Aave proposes comprehensive reform of the risk framework and removal of low-utilization reserves

Aave proposes to remove low-liquidity reserves and retire six blockchain deployments

Aave proposed an agreement-wide risk management reform that plans to remove low-adoption asset reserves and retire six blockchain deployments, a move involving approximately $113.7 million in assets.

LlamaRisk reviewed inactive reserves, cross-chain bridge assets, and expired Pendle tokens to reduce maintenance costs and operational complexity.

Governance changes will freeze affected reserves, tighten lending parameters, and expand risk standards across the process in future asset launches and reviews.

Governance proposal focuses on comprehensive application of risk framework

Aave proposed a governance plan to remove dozens of low-adoption asset reserves and retire six blockchain deployments to strengthen risk management at the protocol level. According to Aave founder Stani Kulechov, the proposal is the first large-scale application of Aave's new risk framework and could affect approximately $113.7 million in supply assets and outstanding debt in multiple deployments.

LlamaRisk prepared the proposal after reviewing the Aave ecosystem against the full process risk criteria proposed in the agreement. The assessment is no longer limited to a single reserve, but rather measures whether each online asset and deployment still meets the operational, liquidity and maintenance requirements under the updated framework.

Under the proposal, Aave plans to remove 50 low-adoption reserves and 21 expired Pendle principal tokens from 11 Aave V3 deployments. In addition, the agreement also intends to phase out deployments on Sonic, Scroll, zkSync, Metis, Soneium and Aptos, and include an additional 25 reserves in the proposed phase-out process.

The reserve removal involved supply assets of approximately US$85.3 million and outstanding debt of approximately US$11.5 million. While six blockchain deployments involve approximately $12.8 million in supply assets and $4.1 million in debt, the scale of the governance proposals is evident despite the relatively small target market.

Targeting inactive assets and high-cost deployments

According to LlamaRisk, the proposal also covers cross-chain bridge assets that duplicate native tokens, as well as expired Pendle principal tokens that no longer generate revenue. In addition, some smaller deployments were selected because agreement revenue no longer covered the costs of maintaining price predictors, clearing infrastructure and ongoing risk monitoring.

If the governance votes to pass the proposal, Aave will freeze the new supply and lending activity of the affected reserves, while reducing both the supply and lending ceilings to one unit. In addition, the agreement will increase the loanable asset reserve factor, making it less attractive for users to maintain positions in markets where tags have been removed.

The entire deployment marked for retirement will face stricter parameter adjustments to speed up the phase-out process. Specifically, the reserve factor will increase to 99%, while a higher base rate will encourage users to close positions and migrate assets to more active Aave deployments.

New Framework Expands Agreement Scope Regulation

This governance proposal stems from the proposed risk framework introduced by Aave in June, which establishes unified standards for asset roll-online, ongoing review and future reserve removal for Aave V3, Aave V4 and Aave Horizon. According to Kulechov, the framework creates a process to continuously evaluate whether assets should remain online to adapt to changes in market conditions and agreement activity.

The framework was launched after the approximately $292 million KelpDAO Cross-Chain Bridge vulnerability incident, which exposed Aave to potential bad debt risk through the deposit of stolen rsETH collateral. Since then, the agreement has expanded the evaluation process from a single asset to a full-process review, aiming to identify operational risks and prevent them from turning into larger governance issues.

In addition, LlamaRisk has submitted a supporting governance proposal for another group of long-term tail reserves that face high Chainlink oracle price risks. The proposal recommends replacing real-time oracle pricing with fixed-price adapters to reduce oracle related risks while improving the consistency of Aave's broader risk management framework.

Conclusion

This proposal reflects Aave's efforts to apply uniform standards to its lending ecosystem while streamlining inactive markets. If approved, the governance changes will reduce operating burdens, remove underutilized reserves, and establish a structured process for future asset launches and agreed decommissioning.