Share
A temporary configuration issue in Aave’s risk-oracle system triggered a wave of liquidations across the decentralized lending protocol, wiping out roughly $26 million worth of wrapped staked Ether (wstETH) positions on March 10, 2026.
The incident stemmed from a misalignment in the protocol’s Correlated Asset Price Oracle (CAPO), which briefly valued wstETH below its real market ratio against Ether. The pricing mismatch caused a cluster of borrowing positions near their safety thresholds to appear under-collateralized, allowing liquidation bots to seize collateral while broader market prices remained stable.
Aave Liquidations Triggered by Oracle Misconfiguration
A misconfigured price feed on @aave triggered about $26 million in liquidations on March 10 after wstETH was briefly valued around 0.9 ETH instead of ~0.99.
The error originated from Aave’s CAPO (Correlated Asset Price… pic.twitter.com/Udo5SJ2ByQ
— Crypto Miners (@CryptoMiners_Co) March 12, 2026
According to a post-mortem published on Aave’s governance forum and analysis from risk manager Chaos Labs, the glitch originated from a mismatch between two parameters used by CAPO: a reference exchange ratio and its associated timestamp.
CAPO is designed to limit how quickly the price ratio between correlated assets such as wstETH and stETH can change. The safeguard helps protect lending markets from sudden price manipulation.
In this case, the oracle system attempted to update the ratio using fresh market data. On-chain constraints limited the adjustment to a 3% change within a 72-hour window, while the stored timestamp referenced older data, causing CAPO to calculate a maximum exchange rate below the real market value.
As a result, Aave temporarily priced one wstETH at roughly 1.19 ETH, while the broader market valued it closer to 1.23 ETH, a deviation of about 2.85%.
The discrepancy pushed several borrowing positions below their liquidation thresholds, triggering automated liquidations.
On-chain data compiled by Chaos Labs shows that about 10,938 wstETH across roughly 34 accounts were liquidated during the incident. The total value of those positions was estimated near $27 million at the time.
Liquidation bots captured roughly 499 Ether in bonuses and value tied to the temporary pricing error, equivalent to roughly $800,000 based on ETH market prices at the time.
Aave founder Stani Kulechov said the protocol itself incurred no bad debt, meaning the system remained solvent despite the oracle issue.
Most of the affected positions were already close to their collateral thresholds.
TL;DR on the stETH CAPO configuration:
– Chaos Risk Oracles is an external tool used within Aave.
– It has historically processed over 1,200 payloads and 3,000 parameters without issues.
– CAPO is a separate defense mechanism designed to protect against inflation attacks or… https://t.co/LcoThpx7p9
— Stani.eth (@StaniKulechov) March 10, 2026
Chaos Labs and Aave developers say the configuration issue has already been corrected.
The protocol recovered 141 ETH from liquidation bonus revenue through BuilderNet refunds and collected another 13 ETH in liquidation fees, which will be used to reimburse impacted borrowers. Any remaining compensation will come from the Aave DAO treasury.
Chaos Labs CEO, Omer Goldberg, said all affected users will receive full reimbursement.
17/ Next Steps
Every affected user will be fully reimbursed.
The Aave DAO SPs are finalizing the reimbursement plan and will publish it shortly.
Notably, 141 ETH have already been recovered via BuilderNet refunds.
— Omer Goldberg (@omeragoldberg) March 10, 2026
Oracle systems play a central role in decentralized lending markets because they determine how collateral is valued. Even small pricing errors can cascade through automated liquidation systems.
The event echoes a similar failure from February 2026, when an AI-generated smart contract error on the Moonwell protocol mistakenly priced cbETH near $1, leaving the lending platform with roughly $1.78 million in bad debt.
The incident didn’t create losses for the protocol, but it shows how even small pricing discrepancies inside automated risk systems can trigger large liquidations in DeFi lending markets.
Share
