Understanding Restaking: How EigenLayer Leverages Staked ETH for Shared Security

For yield strategists, the structure is familiar: stacked slashing conditions on a single capital base. The trade-off is compounding exposure on collateral previously limited to a single validation path.

How EigenLayer partitions the roles

EigenLayer formalized the restaking primitive across three distinct participants. Restakers deposit native ETH or liquid staking tokens (stETH from Lido, rETH from Rocket Pool) into EigenLayer contracts and delegate to an operator. Operators register with EigenLayer, receive delegated stake, and opt into one or more actively validated services. AVSs — the third role — define their own validation logic, reward structure, and slashing conditions. When an operator opts into an AVS, the restaked ETH backing that operator becomes subject to that AVS's specific slashing rules.

The critical architectural point: EigenLayer's contracts enforce delegation and slashing mechanics, but do not define what constitutes a slashable offense. Each AVS sets its own parameters. A restaker's effective exposure is therefore a function of every operator-AVS pairing in their delegation chain, not a single uniform risk profile.

Slashing layers and peg vectors

The baseline case is well-understood. An Ethereum validator deposits 32 ETH, earns roughly 3–4% annualized for block proposals and attestations, and faces slashing for double-signing, conflicting block proposals, or extended downtime. The economic security guarantee is straightforward: attacking consensus requires controlling enough staked ETH that the cost of being slashed exceeds the attack profit. With over 30 million ETH staked, that threshold sits beyond reach for any rational attacker.

Restaking layers additional slashing conditions onto the same collateral. Each opted-in AVS introduces an independent slashing vector tied to the identical 32 ETH. Liquid staking tokens compound the stack further: stETH and rETH add depeg risk on the derivative, plus operational risk on the underlying validator, plus AVS-specific slashing risk on top. The risk surface is no longer flat — it's a multi-dimensional exposure profile that demands separate stress-testing per layer.

Bitmine Immersion Technologies recently disclosed holdings of 5.8 million ETH, with 4.9 million staked, projecting $291 million in annualized staking rewards at scale. That concentration underscores the leverage embedded in large-scale staking operations — and the magnification potential of restaking exposure per operator.

Position-sizing through the stack

Before allocating capital to restaked positions, the practical checklist maps directly to the architecture: identify the operator's full AVS commitments, review each service's slashing conditions independently, and stress-test against correlated failure modes. Utilization rate on the operator — not headline restaking APY — is the cleaner signal for risk-adjusted yield.

The yield-stacking thesis extends well beyond on-chain mechanics. For a concrete off-chain example of layered incentives on a single commitment, this Galaxy Watch Ultra 2 pre-order bonus structure illustrates how front-loaded rewards can reshape purchase decisions much the way restaking reshapes validator economics — same underlying principle, different asset class.