A new Aave Request for Comment proposes increasing borrowing efficiency for major ETH- and BTC-linked collateral across several deployments, including V3 Ethereum Core, Arbitrum and Base, as well as the V4 Ethereum Main Spoke. The suggested changes would raise loan-to-value and liquidation threshold settings for assets such as WETH, wstETH, weETH, WBTC and cbBTC, while also adjusting some related risk parameters on Base.
The proposal says the revisions are based on historical price behavior and Chainlink feed updates, with the key benchmark set at the 99.9th percentile of the worst price move within a one-hour liquidation window. According to the analysis, that one-hour period reflects how long positions are typically cleared through sequential liquidations during fast markets.
What the proposal would change
On V3 Ethereum Core, the ARFC would move WETH to 81% LTV and 84% liquidation threshold, wstETH to 79% and 82%, weETH to 78% and 81%, and both WBTC and cbBTC to 81% LTV and 85% liquidation threshold. On Arbitrum, it proposes WETH at 81% LTV and 81% liquidation threshold, and WBTC at 78% and 82%. On Base, WETH would move to 81% and 84%, while cbBTC would also move to 81% and 84%.
The Base proposal also includes lowering cbBTC’s liquidation bonus to 6.00% and raising the Base cbBTC Stablecoins E-Mode configuration to 82% LTV and 85% liquidation threshold. On the Aave V4 Ethereum Main Spoke, the suggested collateral factors are 84% for WETH, 82% for wstETH, 81% for weETH, and 85% for both WBTC and cbBTC, to be applied through a dynamic configuration update rather than a new collateral deployment.
How the risk model was built
The recommendation is tied to a model that calibrates each threshold against a reserve’s liquidation bonus and the worst observed move over a one-hour window. The proposal describes that window as the critical period because it captures how long a position remains exposed while liquidators process opportunities after a price feed update.
To build the model, the authors examined liquidation behavior from August 2025 to August 2026 and compared liquidations on Aave V3 Ethereum Core, Arbitrum and Base with the price feeds each deployment reads. The review also covered the February 2025 and October 2025 stress events, with Ethereum joined to SVR publications in October 2025 and to pre-SVR standard aggregators in February 2025.
The proposal states that the model ceiling for a liquidation threshold is derived from the 99.9th-percentile excursion and the liquidation bonus. Recommended settings are then placed at or below that ceiling, with WETH at the ceiling, wstETH and weETH one point below it, and BTC settings at least three points inside the implied buffer to account for risks such as market depth, supply caps and concentration that the price model does not directly capture.
Why ETH and BTC are treated differently
The analysis says ETH remains materially more volatile than BTC over the two-year sample, citing annualized volatility of 68.8% for ETH against 44.4% for BTC. It also reports a one-hour p99.9 excursion of -11.85% for ETH and -5.09% for BTC on the binding leg, supporting lower relative efficiency for ETH-linked assets than for BTC collateral.
At the same time, the proposal argues that current blue-chip collateral settings have stayed relatively stable even as volatility and market structure have evolved, creating a trade-off where conservative parameters limit how much users can borrow. The ARFC presents the proposed changes as an effort to align parameters more closely with observed volatility and liquidity rather than simply expanding risk appetite.
Liquidation evidence and implementation context
For wstETH and weETH, the proposal points to 6,486 liquidations and roughly $361 million of seized collateral across all deployments during the two years to August 2026, including about $296 million in wstETH and $47 million in weETH. It says processing speed for those liquid staking tokens was effectively in line with WETH during the stress events, supporting their treatment as blue-chip collateral in practice.
The document also says liquidation processing was measured in seconds across the markets studied, and attributes the slow tail largely to dust debt. It reports that the two stress events produced $0 and $0.39 million of bad debt respectively, with no bad debt tied to the analyzed collateral. The residual scenario outside the model basis, according to the proposal, would be an oracle and liquidation pipeline stall occurring at the same time as a move beyond the 99.9th percentile. The worst realized one-hour moves cited were -24.27% for ETH and -11.15% for BTC.
On Ethereum Core, liquidation bonuses would remain at current levels, including cbBTC’s 7.50% bonus, even though the proposal says depth on Ethereum could also support a 6.00% bonus like Base. It notes that an increase to the liquidation protocol fee for these assets on Ethereum Core is pending, and says cutting the bonus at the same time would further reduce the liquidator’s net share.
Next stage in the governance process
The ARFC is now open for community and service provider feedback. If consensus forms around the proposed parameter changes, the next confirmed step would be escalation to a Snapshot vote.
If Snapshot passes, the changes would then move to an Aave Improvement Proposal for implementation. Until then, the figures in the ARFC remain proposed settings rather than approved protocol changes.
Source: governance.aave.com