Collateral Haircut

The primary function of a haircut is to create a buffer between the loan value and the collateral's market value. This buffer absorbs price fluctuations, preventing the loan from becoming under-collateralized during normal market movements. For example, a 20% haircut on $100 of ETH means only $80 can be borrowed, creating a 20% safety margin before a liquidation is triggered.

The size of the haircut is not arbitrary; it is algorithmically determined based on the asset's risk profile. Key factors include:

• Volatility: High-volatility assets (e.g., memecoins) receive larger haircuts.
• Liquidity: Less liquid assets are discounted more heavily to account for slippage during forced sales.
• Price Oracle Reliability: Assets with less reliable price feeds may have higher haircuts.
• Correlation Risk: Assets highly correlated with the borrowed asset increase systemic risk.

The haircut directly defines the liquidation threshold. If the collateral value (after haircut) falls below the loan value, the position becomes eligible for liquidation. This is tracked via a Health Factor (HF). For instance, with a 25% haircut, the liquidation threshold is 75% of the collateral's value. An HF below 1.0 indicates the collateral's adjusted value no longer covers the debt, triggering liquidation.

Haircuts are implemented differently across DeFi protocols, affecting user strategy. Compound and Aave use distinct collateral factors for each asset. MakerDAO applies a Stability Fee and Liquidation Penalty on top of the collateral's Loan-to-Value ratio. In over-collateralized stablecoins like DAI, the haircut is embedded in the Collateralization Ratio (e.g., 150%), meaning $150 of collateral is needed to mint $100 DAI.

In traditional finance, the Initial Margin requirement in leveraged trading is the conceptual equivalent of a collateral haircut. Both mandate that a borrower/trаder posts more value than they receive. The key difference is operational: a haircut is a value discount applied to posted collateral, while initial margin is the minimum equity that must be deposited to open a position.

Haircuts are a direct trade-off between safety and capital efficiency. A lower haircut allows users to borrow more against their collateral, increasing efficiency but raising protocol risk. Risk managers must balance this to ensure system solvency. During market stress, protocols like MakerDAO can vote to increase haircuts (lower collateral factors) via governance to rapidly de-risk the system.

A collateral haircut is the percentage discount applied to the market value of an asset when calculating its collateral value for a loan or derivative position. Its primary purpose is to create a safety buffer (or overcollateralization) that protects the lender or protocol from:

• Price volatility: Sudden market drops.
• Liquidation risk: The time and cost to sell the asset.
• Oracle inaccuracy: Potential delays or errors in price feeds.

The haircut percentage is determined by a risk assessment of the collateral asset. Key factors include:

• Asset Volatility: Stablecoins (e.g., USDC) have low haircuts (~1-5%), while volatile crypto (e.g., ETH) have higher ones (~10-25%).
• Liquidity Depth: Less liquid assets require larger haircuts to account for slippage during a forced sale.
• Oracle Reliability: Assets with less robust price feeds may get a higher discount.
• Smart Contract Risk: Assets from newer or unaudited protocols may be penalized. The formula is: Borrowing Power = (Collateral Market Value) * (1 - Haircut %).

These are two sides of the same risk coin, often confused.

• Haircut is applied to the collateral's value to determine how much debt it can secure. A 20% haircut means $100 of ETH can back $80 of debt.
• Loan-to-Value (LTV) is the inverse ratio, set against the loan amount. An 80% LTV means an $80 loan requires $100 of collateral. The relationship is: Maximum LTV = 1 - Haircut. A protocol may set a 75% LTV, which is equivalent to a 25% collateral haircut. Haircuts are more common in institutional finance, while LTV is prevalent in DeFi interfaces.

A properly set haircut mitigates specific risks:

• Black Swan Events: If an asset's price drops faster and deeper than the haircut buffer, positions become undercollateralized before liquidation can occur.
• Liquidity Crunch: If many positions are liquidated simultaneously, market liquidity can evaporate, causing slippage that exceeds the haircut's protection.
• Oracle Manipulation: A malicious actor could artificially inflate an oracle price, making undercollateralized positions appear healthy, thereby defeating the haircut's purpose. These scenarios highlight that haircuts are probabilistic buffers, not absolute guarantees.

In Decentralized Finance (DeFi):

• MakerDAO: Uses a Collateralization Ratio (e.g., 150% for ETH) which implies a ~33% haircut. Riskier assets like LINK have higher ratios (165%+).
• Aave & Compound: Use Loan-to-Value (LTV) ratios, which are the inverse of a haircut.

In Traditional Finance (TradFi):

• Repo Markets: The standard mechanism. A treasury bond might have a 2% haircut, while a corporate bond might have 5-10%.
• Central Clearing Counterparties (CCPs): Apply haircuts to margin posted by members based on asset risk profiles.

Advanced protocols use dynamic haircuts that adjust based on market conditions.

• Volatility-Based: Haircuts increase when market volatility (measured by metrics like BVOL) rises above a threshold.
• Utilization-Based: As more of a particular collateral type is used within a system, its haircut may increase to discourage concentration risk.
• Governance Updates: DAOs or protocol governors can vote to change haircuts for assets in response to emerging risks, new audit findings, or changing market structures. This makes risk parameter management a critical, ongoing governance activity.

The Loan-to-Value (LTV) ratio is the maximum amount that can be borrowed against a specific collateral asset, expressed as a percentage of the collateral's value. It is the primary risk parameter set before a haircut is applied.

• Example: An LTV of 75% on $100 of ETH allows a $75 loan.
• The collateral haircut is the inverse: Haircut = 1 - LTV. An LTV of 75% implies a 25% haircut.
• This ratio is the first line of defense against default.

Liquidation is the forced sale of a borrower's collateral to repay an undercollateralized loan. It is triggered when the Loan-to-Value (LTV) exceeds a predefined liquidation threshold.

• The haircut creates a buffer (e.g., 15%) between the initial LTV and the liquidation LTV to protect the protocol.
• Liquidators are incentivized to buy the collateral at a discount (liquidation penalty), which is often related to the haircut size.
• This process ensures the lending protocol remains solvent.

In protocols like Compound and Aave, the collateral factor is synonymous with the Loan-to-Value (LTV) ratio. It determines how much debt can be issued against a specific collateral type.

• A collateral factor of 0.8 (80%) means $100 of collateral can secure $80 of debt.
• The implied haircut is 20%, representing the portion of the collateral's value that cannot be borrowed against.
• This factor is asset-specific and is a core parameter in decentralized money markets.

Overcollateralization is the practice of pledging collateral worth more than the loan value. It is the fundamental security model for most DeFi lending and stablecoins.

• The collateral haircut enforces overcollateralization by discounting the asset's value for loan calculations.
• For example, a 30% haircut on $100 of BTC means it is only valued at $70 for borrowing purposes, ensuring a $70 loan is backed by $100 of assets.
• This buffer protects against market volatility and liquidation delays.

A risk parameter is a configurable variable in a financial protocol that manages exposure to asset volatility, liquidity, and default. Collateral haircuts are a primary type of risk parameter.

• Other key parameters include LTV ratios, liquidation thresholds, liquidation penalties, and reserve factors.
• These are often set by decentralized governance (e.g., token holders) based on asset risk assessments.
• Adjusting these parameters is a core function of protocol management and treasury oversight.

An oracle is a service that provides external data (like asset prices) to a blockchain. Accurate price feeds are critical for calculating collateral values and triggering liquidations.

• The collateral haircut is applied to the oracle price to determine the borrowable value.
• Oracle manipulation or failure can render haircuts ineffective, leading to protocol insolvency.
• Decentralized oracle networks like Chainlink are used to provide robust, tamper-resistant price data for these calculations.