Aave's risk management framework continues to evolve as LlamaRisk, the protocol's independent risk steward, has recommended parameter adjustments across two instances of Aave V3. The recommendations reflect a methodical approach to capacity management, balancing protocol growth with prudent risk oversight. These changes underscore how major lending protocols must dynamically recalibrate their operational parameters as user behavior and market conditions shift on-chain.
The most significant proposal targets USDm on Aave V3 MegaETH, where borrowed amounts have reached saturation at the current 280 million ceiling. LlamaRisk identified structural demand exceeding available borrow capacity, with top positions concentrated among users depositing USDe collateral in e-mode configurations. These leveraged stablecoin positions cluster around health factors of 1.03, indicating users are comfortable operating near liquidation thresholds in a correlated asset environment. This tight distribution actually reduces liquidation risk despite apparent leverage, since both the collateral and borrowed asset move in tandem. Given this sustained demand pattern, LlamaRisk recommends raising the borrow ceiling to 450 million—a 1.6x increase—while maintaining the supply cap at 1 billion, which still carries substantial headroom at 66.7% utilization.
Simultaneously, the risk stewards proposed tightening the supply cap for tETH on Aave V3 Prime, reducing it from 45,000 to 25,000. This asset exhibits the opposite problem: only 16,688 units are currently deposited, leaving substantial idle capacity that serves no productive function and creates unnecessary operational complexity. The new cap maintains a 1.5x buffer above actual deposits, preserving growth room while eliminating excess slack. These opposing adjustments—one expanding, one contracting—demonstrate that effective risk governance requires continuous calibration rather than static policy.
Implementation will proceed through Aave's established Risk Steward process, which enables timely parameter adjustments without requiring full governance votes for routine rebalancing. This streamlined approach has become essential as multi-instance deployments create dozens of individual asset configurations requiring regular attention. The precedent LlamaRisk establishes here—data-driven recommendations backed by behavioral analysis and utilization metrics—will likely inform future parameter management across DeFi protocols seeking sustainable growth frameworks.