LlamaRisk has published an Aave Request for Comment (ARFC) proposing to transform Aave V3 Core into a lending market restricted to liquid blue-chip assets. The proposal would freeze eleven asset reserves that currently hold approximately $1.33 billion in collateral and would route broader asset classes to Aave V4 instead.

What LlamaRisk Is Proposing for Aave V3 Core

The ARFC was posted to Aave’s governance forum on November 10, 2026, and frames Aave V3 Core as a market that has drifted from its original mandate. Rather than supporting a wide catalogue of collateral types, LlamaRisk argues that the instance should concentrate exclusively on assets with deep on-chain liquidity and established market depth — what the proposal calls “established liquid collateral.”

Under the proposal, eleven reserves would be frozen, meaning new deposits and new borrows would stop immediately: rsETH, osETH, LBTC, tBTC, rETH, eBTC, BTC.b, tETH, LDO, mUSD, and PT-srUSDe. Existing positions would not be liquidated by governance action; holders would instead migrate voluntarily or repay their debt over time.

The assets retained on V3 Core cover the most liquid layer of DeFi collateral: WETH, wstETH, weETH, cbETH, WBTC, cbBTC, USDC, USDT, GHO, DAI, XAUt, AAVE, LINK, USDe, sUSDe, and syrupUSDT. According to the proposal, these assets share a common trait: reliable exit liquidity and oracle depth that allow the protocol to liquidate positions without excessive slippage under stress.

According to the governance forum post, the targeted deprecated assets held approximately $1.33 billion in collateral backing roughly $1.03 billion in debt as of November 9, 2026. The post does not state a retrieval time, a USD pricing basis, or a reserve-aggregation methodology, so the figures should be read as proposal estimates rather than independently verified mark-to-market values.

How This Fits the Broader Deprecation Arc

This ARFC is the third phase of a systematic clean-up of the V3 Core market. Two prior deprecation rounds already removed 71 low-adoption reserves, reducing supply by roughly $98 million, according to the governance forum post. This round is different in scale: the reserves being targeted are not marginal positions but active markets, with rsETH alone carrying $661.8 million in collateral — the single largest deprecated position in the proposal.

LlamaRisk’s parallel blue-chip LTV proposal has already been implemented on V3 Core, with WETH set at an 81% loan-to-value ratio and an 84% liquidation threshold, and both WBTC and cbBTC at an 81% loan-to-value ratio and an 85% liquidation threshold, as LlamaRisk noted on X. The current ARFC is the complement to that parameter work: having raised borrowing power for retained assets, the proposal now seeks to clear out the assets that do not meet that liquidity bar.

Why a Liquid Blue-Chip Focus Matters for Aave

Liquidity Risk in Long-Tail Collateral

According to the proposal, the core risk argument for the freeze centers on liquidity concentration. Many of the targeted assets — restaking tokens such as rsETH and osETH, and wrapped Bitcoin variants such as LBTC, eBTC, and BTC.b — carry correlated depegging risk. Under adverse conditions, on-chain liquidity for these assets can thin quickly, extending liquidation windows and increasing the protocol’s exposure to bad debt.

As risk considerations, restaking tokens may also carry smart-contract and collateral-depeg risk, while wrapped Bitcoin variants, including BTC.b, may add bridge or custody and smart-contract risk; oracle reliability and liquidation execution remain relevant across both collateral classes. The proposal does not separately assess these risks for each reserve.

The rsETH and osETH positions also present a structural loop risk flagged in the proposal: users borrow WETH against restaked ETH collateral, compounding exposure. LlamaRisk identifies the WETH borrow-rate lever as a last-resort mechanism to unwind those loops if exposure does not decline naturally after the freeze.

The Revenue Trade-Off

Freezing these reserves is not cost-free for the DAO. The assets proposed for deprecation generate an estimated $3.7 million in gross annual DAO revenue, led by rsETH at $1.85 million, osETH at $0.79 million, tBTC at $0.47 million, LBTC at $0.35 million, and rETH at $0.22 million, according to the governance forum post. The post does not provide a calculation methodology for this estimate.

The proposal implicitly argues that the risk-adjusted cost of maintaining these positions exceeds the revenue they generate. Governance will need to weigh whether $3.7 million in annual revenue justifies ongoing exposure to liquidity and depegging risk across eleven asset classes.

The V4 Migration Path

According to the ARFC, Aave V4 is designated as the destination for the broader asset universe being cleared out of V3 Core. In practical terms, the proposal would keep existing V3 Core positions in a voluntary wind-down while directing future support for the affected collateral categories to a separate Aave version.

Governance Questions and the Path to Implementation

Three-Stage Wind-Down Process

The ARFC outlines a staged execution rather than an immediate full shutdown. Stage one is freezing the reserves, stopping new deposits and borrows. Stage two is issuer engagement, in which LlamaRisk and Aave contributors would work with the asset issuers — Kelp DAO for rsETH, Stader for osETH, and others — to facilitate orderly position migration. Stage three activates follow-up controls if large exposures do not decline on their own timeline, with the WETH borrow-rate adjustment serving as the most direct lever against the leveraged restaking loop positions.

Reserve factors for the frozen reserves would also be raised to 99% to accelerate the wind-down economics. The governance proposal notes rETH as a specific example, with its reserve factor proposed to increase from 15% to 99%. At 99%, nearly all interest generated by remaining borrowers accrues to the DAO rather than to depositors, creating a strong economic incentive for depositors to withdraw and move positions elsewhere. This approach is consistent with how Aave V3 Risk Stewards have previously used interest rate model changes to steer user behavior without forced liquidations.

Parameters and Markets Affected

The eleven targeted reserves span three asset categories: Ethereum liquid restaking tokens (rsETH, osETH, rETH, tETH), wrapped and synthetic Bitcoin variants (LBTC, tBTC, eBTC, BTC.b), and other protocol-native or yield-bearing assets (LDO, mUSD, PT-srUSDe). The breadth of the list signals that this is a portfolio-level restructuring, not a targeted response to any single asset’s risk event.

Each asset’s freeze would be governed by the same ARFC if approved, though implementation timing for individual reserves may vary depending on the outcomes of issuer engagement.

Milestones to Watch

The ARFC is currently at the discussion stage on Aave’s governance forum. The standard Aave governance path moves from ARFC discussion to a Snapshot temperature check, followed by an on-chain AIP vote. AAVE token holders, whose governance power scales with staked AAVE in the Safety Module, will ultimately decide whether the protocol absorbs the revenue reduction in exchange for a structurally cleaner Core market.

The key open question for governance participants is migration mechanics: specifically, whether Aave V4 will have sufficient liquidity depth to absorb the displaced collateral without introducing a gap in protocol utility for restaking-heavy users.