What is restaking, and how does it differ from what's typically understood as staking?
Typical staking refers to locking ETH into Ethereum's validation mechanism, exchanging it for eligibility to participate in network consensus and validate transactions, earning corresponding staking return — this capital's use is singular, only providing security guarantee for the Ethereum mainnet itself. Restaking takes a further step on top of this: ETH already staked (or already converted into a liquid staking token, such as stETH) gets additionally deployed, through a restaking protocol, to provide security guarantee for other services needing it (this kind of service is often called an Actively Validated Service, or AVS) — these services could be a new oracle network, a cross-chain bridge, a data availability layer, or various other infrastructure needing a decentralized trust foundation.
The key difference from typical staking lies in 'the stackability of capital use': typical staking is 'one capital, one use' — your ETH only serves the Ethereum mainnet; restaking, by contrast, is 'one capital, multiple uses' — the same staked ETH (or its tokenized claim) can simultaneously be deployed by multiple different AVSs as security collateral, letting you stack earning from mainnet staking plus the extra rewards each AVS separately offers, theoretically substantially raising the same capital's usage efficiency — but the cost is you also simultaneously bear each of these multiple systems' own risk, not just the Ethereum mainnet's single-system risk.
Why did restaking emerge, and what problem is it trying to solve?
The Ethereum ecosystem has a large number of emerging infrastructure services (such as a new oracle network, cross-chain bridge, or data availability layer) — these services likewise need a decentralized, economically-incentive-constrained validation mechanism to ensure participants within the network operate honestly and don't act maliciously. The traditional approach requires each new service to start from zero, recruiting its own batch of validators and requiring them to stake the service's own native token as collateral — this 'building trust foundation from zero' process is often time-consuming, and a new token's market value usually isn't stable enough, making the deterrent effect of using it as collateral relatively limited.
What restaking aims to solve is exactly this problem of 'the high cost of rebuilding a trust foundation': rather than requiring each new service to build its own validator network from zero, it directly borrows the already-accumulated, massive-scale, long-term-market-validated staked capital Ethereum itself already has, extending a portion of this capital to provide security guarantee for new services. For a new service, this means being able to quickly obtain a certain degree of decentralized security at far lower cost and time than building its own validator network; for an existing ETH staker, this means being able to additionally earn rewards from these new services on top of the capital already staked, raising overall capital efficiency.
How does restaking actually work, and what does the complete flow look like from ETH's base staking to participating in multiple AVSs?
A typical restaking flow involves several steps:
In practice, most retail users don't directly operate this complex process of native staking and choosing AVSs themselves — instead, through a Liquid Restaking Protocol (LRT), the entire process gets packaged into a relatively simple deposit action, with the protocol handling the specific AVS selection and operator management on their behalf.
What's the practical impact of restaking on everyday users, and how should the risk of participating in restaking be assessed?
For a user wanting to raise their ETH staking capital's efficiency, restaking offers a concrete option for stacking a return source — without needing to commit additional new capital, simply extending an already-staked position into a restaking protocol, theoretically letting them additionally earn AVS reward on top of the original return. For a user wanting to maximize capital usage efficiency, this genuinely is an attractive option.
But a few concrete aspects worth watching when assessing restaking risk: stacked penalty risk — as mentioned earlier, restaking means you're simultaneously exposed to each of multiple systems' own penalty rules; the more AVSs participated in, theoretically the higher the stacked risk exposure runs, needing concrete assessment of whether each AVS you choose to participate in has a reasonably designed penalty mechanism and whether the trigger threshold is overly sensitive; operator concentration risk — if a large amount of restaked capital is concentrated and delegated to a small number of operators, once these operators run into trouble (whether technical malfunction or malicious behavior), the impact scope could be fairly broad; and the restaking protocol's own smart contract risk — this is a relatively emerging technical field, and the protocol's own code security likewise needs assessment, not directly committing capital purely based on the surface-level allure of stacked return. A real case also reminds us that even a larger-scale, higher-market-attention restaking-related protocol can genuinely still experience a security incident — once this kind of incident happens, not just the affected protocol itself, but potentially triggering a wavering of capital confidence across the entire sector, concentrating into a large-scale wave of withdrawals — this kind of systemic chain reaction is an important reminder that assessing restaking risk shouldn't just look at a single protocol, but consider the entire sector level.
EigenLayer is currently the largest-scale protocol in the market, considered the representative case of the restaking sector — at the start of 2026, its managed scale at one point approached nearly $20 billion, holding over roughly 90% of the entire restaking market's share; but this sector isn't without risk incidents either — in April 2026, Kelp, a liquid restaking protocol related to the EigenLayer ecosystem, suffered a roughly $300 million-scale security incident, which triggered a roughly $5.4 billion concentrated wave of withdrawals across the entire restaking sector, reflecting that even a large-scale, high-market-attention restaking-related protocol can genuinely still face real security risk, and that risk can, through sector-level wavering confidence, spread to affect other protocols not directly victimized.
The advantage is letting already-staked capital simultaneously provide security guarantee for multiple systems, stacking multiple return sources, substantially raising the same capital's usage efficiency, while also offering emerging infrastructure services a way to obtain a trust foundation at far lower cost than building their own validator network; the drawback is risk stacks equally — the more AVSs participated in, the higher the risk exposure under penalty mechanisms, and the entire sector is relatively emerging — the protocol's own smart contract risk, operator concentration risk, and the systemic risk that a single incident could trigger wavering confidence across the entire sector are all variables needing cautious assessment, not yet having been through sufficiently long-term market validation.