How Lending Protocols Operate
Decentralized Finance (DeFi) lending protocols enable users to borrow and lend digital assets without intermediaries. Users deposit collateral tokens to borrow other assets, with oracles continuously assessing collateral values and loan-to-value (LTV) ratios.
Key metrics:
- Collateral Value:
Price (USD) × Quantity - LTV:
Borrowed Value ÷ Collateral Value - Liquidation Threshold: Predefined LTV level triggering forced repayments.
Example Scenario:
Alice deposits 10 ETH ($1,000/ETH) and borrows 7,500 USDT (75% LTV). If ETH drops to $937.5 (80% LTV threshold):
- A liquidator repays 3,750 USDT (50% debt).
- Receives 4.2 ETH (105% value + 5% bonus).
- Alice’s new LTV: 69%, liquidation price drops to $808.
Why Liquidations Fail
1. Market Impact
Large sell orders depress prices. If selling 4.2 ETH lowers average price below $890 (47.5 USD impact), liquidation becomes unprofitable.
2. Network Congestion
- Ethereum: High gas fees may outweigh rewards.
- Solana: Spam transactions during mass liquidations risk network outages (e.g., Solend whale incident).
Strengthening Protocol Resilience
Solution 1: Optimize Maximum Loan Size
Adjust collateral valuation to account for market impact: V = q × (p - I(q))
Where I(q) = estimated price impact for quantity q.
Solution 2: Dynamic Liquidation Rewards
Formula: Reward = max(a, a/2 + I(q)/p)
a= base reward (e.g., 5%).- Compensates for slippage during large liquidations.
Solution 3: Risk-Based Interest Rates
Charge borrowers extra when:
- Market impact approaches liquidation bonus.
- Price nears liquidation threshold.
Formula: i = C × (I(q)/p × a) × (p - pl)/σ
Outcome: Incentivizes proactive position management.
Estimating Market Impact
Empirical data shows market impact is linear: I(q) = q / 2l
l= liquidity intensity (order book depth).
Example: Binance BTC/USDT liquidity intensity ≈ 2 BTC/USDT → Slope = 0.25 USDT/BTC.
Liquidity Oracle Proposals
Option A: Decentralized Aggregation
- Pros: Transparent, manipulation-resistant.
- Cons: Scalability challenges tracking multiple exchanges.
Option B: Publisher-Aggregated Data
- Pros: Simplified implementation.
- Cons: Less transparency; requires uniform exchange coverage.
Recommended Features:
- Track off-chain liquidity (CEXs, futures).
- Publish buy/sell liquidity separately.
FAQs
1. What triggers liquidation?
When LTV exceeds the protocol’s predefined threshold (e.g., 80% for ETH).
2. How do dynamic rewards improve stability?
They adjust compensation for market impact, ensuring liquidations remain profitable.
3. Why exclude on-chain liquidity?
Susceptible to manipulation (e.g., flash loans). Off-chain data is more reliable.
4. Can risk-based interest prevent bad debt?
Yes, by charging higher rates for riskier positions pre-liquidation.