Full Case Study
Project breakdown
Key Takeaway
A DeFi lending protocol launched on Arbitrum with pre-launch economics modeling catching a liquidation cascade scenario. The fix: tiered liquidation bonuses (5% at HF 0.9–1.0, 8% at HF 0.8–0.9, 12% at HF below 0.8) + minimum CR raised to 150%. At 12 months: $48M TVL, zero bad debt, 8.4% supply APY. $278,000 development cost.
The Pre-Launch Finding That Changed Everything
Our protocol economics engagement ran the standard stress test: what happens if ETH drops 40% in 4 hours (the March 2020 scenario)? Finding: At the proposed 140% minimum collateralization ratio, a 40% ETH price drop over 4 hours would generate $3.2M in undercollateralized positions before the liquidation engine could process them — because the liquidation bonus (flat 5%) did not incentivize liquidators to prioritize the most underwater positions.
The fix: tiered liquidation bonus (5% at HF 0.9–1.0, 8% at HF 0.8–0.9, 12% at HF below 0.8) + minimum CR raised to 150%. Re-ran the March 2020 simulation: all positions liquidated before insolvency. Zero bad debt.
Technical Implementation
Collateral: ETH and WBTC (blue-chip only at launch)
Debt asset: USDC
Minimum CR: 150% for ETH, 160% for WBTC
Liquidation bonus: Tiered (5%, 8%, 12%)
Interest rate model: Two-slope with 80% utilization kink
Oracle: Chainlink with 1-hour staleness check and 15% hourly deviation circuit breaker
Chain: Arbitrum One
12-Month Results
| Metric | Value |
|---|---|
| Total Value Locked (TVL) | $48.2M |
| Supply APY (USDC) | 8.4% |
| Borrow APR (USDC, ETH collateral) | 11.2% |
| Bad debt events | 0 |
| Liquidations processed | 847 |
| Liquidation success rate | 100% |
| Protocol revenue (12 months) | $1.84M |
Development cost: $278,000 (economics modeling + smart contracts + frontend + audit). Timeline: 22 weeks.
