Aave risk manager LlamaRisk has proposed winding down about $1.33 billion of collateral in Aave V3 Core, shifting the Ethereum lending market toward established, liquid assets while giving existing borrowers time to exit.
The governance proposal puts the debt backed by those assets at $1.03 billion as of Nov. 9. It remains a discussion-stage plan, not an approved market change; initial measures would require approval and a subsequent onchain governance proposal.
LlamaRisk argues that Core's pooled borrowing rates leave suppliers exposed to complex collateral without compensating them for its additional risks. Its proposed exclusions span restaking, leveraged wrappers, derivatives-backed strategies and custom bridges, alongside issuers judged insufficiently resourced and reserves with limited demand.
rsETH, osETH, LBTC, tBTC and rETH account for most of the collateral targeted. Smaller reserves in scope are eBTC, BTC.b, tETH, PT-srUSDe-22OCT2026 and LDO. The plan also covers mUSD, which is borrowable but not used as collateral.
Affected Core Collateral and the Debt It Backs
| Subject | Metric | Value |
| rsETH | Collateral | $661.8M |
| rsETH | Debt backed | $605.7M |
| osETH | Collateral | $279.4M |
| osETH | Debt backed | $256.6M |
| LBTC | Collateral | $177.4M |
| LBTC | Debt backed | $72.1M |
| tBTC | Collateral | $117.4M |
| tBTC | Debt backed | $52.9M |
| rETH | Collateral | $81.8M |
| rETH | Debt backed | $35.4M |
Source: data_source (governance.aave.com). Period: 9 November 2026. Source: LlamaRisk, Nov. 9, 2026. Five largest affected collateral reserves, not the full scope. Debt is allocated proportionally to each borrower's collateral mix. (source)
The proposal extends an earlier low-adoption wind-down into larger Core positions. It would retain collateral including WETH, wstETH, weETH, cbETH, WBTC and cbBTC, as well as USDe, sUSDe and syrupUSDT.
Borrowers Would Get Time to Exit
The initial measures would freeze affected reserves and set their loan-to-value limits to zero, including in efficiency modes that allow higher borrowing against correlated assets. That would stop new supply and new borrowing against the collateral, without immediately liquidating existing loans. Holders could still repay, withdraw or migrate.
rsETH and tBTC are already frozen, according to LlamaRisk. The plan would add wstETH, weETH and cbETH to rsETH's efficiency-mode category so borrowers could swap collateral without first repaying their WETH debt in full. LlamaRisk would work with issuers and large borrowers, with no further measures while an asset's exposure declines.
Suppliers would also face lower interest income on CRV, tBTC, rETH, crvUSD, FBTC, UNI and mUSD. Their reserve factors—the share of borrower interest directed to the DAO—would rise to 99%, without an initial increase in borrow rates.
If migration stalls, later controls could gradually lower liquidation thresholds for LBTC, tBTC and rETH. For rsETH and osETH borrowing loops, which LlamaRisk puts at 12 to 14 times leverage, higher WETH borrowing rates would be a last resort, paired with rebates for borrowers holding neither asset. LlamaRisk warns that cutting liquidation thresholds by two to five percentage points could make most of the collateral in those loops liquidatable at once.
V4 would provide another destination: LlamaRisk expects LBTC to continue through a dedicated borrowing module, or spoke, on V4's Ethereum Core Hub. As The Defiant previously reported, V4 can serve separate borrowing markets from shared liquidity. Its documentation describes local risk controls and collateral-dependent interest premiums.
Following approval, LlamaRisk plans to submit the initial changes as an onchain proposal and report on unwinding monthly. Any follow-up steps would be announced on the forum beforehand.