Restaking in 2026: Shared Security, Yield, and Real Risk
Restaking lets staked ETH secure more than one network at once, and it is now one of DeFi's biggest sectors. Here is how it works, who leads, and where the risk sits in 2026.
Staking secures a single blockchain. Restaking tries to make the same staked capital secure several systems at once, turning Ethereum’s validator set into a kind of rentable security layer for other protocols. For supporters, it is the missing piece that lets Ethereum’s vast security budget protect a whole economy of smaller services. For skeptics, it is leverage by another name. The idea went from whitepaper to one of the largest sectors in decentralized finance in under three years, and in 2026 it is also one of the most scrutinized, after a record exploit and a wave of fresh guidance from US regulators.
What restaking actually means
When you stake Ether, you lock it up to help validate the Ethereum network and earn a yield for doing so. Restaking lets you reuse that same staked position to also help secure additional services, from data availability layers to oracle networks and cross-chain bridges. In exchange, you take on extra duties, extra penalties, and extra rewards. The pitch is capital efficiency: one pool of collateral, many jobs, and several income streams instead of one.
The concept was popularized by EigenLayer, which opened its first restaking deposits on Ethereum mainnet in 2023 and now operates under the EigenCloud brand. Within a year it had attracted billions of dollars, proof that stakers wanted more from their collateral than a single yield. Its own documentation frames restaking as a marketplace where stakers supply security and external protocols rent it. That framing has since been copied, forked, and challenged by a growing list of rivals.
How the machinery works
Restaking comes in two broad flavors. Native restaking points an Ethereum validator’s withdrawal credentials at a restaking contract, so the validator’s staked ETH backs both Ethereum and the extra services. Liquid restaking instead takes a liquid staking token, one that already represents staked ETH, and deposits that into the restaking layer. Either way, the collateral now answers to more than one master.
Three roles make the system run:
- Restakers supply the capital and choose how much risk to take.
- Operators run the software that does the actual work for each service; large infrastructure providers, including Google Cloud and Coinbase Cloud, have registered as operators.
- Actively Validated Services (AVSs) are the protocols that buy security, such as data availability layers, oracles, and bridges.
Why would a new protocol want this? Bootstrapping a fresh validator set and token to secure a service is slow and expensive, while renting existing, battle-tested Ethereum collateral is fast. The mechanism that makes the trade credible is slashing: if an operator misbehaves, part of the restaked collateral backing that service can be destroyed. Without slashing, restaking is just a rewards program. With it, restaking becomes real economic security, and real risk for the people supplying the capital. That single design choice, more than any token incentive, is what separates restaking from a points farm.
Liquid restaking tokens and the yield stack
Locking collateral into a restaking contract is capital-inefficient on its own, so the market built a wrapper around it. Liquid restaking tokens (LRTs) represent a restaked position while staying transferable, so holders can lend, trade, or post them as collateral elsewhere while the underlying assets keep earning. eETH from ether.fi, ezETH from Renzo, and rsETH from Kelp DAO (now Kernel DAO) are the best known examples. Together, liquid restaking tokens now hold the bulk of restaked value, because most users prefer a tradeable receipt to a locked position.
This is where the yield stack appears: base Ethereum staking rewards, plus restaking rewards from the services being secured, plus any incentives the LRT issuer layers on, plus whatever the token earns when redeployed in DeFi. Advertised returns have ranged from about 8 percent to 12 percent a year, with brief spikes higher when demand for security runs hot. Every layer in that stack adds a separate point of failure, a tradeoff that is easy to forget when the headline number looks good.
The market in 2026: where the value sits
Restaking grew explosively, then consolidated. At the 2024 peak, EigenLayer alone held more than $15 billion in total value locked. By March 2026 that figure had settled near $8.9 billion, according to DefiLlama, while liquid restaking issuers captured much of the remaining activity. ether.fi has become the largest single LRT issuer, holding the biggest share of that sub-market by a wide margin. The rest is split among a long tail of smaller issuers and the newer base layers trying to copy the model.
The table below sketches the main players in mid-2026. Figures are approximate snapshots from DefiLlama’s restaking dashboard and move daily.
| Protocol | Type | Core token | Approx. TVL (2026) | What sets it apart |
|---|---|---|---|---|
| EigenLayer / EigenCloud | Base restaking layer | EIGEN | ~$8.9 billion | First mover, largest AVS ecosystem |
| ether.fi | Liquid restaking | eETH | ~$5.8 billion | Largest LRT issuer, deep DeFi reach |
| Renzo | Liquid restaking | ezETH | ~$3.3 billion | Strategy-managed exposure |
| Kelp DAO / Kernel | Liquid restaking | rsETH | ~$2 billion (cut after April 2026 hack) | Multi-chain LRT, hit by a major exploit |
| Symbiotic | Base restaking layer | none yet | ~$1.6 billion | Permissionless, any ERC-20 as collateral |
| Karak | Base restaking layer | none yet | early stage | Slashing from day one, broad collateral |
EigenLayer’s slashing era and the great repricing
For its first two years, EigenLayer ran without the one feature that gives restaking its teeth. That changed on April 17, 2025, when the protocol switched on slashing, a step CoinDesk described as completing the project’s original vision. Services had to opt in, but the change finally let misbehaving operators lose collateral.
The market repriced the new risk quickly. Total value locked fell from its $15 billion peak toward roughly $7 billion by late 2025, and the EIGEN token slid with it. By late June 2026, EIGEN traded near $0.25 on CoinGecko, about 95 percent below its all-time high of $5.65. The team kept building through the slump: a July 2025 redistribution feature lets slashed funds be routed back into the ecosystem rather than simply burned, and in December 2025 the EigenLayer foundation said it would steer larger rewards toward active users to defend its lead.
The challengers: Symbiotic and Karak
EigenLayer’s lead pulled in competition. Symbiotic, backed by Paradigm and cyber.Fund, pitches itself as a fully permissionless restaking layer that accepts almost any ERC-20 token as collateral and lets each service configure its own slashing and reward rules. The Block reported a $29 million Series A that took its total funding to about $34.8 million, and the protocol crossed $1 billion in deposits within roughly a month of launch.
Karak took a different angle. It went live on mainnet in January 2025 with slashing enabled from day one, pulled in about $200 million of deposits in its first 24 hours, and accepts a wider menu of collateral, including stablecoins, liquidity-pool tokens, and wrapped Bitcoin. The three approaches map onto three bets: EigenLayer on ecosystem depth, Symbiotic on modular flexibility, and Karak on collateral diversity. None has dislodged EigenLayer yet, but each chips at a different part of its moat.
Where the risk really lives
Restaking concentrates risk in ways plain staking does not. The main exposures stack on top of each other:
- Slashing risk: collateral can be penalized for an operator’s faults, sometimes for events a restaker cannot see or control.
- Smart-contract and bridge risk: every extra contract and cross-chain hop is a new attack surface.
- Rehypothecation and contagion: LRTs flow through lending markets, so trouble in one token can spread to protocols that never touched restaking.
- Depeg risk: an LRT can trade below the value of its backing if confidence or liquidity slips.
- Operator concentration: if a few large operators secure most services, the system inherits their single points of failure.
April 2026 made the abstract concrete. An attacker drained about $292 million from Kelp DAO by exploiting its cross-chain bridge, the year’s largest crypto theft at the time, according to CoinDesk. Roughly 116,500 rsETH, close to 18 percent of supply, was taken or left unbacked. Chainalysis traced the break to a single-point-of-failure verification setup on the bridge and tied the theft to North Korea’s Lazarus Group. The fallout froze rsETH markets on lending venues including Aave and SparkLend, exactly the contagion that critics of restaking had warned about.
What US regulators have said
The regulatory mood in the United States warmed sharply across 2025 and 2026, though not for restaking specifically. In a May 29, 2025 statement, the staff of the Securities and Exchange Commission’s Division of Corporation Finance said that certain protocol staking activities are not offers or sales of securities, a marked shift from the enforcement-first approach of prior years. A follow-up statement on August 5, 2025 extended similar comfort to certain liquid staking arrangements.
The strongest signal came on March 17, 2026, when the SEC and the Commodity Futures Trading Commission issued a joint interpretation on how securities laws apply to crypto assets. It treats staking rewards as payment for validation work and covers self-staking, custodial staking, and liquid staking. Restaking, though, sits outside that comfort zone. Arrangements that involve discretionary management or anything resembling a promised return stay legally uncertain, which keeps the most lucrative corners of restaking in a gray area even as plain staking gets clearer.
What to watch from here
Restaking is past its hype phase and into a harder one: proving that shared security is worth its risks. The questions that will decide its next chapter are concrete. Do AVSs earn enough real fees to pay restakers for the danger they absorb, instead of leaning on token incentives? Can issuers harden the bridges and contracts that the Kelp DAO hack exposed? And will US regulators eventually extend their staking comfort to restaking, or leave it stranded in the gray zone? Institutional money tends to wait for that last answer before it commits at scale.
For now, restaking remains one of the most consequential experiments in crypto: a genuine attempt to let one security budget protect many systems, set against the chance that stacking risk on risk eventually breaks something. Anyone tempted by the yields should size the downside first. None of the above is investment advice.
By the HOGE Wire DeFi desk.