Bridge-Confidence Contagion in Liquid Restaking: A Pre-Registered Null Around the April 2026 KelpDAO Exploit

On April 18, 2026, attackers drained 116,500 rsETH (~$292M, 18% of supply) from KelpDAO's LayerZero bridge by forging a cross-chain message under a 1-of-1 verifier configuration. I test whether the resulting depeg in rsETH spread to other liquid restaking and liquid staking tokens that share LayerZero's Omnichain Fungible Token bridge infrastructure. Using a pre-registered difference-indifferences design on daily prices for nine LRT/LST tokens over February 3 to May 3, 2026, I find a large idiosyncratic depeg in rsETH (+781 bps post-event, t = 212) but no statistically significant differential effect on LayerZero-bridged peers relative to non-LayerZero controls (-9.6 bps, t =-1.41).

Four pre-committed robustness specifications, including volatility, asymmetric response to ETH stress, a tight event window, and signed wedge, all return null at a Bonferroniadjusted threshold. The result is consistent with market participants treating the event as a Kelpspecific failure rather than a LayerZero-wide bridge-protocol shock, even as protocol operators (Ethena, ether.fi) treated it as bridge-specific by precautionary pauses of their own LayerZero bridges. The divergence between operator caution and market pricing is itself the finding.