A governance proposal is under review by the Aave risk management team, aiming to restructure the risk parameters for Bitcoin-backed lending. The initiative, submitted by risk service provider LlamaRisk on September 21st, is currently in the Snapshot review phase, with a vote expected to commence within 24 hours. Although the core objective of the proposal is to optimize capital efficiency by enabling higher leverage ratios, the voting outcome and final implementation status remain undetermined at this time; all elevated parameter limits are merely in a proposed state and have not been activated.
The central point of contention within this governance action lies in the fact that as the pursuit of higher collateral utilization intensifies, the tolerance for price declines faced by borrowers is significantly reduced. This dynamic creates a fresh tension between protocol safety and user experience. Notably, this proposal is not an isolated event but rather an attempt by Aave, following multiple market stress tests, to adapt its risk models to current market liquidity through fine-tuned adjustments; however, the underlying data support and potential systemic risks still require deeper examination.
On the Aave V3 Ethereum Core platform, this proposal materially raises the Loan-to-Value (LTV) ratio for WBTC and cbBTC. The newly proposed LTV is set at 0.81, a substantial increase from the previous 0.73. This means that for every $1 of WBTC or cbBTC supplied as collateral, the maximum amount a user can theoretically borrow increases from $73 to $81. Concurrently, the liquidation threshold is also being raised from 78% to 85%. While these surface-level figures suggest a shift favorable to borrowers, the underlying logic reveals a dramatic contraction in the risk buffer space. In the original parameter framework, there was a 5-percentage-point gap between the LTV and the liquidation threshold; the new proposal compresses this gap to 4 percentage points. More critically, when considering a constant debt size and fluctuating Bitcoin prices, the tolerance for a decline in collateral value relative to the liquidation threshold is now only 4.7%. In essence, despite a higher borrowing ceiling, borrowers are positioned closer to the liquidation point, where even minor market fluctuations could trigger the liquidation mechanism. This structural shift significantly increases the vulnerability of highly leveraged positions during extreme market conditions.
These parameter adjustments are not confined to the Ethereum Core platform but extend across multiple chain environments and asset classes. On the Arbitrum platform, the standard LTV for WBTC is proposed to increase by 5 percentage points. On the Base platform, the LTV increase for cbBTC is more aggressive, rising by 8 percentage points. Additionally, Ethereum-related assets on the Ethereum Core platform, including WETH, wstETH, and weETH, are slated for a 0.5-percentage-point increase in their LTV. Along with these LTV adjustments, several liquidation thresholds are also set to rise; however, the liquidation reward rate for cbBTC on the Base platform is proposed to decrease from 7.5% to 6%, a change aimed at balancing liquidation incentives with market liquidity. It is particularly noteworthy that the E-Mode specifically designed for cbBTC on the Base platform has a proposed LTV of 82% and a liquidation threshold of 85%. These figures represent the maximum debt that each unit of eligible collateral can support, but the proposal does not disclose the complete collateral balances, the debt scale corresponding to each asset type, or the distribution of account health. Because some tokens may not meet collateral requirements, may be disabled, or are unrelated to debt, total reserve figures often overestimate the actual available collateral amount. Consequently, historical liquidation data may not fully reflect the current risk distribution of outstanding positions, injecting uncertainty into the systemic risk assessment following multi-platform expansion.
LlamaRisk conducted a detailed retrospective analysis of liquidation events on Ethereum Core, Arbitrum, and Base platforms spanning from August 2025 to August 2026. On the Ethereum Core platform, there were 7,206 liquidations of Ethereum assets, totaling $618 million, and 2,621 liquidations of Bitcoin assets, amounting to $358 million. The data indicates that, on a value-weighted basis, 99% of liquidation requests were completed at or near the execution price, with a processing time of just 5 minutes. This statistic refers to the window for processing liquidations, not the time for transaction completion following an oracle price update. The research also examined data processing during stress test periods in February and October 2025: during those two events, 100% of seized assets across all listed markets were processed within 5 minutes of an oracle price update that made liquidation profitable. February saw no significant adverse outcomes. In October, although $390,000 in assets became bad debt, the total debt was approximately $128 million, and LlamaRisk noted that this bad debt did not affect the Ethereum or Bitcoin collateral analyzed in this proposal. These results suggest that in historical instances, the speed of liquidators was not a limiting factor. Even with higher leverage ratios, the swift liquidation mechanism proved effective, though this does not guarantee similar outcomes during future extreme market conditions.
However, the model’s effectiveness relies heavily on accurately capturing extreme price volatility, and the existing data exhibits clear internal contradictions and risk blind spots. In the detailed data tables, the 99.9th percentile for the worst 1-hour price decline for Ethereum assets is 11.85%, while for Bitcoin it is 5%. This percentile excludes the most extreme 0.1% of observations from the two-year sample. In reality, historically recorded extreme drawdowns exceed these figures significantly: the worst 1-hour drop for Ethereum was 24.27%, and for Bitcoin it was 10.72%. The proposal summary lists a worst-case 1-hour drop for Bitcoin of 11.15%, which is more than double the 5% percentile figure, a significant discrepancy. This difference constitutes the core of the residual risk. LlamaRisk’s model assumes that regular oracle updates and efficient liquidators ensure highly leveraged positions won't remain unattended for a full hour. But if price data updates stall, the liquidation process slows, or market depth deteriorates simultaneously, price movements exceeding that percentile could have more severe consequences. Therefore, this percentile is actually designed to function as a buffer for bad debt within the protocol, not as protection for borrowers. Based on simplified calculations, a highly leveraged Bitcoin-backed position could hit the proposed liquidation threshold after a collateral price drop of approximately 4.7%. The protocol's model, in turn, assesses post-liquidation coverage by using a price fluctuation margin of roughly 5% plus the liquidation reward rate. Aave's governance team must determine whether current liquidation performance under existing parameters justifies allowing positions to remain closer to the liquidation point in the future. For Bitcoin-collateralized assets on the Ethereum Core platform, the maximum LTV ratio will increase by 8 percentage points. At the highest leverage level, the simple collateral price buffer shrinks from approximately 6.4% to 4.7%. Existing borrowers will not automatically increase their debt, but the new limits will permit borrowers opening new positions or adjusting existing ones to take on more debt.
From a protocol-wide perspective, the situation appears more positive than the borrower's buffer metric alone suggests, given that economically significant liquidations in historical samples have been completed swiftly, and no bad debt was incurred for the analyzed Ethereum and Bitcoin collateral types during the two stress test periods. Nevertheless, past trading conditions do not reflect the systemic failures that can accompany anomalous price swings. In the same dataset, the worst 1-hour price declines far exceeded the levels reflected by the percentile, and the model cannot eliminate liquidity risk, concentration risk, or oracle risk. The forum has stated that implementation guidelines will only be issued following a positive outcome from the Snapshot review. Until the vote and official guidelines determine final values, the 81% Bitcoin LTV ratio remains merely a governance proposal. The unresolved critical question is how Aave should interpret the reference value of historical data when the market experiences unusually large price fluctuations, coupled with pricing dislocations, liquidation difficulties, or insufficient market depth.
Comments