KelpDAO attack via LayerZero verification triggers rsETH loss, Aave outflows and DeFi TVL plunge
A KelpDAO attack this weekend drove a sharp DeFi TVL plunge, with total value locked shrinking by about $13B. The incident is linked to LayerZero’s verification infrastructure (not a standard smart-contract code exploit), and analysts say the attacker may be the Lazarus Group.
After the KelpDAO exploit, rsETH lost its backing. That immediately increased liquidation risk across ETH lending, especially in Aave’s ETH pool. Over the next 48 hours, users exited aggressively, with Aave recording about $8.45B of outflows and total DeFi assets sliding back to the mid-$80B range.
Traders should note the TVL damage can exceed the reported ~$292M stolen amount because DeFi leverage counts the same collateral multiple times. Low yields also likely amplified risk-taking earlier—on Aave, USDC deposits reportedly yielded only ~2.61% annually, making complex leverage less attractive after the shock and speeding position unwinds.
While some commentators said “DeFi is dead,” the later framing emphasizes risk-premium repricing rather than permanent collapse. Aave’s loss-absorption features matter, but the near-term implication is tighter risk budgets. Capital appears to rotate: Spark reportedly scaled back lower-demand rsETH exposure and its TVL rose from about $1.8B to $2.9B.
Bearish
Because the KelpDAO exploit caused rsETH backing to crack, ETH collateral reliability in lending markets deteriorated immediately. The resulting liquidation pressure translated into large Aave ETH-pool outflows and a broad DeFi TVL drawdown, which typically tightens leverage and increases risk premiums in the near term.
In the short run, traders should expect higher funding/liquidation sensitivity around ETH lending collateral and a more cautious stance toward LST-based strategies. In the medium-to-long run, the story is less about a total DeFi shutdown and more about repricing systemic risk from bridge/verification-layer failures, which can raise on-chain capital costs. The rotation effect (Spark reallocating away from rsETH) suggests capital may move rather than vanish, but the overall bias for ETH-linked liquidity remains negative until confidence in collateral is restored.