Aave Proposes Sweeping Cuts to 50 Underperforming Assets and Six Networks

The protocol targets 50 low-adoption asset reserves and full deployments on Sonic, Scroll, zkSync, Metis, Soneium, and Aptos. All measures require DAO approval; nothing executes until on-chain voting concludes.

Reserve deprecation mechanics

The proposal splits into two tracks. The first retires 50 low-adoption reserves and 21 matured Pendle principal tokens across 11 deployments. These positions hold $85.3 million in supply and $11.5 million in debt. Targeted assets include low-use collateral, older bridged tokens, and duplicates where native alternatives now exist—bridged USDC variants would disappear from markets where native USDC is already listed.

The largest single exposure sits on Ethereum: the FBTC and eBTC wrappers. They carry approximately $16.3 million in supply against roughly $63,000 in borrowing. The collateral demand curve never materialized. That is a yield compression signature we have seen repeatedly in Bitcoin-adjacent wrappers that lack native utility beyond passive collateralization.

For affected reserves, the protocol will freeze markets and reduce supply and borrowing caps to one unit. Existing positions remain open, but new deposits, fresh borrows, and additional collateral lockups are blocked. Markets with outstanding loans will see the reserve factor raised, redirecting interest accrual to the Aave treasury and compressing supplier APY. Whole-market closures escalate this further: a 99% reserve factor combined with a 5% base borrowing rate to push borrowers toward repayment and depositors toward withdrawal. Risk stewards retain authority to raise borrowing rates further if positions persist and to reduce liquidation thresholds gradually as residual exposure contracts.

Chain wind-down rationale

The six full chain deployments hold $12.8 million in supply and $4.1 million in debt combined. Sonic carries the heaviest weight at $7.6 million supplied and $2.7 million borrowed; deposits fell 74% over six months. Scroll dropped 86% to $2.2 million; zkSync fell 88% to $844,000. Metis sits at $297,000, Soneium at $173,000. Aptos liquidity contracted 94% over six months, leaving $1.7 million supplied and $719,000 borrowed.

Risk provider LlamaRisk concluded these deployments fail to generate revenue sufficient to cover price feed maintenance, monitoring infrastructure, and operational overhead. That is a straightforward negative unit economics verdict. When the cost of keeping an oracle alive exceeds the fees the market produces, the market is a liability.

Once positions largely unwind, Aave intends to swap live price feeds for fixed-price oracles before retiring the six markets entirely. The proposal remains at the Aave Request for Comment stage.

Risk-to-reward read

We see a protocol pruning tail assets that failed to attract sustainable liquidity. From a capital allocator standpoint, the risk vector is migration: suppliers and borrowers must exit affected markets or face yield compression through elevated reserve factors. The Ethereum wrappers present the most acute case—$16.3 million in supply cannot be withdrawn without depressing on-chain exit liquidity. No systemic insolvency signal here; this is orderly deprecation, not emergency unwinding. The verdict: passive. Watch the vote.