The LlamaRisk risk-management team published a proposal to restructure Aave V3’s Ethereum Core market. The initiative would restrict the use of several crypto assets as collateral, concentrating liquidity in established assets and reducing risks for the decentralized-finance protocol.

As of Nov. 9, 2026, the proposed changes covered $1.33 billion in collateral backing about $1.03 billion in debt positions.

LlamaRisk said Aave V3 Core pools liquidity across different assets, including collateral tokens that depend on restaking, leverage, cross-chain bridges and third-party infrastructure. According to the proposal’s authors, those dependencies create additional risks for liquidity providers, who do not always receive adequate compensation.

Assets covered by the proposal

The proposal targets the largest positions in rsETH, osETH, LBTC, tBTC and rETH. It also covers eBTC, BTC.b, tETH, LDO, mUSD and PT-srUSDe-22OCT2026.

LlamaRisk proposed freezing the relevant reserves by prohibiting new deposits and borrowing against those assets and reducing their loan-to-value ratios to 0%. Existing positions would remain open, allowing users to repay debt, withdraw assets or move collateral into other categories without immediate forced liquidations.

For certain frozen reserves with more than $20,000 in debt, the proposal also calls for increasing the reserve factor to 99%.

The largest concentration of risk is in rsETH and osETH, which account for about $862 million in WETH debt. LlamaRisk said the associated strategies use leverage of about 12 to 14 times.

On April 18, 2026, the restaking protocol KelpDAO suffered an attack that caused about $293 million in losses. The attackers exploited a cross-chain bridge vulnerability and minted unbacked rsETH tokens, some of which were used as collateral on Aave, triggering a liquidity crunch and capital outflows from the protocol.

DeFi participants formed the DeFi United coalition to address a collateral shortfall of about $250 million and developed a plan to restore rsETH’s value. KelpDAO later reported completing a five-week token-recovery process.

Migration to Aave V4

The restructuring would gradually move more complex collateral to Aave V4, where bespoke additions to borrowing rates would allow risks to be assessed individually. Aave Labs announced the launch of V4 on the Ethereum mainnet in late March 2026.

Under the proposal, V3 Core would focus on liquid assets including WETH, wstETH, weETH, cbETH, WBTC, cbBTC, USDC, USDT, GHO and DAI.

If borrowers do not voluntarily reduce their positions, LlamaRisk could gradually lower liquidation thresholds for LBTC, tBTC and rETH. For strategies involving rsETH and osETH, a last-resort measure could involve raising WETH borrowing rates while compensating borrowers who do not use those assets.

LlamaRisk estimated that the affected debt positions generate about $3.7 million in annual revenue for the Aave DAO. That estimate does not account for possible capital reallocation.

Projects object to proposed restrictions

StakeWise founder Dmitri Tsumak asked for the proposed freeze of osETH to receive a separate vote, citing the protocol’s security record.

Rocket Pool representative Darren Langley also asked for rETH to be removed from the proposal and assessed separately under Aave’s risk framework.

“We support the aim of focusing Core on established, liquid collateral and removing risk. We believe rETH falls on the established side of that line, and we ask that it be removed from this ARFC and assessed on its own against the Aave Risk Framework.“

The proposal remains at the Aave Request for Comments, or ARFC, discussion stage, meaning the restrictions have not been introduced. If approved, the initial changes would proceed to an Aave Improvement Proposal, or AIP, and the team plans to report monthly on reductions in the affected positions.

Community response

Aave founder Stani Kulechov supported discussing the initiative. He said it would sharpen V3 Core’s risk profile by removing exposure to restaking assets, leveraged wrapping strategies, issuers that fall short of Aave’s operational and security standards, and low-demand reserves that no longer justify maintenance.

On Nov. 10, Kulechov described the potential scope as $1.52 billion in collateral backing $1.17 billion in debt and said some affected assets could move to isolated V4 markets with more granular risk controls.

An analyst using the name anze also supported separating the risks, citing withdrawal problems for ETH lenders and sharp increases in stablecoin borrowing rates after the rsETH incident.

“Keeping V3 Core simple and pushing the exotic stuff to V4 seems like the right decision. Lenders get a safer pool, and loopers can still use their markets on V4,” anze wrote.

A user identified as Ash criticized V3 Core’s shared-liquidity model, arguing that multiple liquid restaking tokens can carry correlated risks. Ash said isolating those positions in V4 would allow each collateral category to be assessed separately and reduce the risk of problems spreading between protocol segments.

“Isolation isn’t a downgrade for those assets. It’s an admission that one giant shared pool is a contagion design,” Ash emphasized.

Source: Incrypted