Restaking Explained: How Ethereum Rents Out Its Security

Ethereum’s proof-of-stake network pays validators to stay honest, with more than 30 million ETH locked as collateral that can be slashed if anyone tries to cheat. Restaking asks a blunt question: if that collateral already secures Ethereum, why not let it secure other things too, and earn extra yield for the trouble?

The answer became one of the largest experiments in crypto over the past two years. At its 2024 peak, restaking pulled in close to $20 billion, spawned a wave of new tokens, and drew a public warning from Ethereum co-founder Vitalik Buterin. It has since cooled, survived its first nine-figure exploit, and picked up cautious signals from a friendlier US Securities and Exchange Commission. Here is how restaking works, who runs it, and where the risks sit.

What restaking actually is

Staking locks a crypto asset to help secure a proof-of-stake blockchain and pays a reward for the service. Restaking takes an asset that is already staked, or a token that represents a staked position, and commits it a second time to secure additional services. The same capital does two jobs. In exchange, the restaker accepts extra rules and extra penalties.

The services being secured are things Ethereum does not do natively: data availability layers, oracle networks, cross-chain bridges, rollup sequencers, and co-processors. Each of these normally needs its own validator set and its own token to bootstrap trust, which is slow and expensive. Restaking lets them rent economic security from Ethereum’s existing stake instead, an idea builders call shared security or pooled security.

The pitch is efficiency. A new protocol can launch with billions of dollars of stake standing behind it on day one, rather than waiting for its own token to appreciate first. The catch is that the same collateral now backs several promises at once, which is exactly where the argument begins.

How EigenLayer turned staked ETH into rentable security

The concept was formalized by EigenLayer, built by Eigen Labs and its founder Sreeram Kannan. EigenLayer is a set of Ethereum smart contracts that let a staker point stake at extra jobs, and three roles make it run. Restakers deposit ETH or liquid staking tokens and opt in to the added risk. Operators run the software for the services and manage delegated stake. The services themselves, originally called Actively Validated Services, define what they need secured and what counts as misbehavior.

When an operator misbehaves, the restaked collateral behind it can be slashed, and the loss flows back to the restakers who delegated to that operator. In 2025 the project rebranded to EigenCloud and recast itself as verifiable cloud infrastructure, pitched as an AWS for crypto. Its flagship EigenDA sells cheap data availability to rollups, while newer parts, EigenVerify and EigenCompute, handle dispute resolution and off-chain compute. The rebrand arrived with a $70 million EIGEN purchase from a16z, lifting the venture firm’s total backing past $170 million.

Slashing and redistribution: the teeth behind the promise

For most of its life EigenLayer had a conspicuous hole: it promised penalties but could not actually impose them. That changed on 17 April 2025, when slashing went live on mainnet, letting services punish operators who break their rules. CoinDesk described the upgrade as completing the protocol’s original vision.

A second step followed on 22 July 2025 with redistribution. Before it, slashed funds were simply burned. Redistribution lets a service route penalized collateral somewhere useful: to reimburse a victim, cover a bad debt, or pay an insurance claim, which opens restaking to backstops for lending and similar products. It is opt-in, and EIGEN itself was excluded from redistribution at launch.

Slashing is what separates restaking from a plain yield product. Without it, there is no real security being sold; with it, restakers carry genuine downside when the operators they trust get something wrong.

Liquid restaking tokens: the wrapper that took over

Most people never touched EigenLayer directly. They used liquid restaking tokens, or LRTs, which do the restaking for you and hand back a tradeable receipt that keeps earning. Deposit ETH, receive a token, and that token accrues base staking rewards plus restaking rewards while staying liquid enough to trade, lend, or post as collateral.

Ether.fi (eETH) is the largest, with more than $3.2 billion in deposits. Renzo (ezETH), Puffer Finance (pufETH), and Kelp DAO (rsETH) round out the leaders. Total value across liquid restaking platforms climbed to close to $8 billion, according to The Block. During the 2024 points-farming mania, LRTs advertised layered rewards at once: staking yield, restaking yield, EigenLayer points, and each LRT protocol’s own points.

That stacking is also the weakness. An LRT is a claim on a claim on staked ETH, and its peg holds only while redemptions run smoothly. When they do not, the discount can turn violent, as holders would soon learn.

The restaking map: who holds the deposits

Restaking is no longer one protocol. The table below groups the main venues by model and rough deposit size in mid-2026; the figures move daily and are best checked live on DefiLlama.

Two patterns stand out. Ethereum LRTs still hold the bulk of the deposits, and Bitcoin has quietly become the second pillar of the whole sector.

Bitcoin restaking and the flexible challengers

Babylon took restaking to Bitcoin, the one asset larger than ETH and the one that never paid a yield. Instead of wrapping BTC and bridging it, Babylon uses native Bitcoin scripts so holders can lock coins in self-custody and pledge them to secure proof-of-stake chains, forfeiting the coins only if they double-sign. That design pushed Babylon to roughly $3.4 billion in deposits, making it the second-largest restaking network.

On Ethereum, Symbiotic offers a more permissionless take: almost any ERC-20, not just ETH or liquid staking tokens, can back a network, and each network sets its own collateral and rules. It is smaller, near $370 million, but flexible and backed by names close to Lido and Paradigm. Karak, now operating as OpenGDP, spreads multi-asset restaking across several chains yet has stayed a niche worth tens of millions.

The competition matters because it splits the security budget. Every new venue divides operators, auditors, and attention, and each one promises that the same collateral can safely do even more.

Where it breaks: depegs, cascades, and the Kelp DAO hack

Buterin saw the danger early. In a May 2023 essay titled “Don’t overload Ethereum’s consensus”, he warned that stretching Ethereum’s stake and social consensus to backstop outside systems carries high systemic risk. Reusing validators for extra work was acceptable, he argued; dragging Ethereum’s consensus into rescuing failed applications was not.

The first big scare was financial rather than technical. In April 2024, Renzo’s ezETH lost its peg and briefly traded near $688 on Uniswap, a discount of more than 75 percent, after the end of an airdrop farming window triggered a rush for an exit that barely existed. Leveraged loopers were liquidated for tens of millions of dollars on Morpho and Gearbox. The token recovered within hours, but the lesson stuck: an LRT is only as stable as its thinnest pool of exit liquidity.

Then came the technical one. On 19 April 2026, Kelp DAO was drained of about $292 million in the largest DeFi exploit of the year. The attacker abused a one-of-one verifier on Kelp’s LayerZero bridge, minted roughly 116,500 rsETH out of nothing, and used it as collateral on Aave to borrow around $190 million before lenders could react. Aave, SparkLend, and Fluid froze their rsETH markets, and the theft was later tied to North Korea’s Lazarus Group. It was a bridge failure, not a slashing failure, but it struck restaking’s core weakness all the same: a receipt token that dozens of other protocols had wired into their own risk models.

The SEC question: is restaking a securities business?

For years the biggest legal cloud over staking in the United States was the SEC. Under Gary Gensler the agency treated staking-as-a-service as an unregistered securities offering and pushed Kraken to shut its US program in 2023. That posture has since shifted.

On 29 May 2025 the SEC’s Division of Corporation Finance published a staff statement concluding that certain protocol staking activities are not securities transactions. The reasoning: staking that helps run a permissionless network is administrative or ministerial work, not the entrepreneurial effort of others that the Howey test targets. Commissioner Hester Peirce summed it up in a companion note titled “Providing Security is not a ‘Security’.”

Restaking does not clearly fit inside that shelter. The statement covers assets used purely for consensus and warns that anything generating a passive yield, or conveying rights to a business’s profits, falls outside it. An LRT that markets stacked returns and is steered by a team looks much closer to the second description. The guidance is also staff-level and non-binding, and it predates any test case on liquid restaking. Under Chair Paul Atkins the SEC is drafting a broader token taxonomy through its Project Crypto effort, but until that lands, US restakers lean on a narrow reading of a narrow statement.

What to watch in the second half of 2026

Restaking has moved past its hype phase into something more sober. Deposits sit well below the 2024 peak, EIGEN trades near $0.21 against its launch enthusiasm, and the marketing has shifted from points to paying customers for services like EigenDA. The open question is whether shared security has real, recurring demand, or whether it was mostly a farm for token incentives.

Three questions will decide whether shared security becomes lasting infrastructure or fades to a footnote:

• Whether a first real, cascading slashing event tests how losses spread once live slashing and redistribution meet a genuine failure.
• Whether Bitcoin restaking through Babylon keeps growing after its own incentive programs taper.
• Whether US regulators extend their softer staking stance to liquid restaking tokens, or draw a firm line there.

Restaking took the simplest idea in proof-of-stake, reusing collateral, and pushed it about as far as it will go. The next year will show how much weight that collateral can actually bear.

By the HOGE Wire markets desk.