Bad Debt

Bad debt occurs when a loan's outstanding value permanently exceeds the value of its liquidatable collateral. This creates an accounting deficit on the protocol's balance sheet. Key drivers include:

• Extreme market volatility causing collateral value to plummet below liquidation thresholds before keepers can act.
• Oracle failure or manipulation providing incorrect price feeds.
• Liquidity black holes where collateral assets cannot be sold even at a steep discount.

Unresolved bad debt represents a direct liability for the lending protocol. It threatens the solvency of the protocol's treasury or its designated insurance fund. If the deficit grows large enough, it can lead to a bank run where depositors rush to withdraw funds, fearing losses. Protocols often cover deficits via:

• Treasury reserves (e.g., MakerDAO's Surplus Buffer).
• Insurance from staked tokens (e.g., MKR token auctions).
• Socialized losses across remaining depositors (a last-resort mechanism).

Bad debt in one major protocol can propagate risk across the ecosystem. This contagion risk operates through interconnected channels:

• Collateral rehypothecation: The same asset (e.g., stETH) used as collateral across multiple protocols can trigger cascading liquidations.
• Liquidity provider (LP) token depeg: Bad debt causing a protocol's native token to collapse can erode the value of LP positions in DEX pools.
• Loss of confidence that spreads to protocols with similar design or shared asset exposure.

Unlike traditional finance, DeFi bad debt is a transparent, immutable on-chain record. The deficit is visible in the protocol's smart contract state and public ledger. This transparency allows for real-time risk assessment but also means the problem cannot be hidden or easily restructured. Resolution mechanisms, like debt auctions, are also executed on-chain, creating a public record of the recovery process and any final losses.

Significant bad debt events are primary catalysts for protocol governance actions. They force reevaluation of core risk parameters to prevent recurrence. Common post-event changes include:

• Increasing collateralization ratios (e.g., raising Loan-to-Value (LTV) requirements).
• Adding new collateral types or delisting risky assets.
• Enhancing oracle security (e.g., adding delay or using multiple sources).
• Designing new liquidation mechanisms (e.g., Dutch auctions, soft liquidations).

Protocols monitor loan health using a Health Factor (HF), calculated as (Collateral Value * Liquidation Threshold) / Borrowed Value. An HF ≤ 1 indicates an undercollateralized position. For the overall system, bad debt modules or dashboards aggregate the sum of all positions where the debt exceeds the realizable collateral value, even after liquidation attempts. This metric is crucial for risk analysts and auditors.

A price oracle provides the on-chain price data used to determine a loan's collateralization ratio. Failure modes include:

• Oracle lag or staleness during extreme volatility, showing outdated prices.
• Oracle manipulation via flash loans or market attacks to feed incorrect data.
• Dependency on a single, centralized data source that goes offline.

When the oracle reports a higher collateral value than the real market price, liquidations are delayed, allowing debt to become undercollateralized.

Even with accurate price data, the liquidation engine can fail to close an undercollateralized position. Common causes:

• Insufficient liquidation incentives: The liquidation bonus is too small to cover gas costs and slippage for liquidators.
• Network congestion: High gas fees during market crashes make liquidations economically unviable.
• Design flaws: Complex or slow liquidation logic that cannot execute before the position worsens.

This results in 'zombie' loans where the debt exceeds the realizable value of the collateral.

The assumed value of collateral can be unrealizable due to market depth. Key issues:

• Low-liquidity collateral assets: Attempting to liquidate a large position crashes the market price (slippage).
• Wrapped or synthetic assets de-pegging from their underlying value.
• Concentrated positions in a single asset that cannot be sold without significant price impact.

The liquidation may only recover a fraction of the debt, leaving a shortfall.

Direct attacks on lending protocol code can create bad debt by bypassing safeguards.

• Flash loan attacks that manipulate collateral valuation or drain liquidity pools in a single transaction.
• Logic bugs in collateral accounting or interest rate models.
• Governance attacks seizing control to alter critical parameters (e.g., setting collateral factor to 100%).

These exploits can instantly render many positions undercollateralized with no chance for orderly liquidation.

Rapid, multi-sigma price movements can overwhelm all protocol safeguards simultaneously.

• Collateral asset price crashes of 50%+ within a single block or hour.
• Correlated asset collapse where diversified collateral pools all lose value together.
• Cascading liquidations that depress prices further, creating a death spiral.

In these scenarios, the speed and magnitude of the move outpace oracle updates, liquidation capacity, and market liquidity.

Improperly set risk parameters by protocol governance can systematically induce bad debt.

• Excessively high Loan-to-Value (LTV) ratios for volatile assets.
• Incorrectly listing immature or risky assets as collateral.
• Setting liquidation penalties or incentives too low to ensure keeper activity.
• Failing to adjust parameters in response to changing market conditions.

This represents a failure in ongoing risk management rather than a one-time event.

The automated process of selling a borrower's collateral to repay their loan when its health factor falls below a threshold. This is the primary mechanism for preventing bad debt. A failed or delayed liquidation can directly result in a bad debt position.

• Triggered by: Price volatility, collateral depreciation, or debt appreciation.
• Executed by: Keepers or liquidators who earn a fee.
• Risk: If liquidation is impossible (e.g., no liquidators, frozen market), the protocol incurs bad debt.

A numerical metric representing the safety of a loan, calculated as the ratio of the collateral value to the borrowed value. It is the core risk parameter for lending protocols.

• Formula: HF = (Collateral Value * Collateral Factor) / Borrowed Value
• Safe: HF > 1.0 (e.g., 1.5).
• At Risk: HF <= 1.0, triggering liquidation.
• A Health Factor dropping to or below 1.0 without successful liquidation creates bad debt.

A state where the market value of a borrower's collateral is less than the value of their outstanding loan. This is the direct cause of bad debt.

• Causes: Extreme market crashes, oracle failure, or collateral asset de-pegging.
• Protocol Exposure: The difference between the loan value and the collateral value becomes a loss for the protocol's lenders.
• Example: Borrowing $100 of stablecoins with $150 of ETH collateral. If ETH crashes to $80, the loan is undercollateralized by $20.

The maximum Loan-to-Value (LTV) ratio at which a position can be opened. It is a risk parameter set by the protocol for each collateral asset.

• Function: Defines the buffer zone before a position becomes eligible for liquidation.
• Relation to HF: A position's Health Factor falls below 1.0 when its actual LTV exceeds the liquidation threshold.
• Risk Management: A lower threshold (e.g., 75% for volatile assets) provides a larger safety cushion against bad debt.

A service that provides external data (primarily asset prices) to a blockchain smart contract. Reliable oracles are critical for preventing bad debt.

• Role: Feeds current market prices to calculate Health Factors and trigger liquidations.
• Failure Modes: Price feed lag, manipulation (oracle attack), or downtime can cause:
  • Delayed liquidations.
  • Incorrect HF calculations.
  • Undercollateralized positions going undetected.

A systemic state where a protocol's total liabilities (user deposits) exceed its total assets (collateral + reserves). Widespread bad debt is a primary driver of insolvency.

• Mechanism: If bad debt exceeds a protocol's insurance fund or reserve capital, losses are socialized among all lenders, often through mechanisms like debt tokenization or loss redistribution.
• Outcome: Lenders may be unable to withdraw the full value of their deposits.
• Example: The 2022 collapse of several lending protocols was precipitated by massive bad debt from market crashes.