Collateral Auction

An auction is initiated when a borrower's Health Factor (HF) drops below 1.0, indicating their collateral value is insufficient relative to their debt. This is calculated as:

• HF = (Collateral Value * Liquidation Threshold) / Borrowed Value Protocols use Oracles (e.g., Chainlink) for real-time price feeds to determine this. A lower Liquidation Threshold for volatile assets triggers auctions earlier to protect the protocol.

Two main auction models exist to sell seized collateral:

• Fixed Discount / Dutch Auction: Starts at a high price and decreases over time until a buyer accepts. Used by MakerDAO. Maximizes collateral recovery for the protocol.
• English Auction / Reverse Auction: Starts at a low price (often the debt value) and increases with bids. Used by Aave and Compound. Designed to minimize collateral shortfall for the borrower. The choice impacts liquidation penalties and system risk.

Keepers (bots/agents) monitor the blockchain and participate in auctions for profit. They create a competitive landscape crucial for timely liquidations. This activity is a major source of Maximal Extractable Value (MEV), as keepers compete to submit the winning transaction first. Protocols may use Keeper incentives (e.g., a fixed bonus) to ensure auction participation, especially during network congestion.

If an auction fails to cover the debt (e.g., due to a market crash or no keeper participation), it creates a collateral shortfall or bad debt. Protocols manage this through:

• Protocol-Insured Vaults: Using accumulated fees (e.g., Stability Fees) as a backstop.
• Recapitalization Auctions: Minting and auctioning protocol tokens (like MKR in MakerDAO's Debt Auction) to cover the gap, diluting token holders. This is a critical risk management failure state.

• MakerDAO: Uses a complex multi-stage process with Collateral Auction (FLIP) to sell collateral, Debt Auction (FLAP) to sell surplus DAI, and Surplus Auction (FLOP) to recapitalize with MKR.
• Aave V2/V3: Employs a liquidation bonus (e.g., 5-10%) in a single-step, fixed-price liquidation, which is a simplified auction.
• Compound: Uses a similar model to Aave, where liquidators can repay a portion of debt to receive collateral at a discounted price.

Auction design involves critical trade-offs:

• Efficiency vs. Borrower Cost: Faster, more aggressive auctions protect the protocol but increase loss for borrowers.
• Oracle Risk: Dependence on price feeds makes auctions vulnerable to oracle manipulation or staleness.
• Network Congestion: High gas fees can disincentivize keepers, leading to failed auctions and bad debt.
• Collateral Liquidity: Auctions for illiquid assets may result in deeper discounts and larger shortfalls.

Aave V3 employs a liquidation auction model where liquidators can purchase discounted collateral. The process involves:

• A liquidation bonus (e.g., 5-15%) incentivizes liquidators.
• The auction is a single-block, sealed-bid process via the liquidationCall function.
• Liquidators repay a portion of the borrower's debt in the borrowed asset and receive the equivalent value of collateral plus the bonus.
• This mechanism helps maintain the health factor of the protocol by quickly resolving undercollateralized positions.

Compound uses a fixed liquidation incentive model rather than a dynamic auction. When a position's collateral factor is breached, liquidators can repay up to 50% of the borrowed amount in a single transaction. In return, they seize the borrower's collateral at a discount defined by the liquidation incentive (e.g., 8% for most assets). This creates a competitive, gas-auction-like environment where liquidators compete to be the first to execute the profitable transaction, ensuring rapid resolution of insolvent accounts.

Liquity uses a unique two-pronged system instead of traditional auctions. Primary liquidations are handled by a Stability Pool, where LQTY stakers and LUSD depositors absorb debt in exchange for liquidated collateral at a net gain. As a backstop, a recollateralization auction can be triggered if the Stability Pool is depleted. Furthermore, Liquity's redemption mechanism allows anyone to exchange LUSD for underlying ETH at face value, which acts as a perpetual, permissionless auction that anchors the peg.

The Frax Protocol combines algorithmic and collateral-backed stability. For its Frax stablecoin, undercollateralized AMO (Algorithmic Market Operations Controller) positions can be liquidated. The protocol employs a collateral auction for its Frax Lending module (Fraxferry), where liquidated collateral is sold to cover bad debt. The design focuses on minimizing protocol-owned liquidity risk and ensuring the collateral ratio is maintained through automated, incentivized liquidations.

Protocols implement collateral auctions with distinct parameters and goals:

• Discount Rate / Bonus: The incentive for liquidators (fixed vs. dynamic).
• Auction Duration: From single-block (Aave) to multi-hour/days (Maker).
• Bidding Type: Sealed-bid, open ascending, or Dutch auction.
• Settlement Asset: Paid in stablecoin (DAI, USDC) or protocol token (MKR).
• Trigger Condition: Based on Loan-to-Value (LTV), health factor, or collateral ratio thresholds. These variations balance liquidation efficiency, collateral recovery, and market stability.

The liquidation incentive (or bonus) is a reward paid to liquidators for covering a bad debt. If set too high, it can encourage predatory behavior and market manipulation. If set too low, it may fail to attract liquidators, leaving the protocol with bad debt. Attackers can exploit this by artificially triggering liquidations through oracle manipulation or flash loan attacks to claim the incentive, harming the position owner.

An auction fails when no liquidator bids, often due to insufficient incentive, illiquid collateral, or extreme market volatility. This results in bad debt for the protocol, which must be socialized or covered by a treasury reserve. In Dutch auction models, a long, unattractive price decay can increase the risk of failure. Protocols mitigate this with mechanisms like fixed discount sales or liquidation pools.

Collateral auctions are prime targets for Maximal Extractable Value (MEV). Searchers use bots to monitor the mempool for pending liquidations. Risks include:

• Oracle price manipulation via flash loans to trigger false liquidations.
• Sandwich attacks on the auction's closing transaction to profit from price impact.
• Time-bandit attacks where miners reorder transactions to steal profitable liquidations.

During market crashes, a surge in liquidations triggers intense competition among liquidator bots. This leads to gas wars, where bots bid exponentially higher transaction fees to be first in line. The resulting network congestion can:

• Delay critical liquidations, increasing bad debt.
• Make participation prohibitively expensive for smaller actors.
• Cause failed transactions due to gas estimation errors, leaving positions open.

Auctions for volatile or illiquid collateral (e.g., long-tail assets) carry high price slippage risk. A liquidator may win a bid but be unable to sell the asset on a DEX for the expected price, incurring a loss. This disincentivizes participation. Furthermore, rapid price swings during the auction window can make the starting bid instantly outdated, leading to suboptimal recoveries for the protocol.

Auction security depends on correctly configured parameters set by protocol governance, including liquidation penalty, auction duration, and minimum bid decrement. Poorly chosen parameters are a systemic risk:

• Governance attacks could alter parameters to enable theft.
• Static misconfiguration can make the system vulnerable during black swan events.
• Updates often require complex, time-locked proposals, limiting agility during crises.

A reverse auction mechanism used to recapitalize a protocol's system surplus after a collateral auction fails to cover the bad debt. The protocol mints and auctions new protocol tokens (e.g., MKR, LQTY) to raise capital, with proceeds used to burn the outstanding system debt.

• Purpose: Restore the collateralization ratio after a shortfall event.
• Direction: Bidders offer a fixed amount of stablecoins for a decreasing amount of protocol tokens.
• Trigger: Activated when the proceeds from a collateral auction are insufficient.

The automated process of forcibly closing an undercollateralized position to repay its debt before triggering an auction. It is the precursor event to a collateral auction.

• Trigger: Occurs when a position's Loan-to-Value (LTV) ratio falls below the liquidation threshold.
• Mechanism: A liquidation penalty (or fee) is applied, and a portion of the collateral is sold, often via a liquidation engine to keepers or directly to a market.
• Goal: Protect the protocol from insolvency by ensuring debt is always over-collateralized.

A network participant or bot that monitors the blockchain for profitable opportunities to participate in liquidation and auction processes, providing essential liquidity and system health.

• Role: They bid in collateral auctions or trigger liquidations when conditions are met.
• Incentive: Earns a liquidation bonus or auction discount for their service.
• Function: Acts as a decentralized agent maintaining protocol solvency by ensuring auctions are executed efficiently.

A common format for collateral auctions where the price of the asset starts high and decreases over time until a bid is received. This design encourages rapid price discovery.

• Mechanism: Also known as a descending price auction. The starting price is often set above market value.
• Advantage: Maximizes recovery for the protocol by capturing bidders' willingness to pay.
• Example: Used by MakerDAO's collateral auction system for selling seized collateral like ETH or WBTC.

The variable interest rate accrued on debt generated against locked collateral. It is a key risk parameter that influences the likelihood of liquidation and subsequent auctions.

• Function: Paid in the protocol's native stablecoin (e.g., DAI) or governance token.
• Purpose: Compensates the protocol for risk and manages demand for credit.
• Impact: Higher fees increase borrowing costs, potentially leading to more positions becoming undercollateralized during market downturns.

The primary metric determining the health of a collateralized debt position (CDP), calculated as (Value of Collateral / Value of Debt) * 100%.

• Minimum Ratio: The liquidation ratio is the threshold below which liquidation is triggered.
• Safety Buffer: Users maintain a ratio above the minimum to avoid market volatility triggering an auction.
• Monitoring: Continuously tracked by oracles; a falling ratio is the direct precursor to auction processes.