Aave Core Repositioning Proposal Targets $1.33 Billion in Collateral

Key Facts

LlamaRisk published a proposal on the Aave governance forum on November 10, 2026, at 7:33 p.m., to reposition the Aave V3 Core market around established liquid collateral. The collateral in scope is worth $1.33 billion and backs $1.03 billion in debt, according to November 9 data. The proposal is open for discussion at the ARFC stage and does not call for a fully completed, immediate exit from the assets. The plan proposes freezing the reserves in scope, halting new supply and reducing the loan-to-value ratio to zero. Existing positions would remain open, with borrowers able to repay, withdraw collateral or migrate their positions.

The proposal covers rsETH, osETH, LBTC, tBTC and rETH on Core, along with a group of less-exposed reserves. November 9 data show that $661.8 million in rsETH collateral backs $605.7 million in WETH debt. The $279.4 million in osETH collateral backs $256.6 million in WETH debt. WETH debt backed by rsETH and osETH totals about $862 million, most of it arising from leveraged borrowing loops in Ethereum-related E-Mode categories. By contrast, LBTC, tBTC and rETH back $160 million in debt, mostly in stablecoins, with debt attributed to each asset in proportion to the borrower’s collateral mix.

LlamaRisk links its proposal to collateral risks that depend on infrastructure and services beyond the token itself, and aims to reduce exposures where leverage loops overlap with asset risks. The company says rsETH and osETH loops operate at leverage of 12 to 14 times. It warns that lowering the liquidation threshold by 2 to 5 points could make most collateral in these positions liquidatable at once. The plan therefore does not prioritize lowering liquidation thresholds for these positions. During the transition, borrowers could repay debt, withdraw collateral or migrate positions, while raising the WETH borrow rate remains a last resort if the transition fails to reduce exposure.

The specifications provide special treatment for frozen reserves that remain borrowable and carry outstanding debt. If debt exceeds $20,000, the proposal recommends raising the reserve factor to 99%. This mechanism shows that halting new supply does not necessarily mean all new borrowing from a frozen, borrowable reserve will stop. For LBTC, tBTC and rETH, liquidation thresholds could be lowered if exposure does not decline after outreach and an orderly transition. Each step would be capped at 2 percentage points, with the collateral that becomes liquidatable kept below $2 million or 5% of the asset’s remaining collateral.

LlamaRisk estimates that debt backed by the collateral in scope generates about $3.7 million a year in gross DAO revenue. The estimate is based on average realized borrow rates during the three months before the proposal and does not account for capital migration or inflows that could offset existing borrowing. The company places the plan in the context of earlier work to disable low-utilization assets, covering 71 reserves across 11 deployments, as well as 6 full markets with total supply of about $98 million. LlamaRisk cites those measures as an implementation precedent, while the current proposal covers a larger amount of collateral and debt. The assessment balances reducing exposure with accounting for revenue generated by outstanding debt.

LlamaRisk presents Aave V4’s architecture as a potential path for hosting assets that may not remain within the proposed focus of Core. In the design described, liquidity is concentrated in hubs, while collateral sits in separate spokes with their own supply and withdrawal caps. A collateralRisk factor also adds a premium to each borrower’s interest rate based on the risk of the collateral backing the loan. This separation is intended to address liquidity and collateral risk through different modules. The proposal presents this architecture as an option for a broader range of assets, while keeping the Core market focused on established liquid collateral.

Bridge risks are part of the backdrop to the discussion. In an update on the April 18, 2026, incident, Aave said a vulnerability enabled the unauthorized minting of rsETH from the LayerZero adapter on Ethereum. The update said the incident broke the collateral relationship between assets locked on Ethereum and versions on other networks. It estimated that 152,577 rsETH had been extracted, equivalent to an initial shortfall of about 163,183 ETH at a conversion ratio of 1.0696 rsETH per ETH. In April 2025, the BGD team had also recommended a precautionary freeze of rsETH across Aave V3 Core, Prime, Arbitrum and Base markets after identifying a flaw in Kelp DAO’s contract architecture, before later reporting that the freeze had been lifted and the asset had resumed operations. LlamaRisk includes reliance on external infrastructure among the layers of risk its proposal seeks to reduce.

After approval, LlamaRisk plans to put the initial measures into an onchain AIP, then contact asset issuers and the largest borrowers and publish monthly updates on progress in reducing exposure. If outreach and an orderly transition do not reduce exposure to LBTC, tBTC and rETH, liquidation thresholds could be lowered gradually within the limits set out in the proposal. For rsETH and osETH loops, the variable base WETH borrow rate could be raised by up to 100 basis points, in 50-basis-point steps and no more than once a month. The proposal says this increase must remain a last resort after a transition stalls, and could be accompanied by discounts for borrowers who do not hold either asset. EL7 data recorded AAVE closing at $131.64 on November 9 and $123.64 on November 10.