocol itself acts as the auctioneer and rule-setter. Its smart contracts define the auction mechanics, enforce rules, and facilitate the transfer of funds and collateral.

• Responsibilities:
  • Defines the liquidation threshold and health factor.
  • Enforces the auction type (e.g., Dutch auction, English auction).
  • Distributes the liquidated collateral and repays the vault's bad debt.
• Goal: Ensure system solvency by converting insolvent collateral into stable debt repayment with minimal bad debt.

In open auction models (like English auctions or Dutch auctions), bidders compete to purchase the collateral. They are distinct from the initial liquidator who may only start the process.

• Function: Submit bids for the collateral, driving price discovery.
• Auction Types:
  • Dutch Auction: Price starts high and decreases; first bidder wins.
  • English Auction: Bidders compete, driving the price up.
• Outcome: The highest bidder (or first taker) wins the collateral, and proceeds are used to repay the debt.

The specific vault or lending pool that holds the debt is a key beneficiary. It represents the collective lenders whose funds are at risk.

• Primary Interest: Recovers the bad debt owed to it.
• Mechanism: Receives the stablecoins or base asset from the auction proceeds to cover the loan principal and interest.
• Systemic Role: The health of the vault depends on efficient liquidations to minimize losses and maintain liquidity for other users.

While not a direct bidder, the oracle network (e.g., Chainlink, Pyth) is a foundational participant. It provides the external price feeds that determine when a position is undercollateralized and the value of assets during the auction.

• Critical Function: Supplies the fair market price used to calculate health factors and auction starting prices.
• Risk: Oracle manipulation or latency can lead to incorrect liquidations or insufficient auction proceeds.
• Design Impact: Auction mechanisms often include delays or price buffers (oracle safety margins) to mitigate oracle risk.

A borrower's collateral is liquidated when their health factor falls below 1. This occurs due to:

• Collateral value decline (e.g., ETH price drop).
• Debt value increase (e.g., stablecoin de-pegging).
• Interest accrual increasing the debt over time. Borrowers face a liquidation penalty (a fee added to their debt) and lose a portion of their collateral, which is sold at a discount to liquidators.

Liquidators are incentivized by a liquidation bonus (or discount) on the seized collateral. This creates a competitive, time-sensitive environment prone to Maximal Extractable Value (MEV). Bots use tactics like frontrunning and backrunning to win auctions, which can lead to:

• Network congestion and high gas fees.
• Centralization risk, as sophisticated bots dominate.
• Inefficient prices if competition is insufficient.

The auction mechanism itself must be designed to minimize losses. Key risks include:

• Dutch auction vs. fixed discount: Dutch auctions (descending price) can better discover market price but are more complex.
• Slippage and market impact: Large liquidations can move oracle prices, causing cascading liquidations.
• Oracle risk: Reliance on a single price feed can be manipulated, leading to unfair liquidations (oracle manipulation).

If a liquidation auction fails to attract sufficient bids, the protocol may be left with undercollateralized debt, threatening its solvency. This bad debt must be covered by a protocol treasury or a safety module. Failure modes include:

• Flash crash scenarios where oracle prices diverge from market prices.
• Liquidity crunches where no liquidator can absorb the sale.
• Black swan events that overwhelm the system's designed parameters.

Front-end applications must clearly communicate liquidation risk. Poor UX can lead to unexpected losses. Essential features include:

• Real-time health factor monitoring and alerts.
• Clear display of liquidation price thresholds.
• Transparency on the auction process, including pending liquidations and historical prices. Lack of these can result in user error and loss of funds.

The process of automatically selling a borrower's collateral to repay their outstanding debt when their health factor or collateralization ratio falls below a predefined threshold. This is the core risk-management mechanism in lending protocols.

• Trigger: A price drop or debt increase pushes the position into an undercollateralized state.
• Purpose: Protects the protocol and lenders from bad debt by ensuring loans are always overcollateralized.

A bot or user who participates in a liquidation auction by repaying a portion of the borrower's debt in exchange for the collateral at a discounted rate. Their activity is essential for system solvency.

• Incentive: Earns a liquidation bonus (or penalty) on the seized collateral.
• Role: Provides immediate liquidity to close unhealthy positions, competing in auctions for the most profitable opportunities.

A numerical representation of the safety of a borrowed position, calculated as (Total Collateral Value * Liquidation Threshold) / Total Borrowed Value. It determines when a position can be liquidated.

• > 1: Position is safe.
• <= 1: Position becomes eligible for liquidation.
• Key Metric: Continuously monitored by protocols and liquidator bots to trigger auctions.

The ratio of the value of deposited collateral to the value of the borrowed assets. It is the inverse of Loan-to-Value (LTV) and is a fundamental measure of a position's risk.

• Formula: (Collateral Value / Debt Value).
• Minimum Ratio: Set by the protocol; falling below this triggers a liquidation.
• Example: A 150% ratio means $150 of collateral backs $100 of debt.

A common type of liquidation auction where the price of the collateral for sale starts high and decreases over time until a liquidator accepts the price. This design efficiently discovers the market-clearing price.

• Mechanism: Begins at a price above market value, descending in steps.
• Goal: To sell the collateral quickly while maximizing recovery for the protocol and minimizing loss for the borrower.

Debt that cannot be fully recovered through the liquidation process, typically occurring when collateral value falls too rapidly or auction mechanisms fail to attract liquidators. This represents a systemic risk.

• Cause: Extreme market volatility, oracle failure, or insufficient liquidation incentives.
• Impact: Losses are socialized or absorbed by a protocol's insurance fund or treasury.