TLDR
Aave (AAVE) saw about 2627 million dollars of forced liquidations after a misconfigured risk oracle briefly undervalued wstETH collateral on its lending markets.
- A CAPO risk oracle configuration error priced wstETH roughly 2.85 percent below market, triggering automatic liquidations across 34 accounts.
- Around 10,938 wstETH was liquidated and liquidators earned roughly 499 ETH, but Aave incurred no bad debt and Lidos wstETH and protocol remained sound.
- Aave governance is preparing full reimbursement and risk control changes, and future debates will likely focus on how DeFi protocols design and audit oracle layers.
Deep Dive
1. What Went Wrong
According to Chaos Labs and Coindesk, Aaves CAPO risk oracle temporarily valued wstETH at about 1.19 ETH while the actual market rate was closer to 1.23 ETH, a roughly 2.85 percent underpricing of collateral value.
CryptoBriefing explains that the bug came from a mismatch between a seven day old snapshot ratio and an on chain constraint that only allowed a 3 percent change every three days, causing the CAPO formula to output an artificially low maximum exchange rate for wstETH.
This did not come from Lidos underlying price feed; Chaos Labs and Lido both emphasize that the base oracle reported correct market values, and Lido said the incident was unrelated to wstETH itself, pointing instead to Aaves CAPO configuration issue.
The failure was in Aaves internal risk oracle logic, not in Ethereum, Lido, or the main price oracle, highlighting how extra risk layers can fail even when base feeds are accurate.
2. Who Was Affected
The mispricing made some leveraged wstETH borrowers look undercollateralized, which automatically triggered liquidations of about 10,938 wstETH across 34 accounts, worth roughly 2627 million dollars at the time.
Liquidators captured about 499 ETH in bonuses and profits, yet reports note that Aave itself did not take on any bad debt because the system liquidated positions fast enough to keep the protocol solvent.
Chaos Labs and community summaries indicate that borrowing caps were temporarily tightened and that some ETH has already been recovered via BuilderNet refunds, with the Aave DAO treasury expected to cover the remaining user losses.
Users bore the direct losses, liquidators profited, and the protocol stayed solvent, so the risk showed up as user level liquidation pain rather than a protocol insolvency event.
3. What To Watch Next
Lido and risk contributors have already pushed fixes for the CAPO configuration, but the key open item is the Aave DAO vote that will formalize full reimbursement from treasury and recovered funds.
Expect governance discussions around adding safeguards such as tighter bounds on risk oracle adjustments, more frequent audits of oracle configuration, or diversified oracle designs that reduce the impact of misconfigured parameters.
Market wise, coverage notes that AAVE traded down with other DeFi names during the broader risk off move, but the structural risk question is whether users and treasury are comfortable with a model where configuration bugs can cause eight figure liquidations that are later socialized.
The main signal to track is how quickly and transparently Aave compensates users and hardens its oracle stack, because that will shape long term trust in its risk engine.
Conclusion
Aaves rare 27 million dollar liquidation spike came from a narrow but impactful risk oracle configuration bug that temporarily mispriced wstETH collateral.
The protocol stayed solvent, Lidos wstETH remained fundamentally sound, and governance is moving toward full user reimbursement, but the incident puts a spotlight on oracle design and configuration risk across DeFi lending.
CoinsKid.Com