Automated Underwriting

The complete, timestamped record of all transactions associated with a specific blockchain address. This is the foundational data layer for assessing financial behavior.

• Key Metrics: Transaction volume, frequency, counterparties, and gas expenditure.
• Analysis Use: Identifies patterns of income (e.g., regular DEX yields), spending habits, and overall on-chain activity level.
• Example: A wallet with a two-year history of consistent DeFi interactions is a stronger candidate than a newly created, inactive wallet.

Real-time data on existing debt positions and collateral locked within lending protocols like Aave, Compound, and MakerDAO.

• Key Metrics: Loan-to-Value (LTV) ratios, health factors, liquidation thresholds, and types of collateral assets.
• Analysis Use: Assesses current leverage, risk of liquidation, and overall debt management. A user with multiple, healthy positions across protocols demonstrates responsible credit usage.
• Protocols: Data is sourced directly from smart contracts of major money markets.

A record of a wallet's engagement with decentralized applications, beyond simple transfers. This includes liquidity provision, staking, yield farming, and governance participation.

• Key Metrics: Total Value Locked (TVL) across protocols, duration of engagements, and yield earned.
• Analysis Use: Demonstrates sophistication, capital commitment, and generates a reputation score. Long-term liquidity providers in established pools are seen as lower-risk, engaged participants.
• Examples: Uniswap V3 LP positions, Curve gauge voting, Convex staking.

Persistent, verifiable identifiers and attestations that build a pseudonymous financial identity. These are not controlled by a central authority.

• Data Sources: Soulbound Tokens (SBTs), attestation platforms like Ethereum Attestation Service (EAS), and decentralized identity protocols (e.g., ENS names with a history).
• Analysis Use: Provides social proof, credential verification (e.g., KYC attestation from a trusted provider), and proof of membership in reputable DAOs or communities.
• Impact: Mitigates the 'cold start' problem for new addresses by allowing them to port a reputation.

A snapshot of all tokens (fungible and non-fungible) held in a wallet across multiple chains and layers, aggregated via indexers.

• Key Metrics: Portfolio diversification, concentration risk, asset volatility profiles, and net worth.
• Analysis Use: Evaluates financial stability and risk tolerance. A diversified portfolio of blue-chip assets is less risky than one concentrated in a single, volatile meme coin.
• Tools: Services like Zerion, Zapper, and DeBank provide aggregated views, which underwriting models can query via APIs.

Records of past credit extended and repaid within decentralized credit markets, such as those facilitated by credit delegation vaults on Aave or standalone credit protocols.

• Key Metrics: Borrowing limits granted, repayment history (on-time/default), total credit utilized, and delegatee performance.
• Analysis Use: Creates an on-chain credit score. A delegatee with a perfect repayment history across multiple loans is a prime candidate for increased credit limits. This is the most direct analog to traditional credit history.
• Protocol Example: Aave V3's Credit Delegation feature.

The foundational model for automated underwriting in DeFi. Borrowers must deposit collateral worth more than the loan value, with the collateral factor or loan-to-value (LTV) ratio algorithmically determining borrowing capacity. Liquidation is triggered automatically if the collateral value falls below a set threshold.

Key Examples:

• Aave & Compound: Use pooled liquidity and dynamic interest rate models.
• MakerDAO: Issues the DAI stablecoin against locked collateral like ETH.

Protocols that extend credit based on assessed borrower risk, requiring less than 100% collateral. They use on-chain reputation, credit delegation, or identity verification to underwrite loans.

Key Examples:

• Goldfinch: Uses Pool Delegates to perform off-chain due diligence on real-world borrowers, with underwriting results stored on-chain.
• Maple Finance: Institutional Pool Delegates underwrite loans to professional trading firms, with terms enforced by smart contracts.

A design that contains risk by isolating assets into separate vaults or markets. The failure of one vault does not impact others, allowing for the automated underwriting of more exotic or volatile collateral types.

Key Examples:

• Euler Finance: Features isolated tier assets and risk-adjusted borrow factors.
• Morpho Blue: Enables permissionless creation of isolated markets with custom oracles, LTV, and liquidation parameters set by the market creator.

Specialized protocols that develop automated frameworks for non-fungible and off-chain assets. This involves unique valuation oracles, liquidity curves, and liquidation mechanisms.

Key Examples:

• NFTfi & Arcade: Use peer-to-peer negotiations or automated offers to underwrite NFT-backed loans.
• Centrifuge: Tokenizes real-world assets (e.g., invoices, royalties) as NFTs and uses asset-specific risk assessment pools for underwriting.

The interest rate model is a core underwriting mechanism. It automatically adjusts borrowing costs based on real-time utilization rates of a liquidity pool, managing supply/demand and risk.

How it works:

• Low utilization: Low rates to encourage borrowing.
• High utilization: Rates rise sharply to incentivize repayment and additional supply, acting as a automated brake on risk.

Automated underwriting is dependent on price oracles (e.g., Chainlink) for real-time asset valuation. Key automated functions include:

• Collateral Value Calculation: Continuously updating loan health.
• Liquidation Triggers: Automatically initiating liquidations when health factor < 1.
• Parameter Updates: Governance can adjust LTV, liquidation thresholds, and bonuses based on oracle-fed market data.

Automated underwriting relies on price oracles (e.g., Chainlink, Pyth) to determine collateral value. Key risks include:

• Oracle failure: If the oracle feed is delayed, halted, or provides incorrect data, loans may be undercollateralized without triggering liquidation.
• Manipulation attacks: An attacker could artificially manipulate the price on a smaller DEX to trigger unfair liquidations or create bad debt.
• Solution: Protocols use time-weighted average prices (TWAPs) and multiple oracle sources to mitigate single-point failures.

The core logic for collateral checks, liquidations, and interest accrual is encoded in immutable smart contracts.

• Code vulnerabilities: Bugs or exploits in the contract logic can lead to the loss of user funds, as seen in historical exploits.
• Upgradability vs. Immutability: Some protocols use proxy patterns for upgrades, introducing centralization risk if admin keys are compromised. Fully immutable contracts cannot be patched if a bug is found.
• Formal verification and extensive auditing are critical to mitigate this risk.

The liquidation mechanism is the primary safety net for overcollateralized loans. Risks include:

• Network congestion: During market volatility, high gas fees can prevent liquidators from executing profitable transactions, allowing bad debt to accumulate.
• Insufficient liquidity: If there isn't enough market depth to absorb the liquidated collateral, it may be sold at a steep discount (slippage), potentially leaving the protocol undercollateralized.
• Liquidation incentives: If the liquidation bonus is too low, it may not attract enough liquidators to keep the system healthy.

The stability of the collateral asset is paramount. High volatility or a depegging event (e.g., a stablecoin losing its peg) can instantly collapse the loan-to-value (LTV) ratio.

• Example: If a protocol accepts a volatile asset as collateral and its price crashes 30% in minutes, loans can become undercollateralized before liquidations occur.
• Protocols mitigate this by setting conservative LTV ratios for volatile assets, using oracle delay (circuit breakers), or only accepting highly liquid, stable assets.

Many automated underwriting protocols are governed by decentralized autonomous organizations (DAOs) that vote on key parameters.

• Risk of malicious proposals: Governance tokens could be concentrated, allowing a small group to pass proposals that harm users (e.g., stealing collateral).
• Parameter misconfiguration: Incorrect settings for interest rates, liquidation thresholds, or fees set via governance can destabilize the entire protocol.
• Timelocks and multisig safeguards are used to prevent instant execution of harmful governance decisions.

DeFi's composability ("money legos") allows protocols to build on each other, but creates interconnected risks.

• Contagion: A failure or exploit in one automated lending protocol can spill over to others that use its tokens as collateral or that share liquidity.
• Example: If a major stablecoin depegs, every protocol using it as collateral could face simultaneous insolvency, triggering a cascade of liquidations.
• This creates systemic risk where the failure of one component threatens the entire DeFi ecosystem.