Multi-Collateral Debt

Unlike single-collateral systems, a Multi-Collateral Debt Position (MCD) can be backed by a portfolio of different assets. This reduces risk concentration and allows borrowers to leverage a wider range of assets. Common collateral types include:

• Volatile Assets: ETH, WBTC
• Stable Assets: LP tokens, staked assets (e.g., stETH)
• Real-World Assets (RWAs): Tokenized treasury bills, private credit Each asset has a unique Collateralization Ratio and Stability Fee set by governance.

Each collateral type has specific, adjustable risk parameters managed by protocol governance. Key parameters include:

• Liquidation Ratio: The minimum collateral-to-debt ratio before a position becomes eligible for liquidation.
• Liquidation Penalty: A fee applied to the debt of liquidated positions.
• Debt Ceiling: A maximum amount of debt that can be generated against a specific collateral type. These parameters create a risk-tiered system, where more volatile assets require higher overcollateralization.

A key feature of multi-collateral systems like MakerDAO is the DAI Savings Rate (DSR). This is the interest rate paid to users who lock DAI in a savings contract within the protocol. The DSR is a primary tool for monetary policy:

• Increasing the DSR encourages holding DAI, reducing supply and increasing its market price.
• Decreasing the DSR discourages holding, increasing supply. This mechanism helps maintain the peg of the generated stablecoin to its target value (e.g., 1 USD).

The addition of new collateral types and adjustments to risk parameters are not automated; they are executed through decentralized governance. Token holders (e.g., MKR holders in MakerDAO) vote on:

• Collateral Onboarding: Adding new asset types via Risk Core Units and Spell contracts.
• Parameter Changes: Adjusting liquidation ratios, stability fees, and debt ceilings.
• Emergency Shutdown: A last-resort mechanism to settle all positions at a fixed collateral price.

To manage systemic risk, each collateral asset has a Debt Ceiling, which is the maximum amount of stablecoin debt that can be generated against it. This prevents over-concentration in any single asset. The Global Debt Ceiling is the sum of all individual ceilings. When a specific asset's ceiling is reached, no new debt can be minted against it until existing debt is repaid or the ceiling is raised via governance. This acts as a circuit breaker for risk.

Multi-Collateral Debt Positions (MCD) evolved from Single-Collateral DAI (SAI), which only accepted ETH. The key advancements are:

• Risk Management: Diversification across uncorrelated assets.
• Scalability: Multiple debt ceilings allow for greater total system debt.
• Monetary Policy: Tools like the DSR provide more precise control over the stablecoin's economics.
• Accessibility: Allows users with diverse portfolios to participate without first converting assets to a single type.

A Collateralized Debt Position (CDP) or Vault is the fundamental unit. Users lock approved assets (e.g., ETH, WBTC) to mint a stablecoin like DAI. The system calculates a collateralization ratio (collateral value / debt value). If this ratio falls below a liquidation ratio, the position is subject to liquidation to ensure the protocol remains solvent.

Each collateral type has unique risk parameters set by governance:

• Debt Ceiling: Maximum debt that can be minted against that asset.
• Liquidation Ratio: Minimum collateralization level before liquidation.
• Stability Fee: An interest rate on the generated debt.
• Liquidation Penalty: A fee applied during liquidation. These parameters manage systemic risk and volatility exposure from different assets.

When a vault becomes undercollateralized, a liquidation engine triggers an auction. Keepers (automated bots) bid to purchase the collateral at a discount, using the stablecoin debt to repay the vault. Common models include:

• Fixed Discount / English Auctions: Collateral is sold to the highest bidder.
• Dutch Auctions: The price starts high and decreases until a keeper buys. This process ensures bad debt is covered and the stablecoin remains overcollateralized.

• MakerDAO: The pioneer, accepting ETH, WBTC, and real-world assets (RWAs) to mint DAI.
• Liquity: Accepts only ETH as collateral for LUSD, using a Stability Pool for liquidations.
• Abracadabra.money: Accepts interest-bearing tokens (like yvUSDC) as collateral to mint MIM.
• Aave & Compound: While primarily lending markets, their stablecoin borrowing functions similarly, using multiple assets as collateral for loans.

Benefits:

• Diversification: Reduces systemic reliance on a single asset.
• Capital Efficiency: Users can leverage diverse portfolios.
• Stability: Broader collateral base can dampen volatility impacts.

Trade-offs:

• Complexity: Risk parameters must be actively managed per asset.
• Governance Overhead: Adding new collateral requires extensive analysis and voting.
• Oracle Risk: Reliance on price feeds for multiple assets increases attack surface.

Beyond collateral, protocols employ secondary mechanisms:

• Surplus/Deficit System: Fees (Stability Fees, liquidation penalties) accrue as system surplus, used to cover bad debt (system deficit).
• Emergency Shutdown: A governance-triggered failsafe that settles all positions at a fixed price, redeeming users.
• Collateral Rebalancing: Protocols like MakerDAO use PSM (Peg Stability Module) modules with off-chain assets to directly defend the peg, complementing the multi-collateral system.

A liquidation cascade occurs when a sharp price drop in one collateral asset triggers mass liquidations, causing a feedback loop that depresses prices further. This systemic risk is amplified in multi-collateral systems where correlated assets (e.g., wBTC and ETH) can fail simultaneously. Key factors include:

• Oracle latency: Stale price feeds can trigger unnecessary liquidations.
• Liquidity depth: Insufficient market depth for the collateral being sold.
• Correlation risk: High price correlation between different accepted collateral types.

The integrity of the price oracle is paramount, as it determines collateral valuations and liquidation thresholds. Attack vectors include:

• Flash loan attacks: Borrowing large sums to manipulate spot prices on a DEX that feeds the oracle.
• Front-running bots: Monitoring the mempool to execute liquidations before users can add collateral or repay debt.
• Oracle delay exploits: Exploiting the time lag between a market price drop and the oracle update to drain undercollateralized positions.

Not all collateral is created equal. Multi-collateral systems must assess and mitigate the specific risks of each asset type:

• Stablecoin depegs: If a system accepts algorithmic stablecoins, a depeg event can instantly render many positions undercollateralized.
• Liquidity risk: Long-tail or illiquid assets may have no viable market during a liquidation, causing bad debt.
• Smart contract risk: Collateral in the form of yield-bearing tokens (e.g., stETH, aTokens) inherits the risks of its underlying protocol.

Governance decisions on risk parameters directly impact system stability. Poorly set parameters can lead to insolvency or inefficient capital use.

• Liquidation ratio (LR): If set too low, positions become undercollateralized too quickly during volatility.
• Liquidation penalty: If too high, it can exacerbate debt deficits; if too low, it may not incentivize liquidators.
• Debt ceiling: Per-asset limits prevent over-concentration in a single, potentially risky collateral type.

The underlying code and its evolution present technical risks.

• Implementation bugs: A flaw in the core CDP logic or price feed integration can lead to fund loss.
• Governal attack: Malicious governance proposals could alter parameters to benefit attackers.
• Upgrade complexity: Upgradable contracts in a multi-collateral system are highly complex, increasing the risk of introducing new bugs during migrations.

Multi-collateral interfaces increase cognitive load, leading to operational mistakes.

• Cross-position confusion: Users managing multiple vaults with different collaterals may misjudge their overall health.
• Gas optimization risks: During network congestion, transactions to top up collateral or repay debt may fail, resulting in avoidable liquidation.
• Approval risks: Interacting with numerous token contracts increases the attack surface for phishing or malicious contract approvals.

The Collateralization Ratio (CR) is the value of a user's locked collateral relative to their borrowed debt, expressed as a percentage. It is the primary metric for assessing liquidation risk.

• Minimum CR: The threshold below which a position is eligible for liquidation.
• Health Factor: A related metric (inverse of CR) used in protocols like Aave, where a value below 1 triggers liquidation.
• Dynamic Management: Users must actively manage their CR by adding collateral or repaying debt to avoid market volatility triggering a liquidation event.

A Debt Ceiling is a protocol-level parameter that limits the total amount of stablecoin debt (e.g., DAI) that can be generated against a specific collateral type. It is a critical risk management tool.

• Risk Isolation: Prevents overexposure to any single, potentially volatile asset.
• Governance Control: Set and adjusted via decentralized governance (e.g., MakerDAO's MKR holders).
• Example: A debt ceiling for wrapped Bitcoin (WBTC) might be set at $500M to cap systemic risk from that collateral pool.

The Stability Fee is the variable annual interest rate charged on borrowed debt in a multi-collateral system (e.g., the DAI Savings Rate adjustment in Maker). It is distinct from a liquidation penalty.

• Monetary Policy Tool: Used by governance to control the supply and demand for the generated stablecoin, helping to maintain its peg.
• Accrues in Debt: The fee is not paid periodically but is continuously added to the user's outstanding debt balance.
• Risk-Based: Can vary per collateral type, reflecting its volatility and liquidity risk.

The Liquidation Engine is the automated mechanism that seizes and sells undercollateralized positions. It typically uses a Dutch Auction or Fixed Discount model to sell collateral.

• Liquidation Penalty: A fee (e.g., 13% in Maker) added to the debt, paid to the liquidator.
• Keeper Network: A decentralized set of bots (keepers) that monitor positions and trigger liquidations for profit.
• Goal: To repay the bad debt and return excess collateral to the user, ensuring the protocol remains solvent.

Each approved Collateral Type (e.g., ETH, WBTC, Real-World Assets) has unique Risk Parameters set by governance. These define the asset's role and risk within the system.

• Key Parameters: Debt Ceiling, Liquidation Ratio, Stability Fee, and Liquidation Penalty.
• Risk Assessment: Based on asset price volatility, liquidity, and custodial risk.
• Collateral Onboarding: A formal governance process (through MakerDAO's Governance or similar) is required to add new asset types, involving risk reports and community votes.