Overview
Following our previous report, systemic volatility across the broader crypto market has increased materially. Several major assets have experienced sustained drawdowns in excess of 20%, driving elevated liquidation activity across Aave deployments.
Over the past seven days, approximately $467M of collateral has been seized across 18,860 liquidation events, with an estimated $22M paid in liquidation bonuses. Aave’s liquidation mechanism has operated as intended during this period: liquidations were executed continuously across markets, and bad-debt accruals remained minimal, with any residual impact disproportionately offset by protocol revenue stemming from liquidation fees and SVR recapture.
Oracles
Throughout the stress period, Aave’s primary oracle infrastructure behaved in line with expectations. The majority of price update events remained within configured heartbeat and deviation thresholds. A subset of updates exceeded the mechanistic deviation bounds; however, these instances are consistent with abrupt spot repricing during high-volatility regimes rather than systematic oracle underperformance.
Overall, no evidence was observed of prolonged staleness, persistent mispricing, or update failure modes that would impair liquidation execution at the protocol level.
Liquidations
Since our last update, liquidators have seized over $340M in collateral, with activity concentrated in WETH, WBTC, and wstETH across major Aave v3 deployments. Additional liquidation volume was driven by the adverse price performance of higher-volatility collateral assets, including AAVE and LINK. Notably, USDT appears among the larger seized-collateral contributors despite limited direct price volatility. This is best explained by cross-collateralized positions that combined stablecoin collateral with volatile assets: as volatile collateral depreciated, liquidation became necessary at the portfolio level, and liquidators frequently elected to seize stablecoin collateral to reduce their inventory risk or swap associated costs.
SVR Performance
During the observed volatility window, SVR contributed materially to protocol-level outcomes by recapturing a portion of liquidation-driven value flows and converting them into net protocol revenue. Over the period covered, $297.51 million of collateral was seized, and $283.61 million of debt was repaid, with $13.89 million paid out in liquidation bonuses. In aggregate, SVR resulted in approximately $3.11 million of net profit attributable to Aave. These results suggest that liquidation execution remained efficient.
AAVE Holder Liquidation
During the liquidation window, we monitored a specific large position in detail. As prices began to decline, we identified a high-risk address with USDC debt collateralized primarily by AAVE. Notably, the account represented approximately 2% of total AAVE supply, making it systemically relevant in the context of stressed market conditions and liquidation feasibility.
Despite the position becoming eligible for liquidation, the effective on-chain liquidity for AAVE was insufficient to support usual liquidation sizes which typically account for 50% of the position. As a result, liquidation activity fragmented into a sequence of smaller fills, which is evident both in the transaction table and in the stepwise evolution of collateral and debt presented below. Most liquidations were executed in the $0.5 - $1.5 million seized collateral range, whereas a more liquid market would typically incentivize liquidators to seize a larger portion of the position per transaction to maximize net liquidation profit. While the arbitrage of AAVE/stablecoin pools exhibited limited efficiency, with liquidity pools often not being rebalanced to reflect the instantaneous price of the asset, the majority of solvers utilized AAVE/WETH or AAVE/wstETH pools, which given the high correlation of the assets, moved in line, and required less external involvement to stay relatively balanced. During the events the position has largely been out of risk, while the liquidations were chipping away 2-6% of the position at a time, the health factor returned to healthly values after the majority of collateral seizure events, when this did not occur a rapid subsequent liquidation would typically happen.
The position was ultimately cleared via a final, substantially larger liquidation executed by Wintermute, who effectively assumed the inventory risk and appears to have offloaded exposure through alternative channels. This behavior is atypical in Aave liquidations, as taking inventory typically requires liquidators to internalize additional costs and risks (including MEV compensation, hedging requirements, and balance sheet constraints) that most searchers avoid under stressed liquidity conditions.
| Block |
Seized AAVE ($) |
Repaid USDC ($) |
Transaction Hash |
| 24393011 |
$1,136,166 |
$1,064,324 |
0x25784da3e88e979dbe4257cfe9cd3dd9b1c40a7f7d317be2132abf43f38f48ea |
| 24393012 |
$519,459 |
$486,612 |
0x0ec3a4ce06ab809e82cb454a6e41b149d3bd8c5e72d0606a17fd039a2240753d |
| 24393012 |
$487,720 |
$456,881 |
0xa5821d691a36055014bd53731ed355585934031c5909b159fbec437bf41205fb |
| 24393136 |
$1,477,013 |
$1,383,619 |
0xc9f368231afd429d992ac444fb21d34a19c8c2a6b881a9bc38b0b95d53a20630 |
| 24393462 |
$1,268,481 |
$1,188,273 |
0x8ddc5fd2380a4c95a4055262b22e619e83e71ffb9c950423e99e0d0f3daa3324 |
| 24393476 |
$761,068 |
$712,945 |
0x884454663e666602b0266f1306bd030e46a99a05577e90a54c82c6d0c79b5f3b |
| 24393477 |
$491,399 |
$460,327 |
0x98d77b8bb1ae2e24a780e66141641d16c33dd558be7214c5f82164278a527ab0 |
| 24393577 |
$1,262,741 |
$1,182,896 |
0x17946b973cd90443605eade557b4634cc710e866b4ad9a517acb579f111459fc |
| 24393595 |
$1,222,914 |
$1,145,587 |
0xff8e99dfe6a54fb387c2f57a2783dbf7decc25923d3ced40377c09e32f9d2062 |
| 24394116 |
$10,823,247 |
$10,138,873 |
0x29c91c20c028761bac0b3c78ce354543ea61ab4e198ddec2c56d7af5a2b2b083 |
To further illustrate the execution constraints associated with liquidating the position, we also examined post-liquidation balance trajectories for addresses that appear to be associated with liquidation execution and subsequent distribution. As shown below, the observed balance changes are consistent with a workflow in which seized AAVE is initially accumulated by a solver/searcher address and subsequently transferred to a counterparty that resembles an OTC facilitation venue or desk. We emphasize that these labels are based solely on observed on-chain balance movements and transfer patterns; they should not be interpreted as definitive attribution.
In practical terms, this pattern provides a plausible explanation for how the final liquidation cleared. Rather than unwinding the seized inventory via liquidity pools, where slippage would compress or eliminate liquidation profitability, the liquidator appears to have relied on off-chain or alternative liquidity channels to distribute the asset.
Conclusion
Overall, Aave’s core liquidation mechanics have performed robustly through a period of elevated volatility. However, the AAVE-collateralized USDC position highlighted above demonstrates that, for highly concentrated collateral, secondary-market depth can become the binding constraint on liquidation sizing and latency. In such cases, clearing the position may depend on participants willing to assume inventory risk and source liquidity beyond immediate on-chain execution. While this outcome has resolved in no bad debt for the protocol, it provides a useful empirical reference for stress scenarios and liquidation outcomes.
Disclosure
Chaos Labs has not been compensated by any third party for publishing this analysis.
Copyright
Copyright and related rights waived via CC0.
