DeFi: Aave considers largest wave of reserve removals in its history

Aave wants to withdraw several reserves that have become too little used. This transaction would involve $98.1 million in deposited assets and $15.6 million in debt. For the crypto protocol, it is not a question of responding to an immediate emergency. The objective is rather to reduce the risks and costs linked to markets that have become unprofitable.

Aave wants to remove little-used crypto reserves

The proposal provides for the withdrawal of 50 reservations with limited adoption. It also targets 21 mature Pendle Principal Tokens. This reorganization comes as Aave prepares its V4, a version designed to better separate liquidity and risk management. The affected reserves are spread across 11 Aave V3 deployments. They represent approximately $85.3 million in provided assets and $11.5 million in debt.

These amounts remain modest on the scale of the crypto protocol. Their maintenance, however, mobilizes several technical services. Each market notably requires a price oracle, a liquidation mechanism and regular monitoring. When activity becomes too low, the revenue generated no longer always compensates for these costs. Aave therefore prefers to concentrate its resources on the most active reserves.

The plan goes further than a simple deletion of tokens. Aave is also considering fully shutting down its deployments on Sonic, Scroll, zkSync, Metis, Soneium and Aptos. These six markets bring together 25 additional reserves. Their withdrawal would affect approximately $12.8 million in deposited assets and $4.1 million in debt. The crypto protocol believes that the revenue from these deployments no longer justifies the costs necessary to maintain them.

This decision illustrates the limits of too rapid multi-chain expansion. Deploying a protocol across multiple networks can attract new users. But each additional blockchain adds technical risks, liquidity needs and operational expenses. Aave therefore does not renounce its multi-chain model. The crypto protocol instead seeks to close markets that have become peripheral. This logic also occurs in a context of tensions around risk management.

Progressive measures to push users to withdraw

Aave does not plan to immediately remove the affected reservations. The process would be gradual. Lenders and borrowers would then have the time necessary to close their positions. The protocol would start by blocking new deposits and new borrowing. The procurement and borrowing limits would then be reduced to a single unit. This measure would virtually prevent new activity without abruptly disrupting existing positions.

For assets that can still be borrowed, Aave would also increase the reserve factor. A larger share of the interest paid by borrowers would then accrue to the protocol. Maintaining these positions would gradually become less attractive. The six deployments earmarked for complete closure could face stricter measures. The reserve factor would rise to 99%, while base interest rates would increase. The objective is clear: to encourage users to naturally close their positions.

This reorganization stems from the new risk framework proposed by Aave in June 2026. This was presented after the KelpDAO bridge attack, which exposed the protocol to significant bad debt risk. In this incident, stolen rsETH was deposited into Aave as collateral. The attacker then borrowed other assets against this collateral. Even if Aave was not directly responsible for the breach, the episode showed that an external asset could transmit its risk to the protocol.

The platform then suffered significant withdrawals, while continuing to operate. Tremplin.io had analyzed this liquidity crisis and the measures envisaged to better isolate the risks of contagion. The proposed withdrawal therefore does not only target assets deemed dangerous. It also concerns tokens reproducing native assets, Pendle products that have expired and markets whose activity has become insufficient.