Over-Collateralized Loan

The Collateralization Ratio (CR) is the core risk parameter, calculated as (Value of Collateral / Value of Loan) * 100. A 150% CR means $150 of collateral backs a $100 loan. Protocols set minimum collateralization ratios (e.g., 110% for MakerDAO's ETH-A vault) to create a safety buffer against price volatility. Falling below this triggers liquidation.

This is the automated enforcement of the loan's safety. If the collateral value drops too close to the loan value (hitting the liquidation threshold), a liquidation event occurs. Liquidators can purchase the undercollateralized collateral at a discount, repaying the loan and keeping the difference. This process protects lenders from bad debt and is a critical, programmatic feature of protocols like Aave and Compound.

The cost of borrowing, often called a Stability Fee (in DAI loans) or variable interest rate. This is typically paid in the borrowed asset and accrues over time. Rates are often set by governance and can be dynamic based on market utilization. For example, borrowing DAI from a Maker Vault incurs a stability fee, while borrowing USDC on Aave incurs a variable APY.

A protocol-level risk management tool that limits the total amount of a specific asset that can be borrowed against a particular collateral type. For instance, MakerDAO may set a debt ceiling of 100 million DAI for its wstETH-B vault. This prevents overexposure to any single collateral asset and manages systemic risk within the protocol.

Protocols whitelist specific collateral assets (e.g., ETH, wBTC, stETH) based on liquidity, volatility, and market depth. Each asset has unique risk parameters:

• Liquidation Threshold: The CR at which liquidation begins.
• Liquidation Penalty: The discount liquidators receive.
• Loan-to-Value (LTV) Ratio: The maximum borrowing power (e.g., 70% LTV means you can borrow up to $70 against $100 of collateral).

Why lock up more value than you borrow?

• Access Liquidity Without Selling: Borrow against long-term holdings for expenses or other investments, maintaining exposure to asset appreciation.
• Leverage: Re-borrow and re-deposit collateral to increase exposure (a risky strategy).
• Yield Farming: Borrow a stablecoin to provide liquidity in a farming pool, aiming for a higher yield than the borrowing cost (yield arbitrage). The over-collateralization is necessary due to the absence of credit checks and the need for instant, automated settlement.

The primary risk for a borrower is liquidation. If the value of the posted collateral falls below the required collateralization ratio (e.g., 150%), a portion of the collateral is automatically sold to repay the loan, often at a penalty. This can occur due to high market volatility.

• Liquidation Penalty: Borrowers lose a percentage of their collateral (e.g., 5-15%) as a fee.
• Price Oracle Risk: Liquidations rely on price feeds; stale or manipulated data can trigger unfair liquidations.

Not all collateral is equally stable. Borrowing against highly volatile assets (e.g., altcoins) or assets that can depeg (e.g., algorithmic stablecoins) significantly increases liquidation risk. A sharp, correlated market downturn can trigger mass liquidations across a protocol, overwhelming the liquidation engine and potentially leading to bad debt.

The loan agreement is enforced by smart contracts. Borrowers are exposed to:

• Code Exploits: Bugs or vulnerabilities in the lending protocol's contracts can lead to loss of funds.
• Governance Risk: Protocol parameters (like collateral factors, interest rates) can be changed by governance votes, affecting loan safety.
• Admin Key Risk: Some protocols retain emergency admin keys, creating a centralization vector.

Over-collateralization locks up capital that could be deployed elsewhere. This opportunity cost is a fundamental economic consideration. A borrower locking $15,000 in ETH to borrow $10,000 in stablecoins foregoes potential yield or gains on the full $15,000. This inefficiency is a key driver for the development of under-collateralized and credit-based lending models.

In stressed market conditions, a wave of liquidations can create a positive feedback loop: forced sales drive collateral prices down further, triggering more liquidations. This systemic risk can destabilize the entire protocol and spill over to connected DeFi ecosystems, as seen during major market crashes like March 2020 ("Black Thursday").

Borrowers and protocols employ several strategies to manage these risks:

• Health Factor Monitoring: Borrowers must actively monitor their loan's health factor or collateral ratio.
• Using Stable Collateral: Preferring less volatile collateral like wrapped Bitcoin (wBTC) or stablecoins.
• Protocol Audits & Insurance: Relying on audited code and considering DeFi insurance coverage.
• Circuit Breakers: Some protocols implement mechanisms to pause liquidations during extreme volatility.

Liquidation is the enforced closure of an under-collateralized loan to protect lenders. It triggers when a borrower's collateralization ratio falls below the protocol's minimum threshold. A portion of the borrower's collateral is automatically sold, often at a discount, to repay the debt plus a liquidation penalty. This process is typically performed by bots or keepers, ensuring the lending pool remains solvent.

A debt ceiling is a protocol-level parameter that limits the total amount of debt (e.g., DAI) that can be generated against a specific type of collateral. It is a critical risk management tool that prevents over-exposure to any single asset. For instance, MakerDAO governance may set a debt ceiling of $500 million for USDC as collateral. This cap helps manage systemic risk within the lending protocol.

A liquidity pool is a smart contract containing funds that facilitate lending, borrowing, and trading. In lending protocols like Aave, users supply assets to a pool to earn interest, while borrowers draw loans from the same pool. Over-collateralization is enforced at the individual position level, but the pooled structure aggregates capital and risk, determining interest rates algorithmically based on supply and demand.