Aave Cuts 75 Reserves Across Six Chains as $98M in Deposits Prove Unprofitable

Aave, the largest decentralized lending protocol by total value locked, has published a governance proposal calling for the complete wind-down of deployments on six blockchains – Sonic, Scroll, zkSync, Metis, Soneium, and Aptos – alongside the deprecation of 75 reserves holding a combined $98.1 million, in a move that frames underperforming multichain expansion as a liability rather than a growth lever.

Six Chains, $12.8 Million, and Revenue That Does Not Cover Maintenance

The proposal, filed by risk manager LlamaRisk on July 29, 2026, is structured in two tracks. The first covers 50 low-adoption reserves plus 21 matured Pendle principal token listings across 11 Aave V3 deployments, with approximately $85.3 million supplied and $11.5 million outstanding in debt, according to LlamaRisk data. The second track shuts six entire markets carrying 25 additional reserves, $12.8 million in deposits, and $4.1 million in open loans.

The chain-level breakdown illustrates how badly usage has fragmented. Sonic leads the closures with $7.6 million deposited, but that figure represents a 74% decline over six months. Scroll carries $2.2 million after falling 86%. Aptos holds $1.7 million following a 94% collapse in available liquidity, and at current levels the deployment generates less than $1,000 per quarter for the protocol, per LlamaRisk figures. zkSync is at roughly $844,000 after an 88% decline, Metis at $300,000 after dropping 79%, and Soneium at $200,000 following a 95% fall.

Each of the six deployments now generates under $5,000 per quarter. Metis, Soneium, and Aptos each bring in less than $1,000. For context, according to reporting by CoinDesk, Aave’s Ethereum mainnet deployment generates over $142 million annually and its Base deployment produces roughly $4.7 million. The six chains being cut together account for under 1% of Aave’s approximately $14 billion in total assets across 23 chains.

The economics are compounded by a deteriorating top line. Gross protocol revenue fell from $198 million in Q1 to $156 million in Q2, a decline of roughly one-fifth. Liquidation fees dropped from $27 million in Q2 to under $200,000 through the first month of Q3, per CoinDesk figures. Against that backdrop, maintaining oracle coverage, liquidation infrastructure, and security monitoring for markets producing a few hundred dollars in quarterly protocol revenue is structurally indefensible.

Aave Co-Founder Kulechov Frames It as Orderly Risk Reduction

In a post on X on July 30, 2026, Aave co-founder Stani Kulechov described the operation as a comprehensive review resulting in an orderly reduction of economic and technical risk, covering 50 low-adoption reserves across multiple deployments and 25 additional reserves in the six markets being wound down entirely. The framing is accurate as far as it goes. What it sidesteps is that several ecosystems simply failed to generate enough demand to justify the ongoing cost of oracles, monitoring, upgrades, and liquidation pathways – a cost that is fixed regardless of whether a market holds $200,000 or $200 million.

This is not Aave’s first signal on the subject. In December 2025, the Aave Chan Initiative backed a temp-check to raise reserve factors on weak markets, close the zkSync, Metis, and Soneium instances, and establish a rule requiring any future deployment to commit to at least $2 million in annual revenue. The current proposal formalizes and extends that threshold into a broader portfolio pruning exercise. The full ARFC is available on the Aave governance forum.

How the Wind-Down Works: No Forced Liquidations, But a Deliberately Hostile Rate Environment

Aave cannot delete user positions, so the mechanism is coercive economics rather than forced closure. The plan calls for freezing affected reserves to new deposits, borrows, and collateral use, then cutting supply and borrow caps to a single token. For reserves carrying debt, the reserve factor will be raised to 99% – meaning nearly all interest paid by borrowers flows directly to the Aave treasury – with a 5% base borrowing rate introduced. That combination squeezes depositor yields toward zero while making borrowing expensive enough that remaining users find it rational to exit on their own terms.

Existing positions are not automatically liquidated. The environment is made progressively less attractive until users unwind voluntarily. If borrowers do not repay, the rate curve can be tightened further. The design is deliberate: Aave protects existing users from abrupt closure while ensuring the protocol incurs no ongoing subsidization of dead markets.

A companion LlamaRisk proposal, noted in the governance filing, targets long-tail assets that Chainlink has flagged as carrying high or very high operational risk due to deteriorated underlying token liquidity. That package covers 10 deployments with $6.76 million supplied and $4.29 million borrowed. Rather than shutting price feeds immediately, it would freeze reserves, cut caps to one, and replace live oracles with static price adapters so that remaining loans can still be safely liquidated.

The Strategic Shift: Depth Over Presence, and V4 as the Architecture Behind It

For years, multichain expansion was framed as frictionless growth: more chains, more users, more deposits. The operational reality was fragmented liquidity spread across small markets that demand the same overhead as large ones. The collapse in Layer 2 TVL across underperforming networks has made that tradeoff increasingly visible across DeFi, and the Aave governance data gives it a specific dollar figure.

The risk is not theoretical. The primary source notes that after an incident involving rsETH linked to the Kelp DAO exploit, Aave had to freeze markets and manage contagion across multiple networks simultaneously – a direct illustration of how thin-liquidity deployments can generate operational problems far larger than the revenue they produce.

Aave V4 is the structural answer to the fragmentation problem. Its architecture, built around shared liquidity hubs and specialized markets connected to the same capital pool, is designed to prevent each new use case from requiring an isolated, capital-inefficient deployment of its own. Deposits on V4 nearly doubled month over month as of the governance filing, with deposit and borrow limits raised for the eleventh time. For investors tracking Aave’s V4 architecture and lending market expansion, the deprecation of underperforming V3 chains is the necessary clearing operation that precedes deeper capital concentration on the networks where Aave actually has pricing power.

The protocol has also been converting revenues into automatic AAVE token buybacks under the updated Aavenomics framework, which raises the internal cost of maintaining markets that cannot cover their own overhead. The deprecation proposal and the buyback program are mechanically linked: every dollar not wasted on a $3,000-per-year Metis deployment is a dollar available for buybacks or deployment on productive markets.

What Comes Next: Governance Vote, Then Parameter Changes

The ARFC is classified as an advanced technical proposal that must clear a preliminary governance vote before moving to on-chain execution. Until that vote passes, no freezes, cap changes, or rate adjustments take effect – characterizing any closure as final at this stage would be premature. The next milestone to watch is the preliminary vote outcome and whether any AAVE token holders push back on the scope of the wind-down or the specific chains included.

If the proposal passes, the more consequential question for the sector is whether the $2 million annual revenue floor becomes a standing protocol standard that shapes where Aave deploys in the future – and whether other major DeFi protocols follow with comparable pruning exercises of their own multichain footprints. For AAVE token holders, the market will be forced to price whether the cost savings and capital concentration this cleanup enables are sufficient to offset the top-line revenue decline that has been running across Q2 and into Q3.

Follow CoinNews on X and Telegram for ongoing updates on Aave governance and DeFi market structure.