Aave governance has proposed adjusting collateral efficiency metrics for its two largest asset classes, reflecting updated risk analysis grounded in historical volatility data and recent market stress events. The adjustments span Ethereum mainnet, Arbitrum, and Base, with separate parameter modifications coming to the protocol's V4 implementation. Rather than a blanket increase, the proposal demonstrates sophisticated risk calibration, with different thresholds applied based on each asset's empirical behavior and network-specific conditions.
The core methodology underpinning these changes relies on analyzing the 99.9th percentile of worst-case price excursions within one-hour windows—the approximate window for liquidation execution through sequential calls. This stress-testing approach incorporates data from two significant market dislocations: the February 2025 volatility spike and October 2025 price pressures. Ether exhibits substantially higher volatility than Bitcoin, with annualized two-year figures of 68.8% versus 44.4%, a gap that justifies tighter collateral constraints on the ETH family despite broader increases. The maximum hourly drawdown for Ether reaches 11.85%, compared to just 5.09% for Bitcoin, reflecting the structural differences between these assets and their respective liquidation risks.
The specific parameter adjustments reveal careful calibration rather than uniform loosening. WETH moves to 81% loan-to-value and 84% liquidation threshold across three major networks, representing a substantial but bounded increase. Liquid staking derivatives like wstETH and weETH sit one percentage point below their calculated ceilings on mainnet, introducing deliberate conservatism. Bitcoin positions receive even more cautious treatment on Arbitrum, where WBTC tops out at 78% LTV—eight percentage points below Bitcoin's higher maximum—due to acknowledged concerns around market depth, liquidity concentration, and price feed reliability on that network. The cbBTC parameters on Base include a concurrent reduction in liquidation bonus to 6%, tightening the overall risk surface despite higher LTV allowances.
These changes matter because they directly influence capital efficiency for leveraged strategies and collateral stacking, mechanisms that drive substantial borrowing volumes on Aave. Tighter parameters reduce maximum leverage but enhance protocol stability; looser parameters increase yield capture opportunities but extend tail risks. The analysis explicitly acknowledges that Bitcoin thresholds sit deliberately inside protective buffers, suggesting governance's preference for prudence over marginal efficiency gains. As Aave continues scaling across multiple chains and exploring new collateral types, this risk-calibration framework—grounded in quantified stress scenarios rather than intuition—may set a template for how large lending protocols navigate the perpetual tension between accessibility and safety.